Skip to main content

Surgical Evidence Atlas

How are surgical methods compared in scientific research?

Explore, in plain language, the results of research on hernia, gallbladder, bowel, stomach, pancreas, liver and other areas of surgery.

  • 63 studies
  • ✓ 63 physician-approved (39 conditional)

The results of these studies may not apply to every patient. Treatment decisions should be based on an individual assessment.

What do the approval badges mean?

✓ Physician-approved: A physician has reviewed the summary together with its sources and found it suitable for publication.

Conditionally approved: These summaries have also been approved by a physician. “Conditional” does not mean the study failed or is unreliable; it is this site’s publication condition. Because some source checks (for example, the full text of the paper) are not yet complete, or because the study has important limitations, those limitations are kept clearly visible on the card. Each card explains this in its “Why conditional approval?” box.

Find a study

Showing 63 of 63 studies

✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

STITCH

Can the way an abdominal incision is closed change the risk of a hernia?

Patients who developed a hernia at the incision within one year (per 100 patients, approximately)
Time point: 1 yearOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With small, closely spaced stitches, about 8 fewer hernias per 100 patients.

The difference was statistically significant. Applies to planned midline abdominal surgery.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

A hernia can develop at the incision after open surgery through the middle of the abdomen. The study asked whether closing the incision with smaller, more closely spaced stitches (small-bites closure) reduces this hernia compared with larger, wider-spaced stitches.

What did it find?

Within one year, an incisional hernia occurred in about 13 in 100 patients with small-bites closure and about 21 in 100 with large-bites closure; this difference was statistically significant. The advantage of small-bites closure persisted with longer follow-up.

What did longer follow-up show? The advantage of closure with small, closely spaced stitches persisted in longer follow-up: the estimated chance of an incisional hernia by year 13 was lower in this group (about 34 vs 49 in 100, estimated). These results apply to planned (non-emergency) open operations through the middle of the abdomen and cannot be generalised to every abdominal operation.

What does it mean for me?

In planned midline abdominal surgery, the way the incision is closed can affect the chance of a later hernia.

Which patients does it apply to?

Patients having a planned (non-emergency) open operation through the middle of the abdomen. It does not automatically apply to emergency or contaminated (infected) abdominal surgery.

Show scientific details

SURG-STITCH

Full scientific title

STITCH

Clinical question

In elective midline laparotomy, do small-bite closures reduce 1-year incisional hernia versus large bites?

Themes

Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Incisional hernia at 1 year

Primary endpoint met?

Met

Clinical interpretation

Benefit

Population

Elective midline laparotomy closure; 560 randomized (small bites 276, large bites 284).

Intervention

Small-bite closure

Comparator

Large-bite closure

Results

In analyzed patients, small bites 35/268 (13%), large bites 57/277 (21%). Adjusted OR 0.52 (95% CI 0.31–0.87).

Long-term update

2026 long-term follow-up: Long-term assessment of the original STITCH randomized population was published in JAMA Surgery (van den Berg et al.). At 13 years, estimated cumulative incisional hernia incidence was 34% with small bites vs 49% with large bites (HR 0.61; 95% CI 0.43–0.86). Cumulative incidence of hernias wider than 20 mm was 17% vs 34%. Hernia repair rates were similar. Median censoring time was 8 years; incomplete follow-up and competing-risk analysis should be considered. This record is not the 2017 BJS Open STITCH-related publication.

Sample size

560

Registry

NCT01132209

PMID

26188742

DOI

10.1016/S0140-6736(15)60459-7

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Primary: PubMed abstract (Deerenberg Lancet 2015, PMID 26188742). Long-term: PubMed abstract + Conflict of Interest Disclosures (van den Berg JAMA Surg 2026, PMID 41920572). The related 2017 BJS Open STITCH publication was not confused with this long-term record.

Funding

Erasmus University Medical Center and Ethicon (as reported in the PubMed abstract — 2015 primary publication).

Conflicts of interest

2026 long-term publication (JAMA Surgery, PMID 41920572) Conflict of Interest Disclosures: Dr Deerenberg reported support from Ethicon and Johnson & Johnson during the conduct of the study, and personal fees from Medtronic and Gore outside the submitted work. No additional disclosures were reported for the other authors. This line is not the conflict-of-interest statement of all authors of the original 2015 STITCH article.

Limits

Studied elective midline closure; findings do not automatically apply to all emergency or contaminated abdominal closures.
✓ Physician-approved · 9 Nov 2026Gallbladder

CHOCOLATE

For high-risk gallbladder inflammation: surgery or a drain?

Patients with a major complication within one year (per 100 patients, approximately)
Time point: 1 yearOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With keyhole gallbladder removal, about 53 fewer major complications per 100 patients.

The difference was statistically significant. The difference in death was not significant. The trial was stopped early after an interim analysis.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For patients with gallbladder inflammation (acute cholecystitis) who are high-risk because of other illnesses, the study asked whether removing the gallbladder by keyhole surgery (laparoscopic cholecystectomy) is better than placing a drain into the gallbladder through the skin (percutaneous drainage).

What did it find?

Within one year, major complications occurred in about 12 in 100 patients after surgery and about 65 in 100 after drainage. Death occurred in about 3 in 100 after surgery and about 9 in 100 after drainage, but this difference was not statistically significant. The trial was stopped early after a planned interim analysis.

What does it mean for me?

In high-risk patients who can tolerate surgery, removing the gallbladder was associated with fewer major complications than drainage. A survival advantage was not proven.

Which patients does it apply to?

Patients with gallstone-related acute cholecystitis and an intensive-care illness-severity score (APACHE II) of 7–14. Patients with a score of 15 or more, too ill to tolerate surgery, were not included.

Show scientific details

SURG-CHOCOLATE

Full scientific title

CHOCOLATE

Clinical question

In high-risk acute calculous cholecystitis, does laparoscopic cholecystectomy reduce 1-year major complications versus percutaneous drainage?

Themes

Timing and strategy · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Separate primary endpoints: major complications at 1 year; death at 1 year

Study outcomes (by outcome)

  • Major complications

    Time point: 1 year

    Result: Cholecystectomy 8/66 (12%) vs drainage 44/68 (65%); RR 0.19 (95% CI 0.10–0.37).

    Statistical interpretation: Superiority shown for major complications.

    Endpoint met?: Met

  • Death (mortality)

    Time point: 1 year

    Result: Cholecystectomy 3% vs drainage 9%; p=0.27.

    Statistical interpretation: No statistically significant mortality difference; mortality superiority not shown.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Acute calculous cholecystitis with APACHE II ≥7; APACHE II ≥15 excluded. 142 randomized; reported comparison cholecystectomy n=66 vs drainage n=68; stopped early after interim analysis.

Intervention

Laparoscopic cholecystectomy

Comparator

Percutaneous drainage

Results

Death and major complications were reported as separate primary endpoints. Major complications: 12% vs 65%; RR 0.19 (95% CI 0.10–0.37). One-year mortality: 3% vs 9%; p=0.27. The trial was stopped early after a planned interim analysis. Patients with APACHE II ≥15 were excluded.

Sample size

142

Registry

NTR2666

PMID

30297544

DOI

10.1136/bmj.k3965

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

BMJ open-access full text + PubMed abstract (PMID 30297544). Funding and ICMJE interest statement from the BMJ article page (https://www.bmj.com/content/363/bmj.k3965).

Funding

St Antonius Research Foundation. According to the article, the funder had no role in study design, data collection, analysis, interpretation or writing of the report. The first author received support from the Dutch Endoscopic Surgery Society for the thesis publication.

Conflicts of interest

According to the ICMJE statement in the BMJ article, the authors declared no relevant financial or other interests (https://www.bmj.com/content/363/bmj.k3965).

Limits

Not generalizable to patients unfit for surgery; early termination should be noted. APACHE II ≥15 was excluded.
✓ Physician-approved · 9 Nov 2026Pancreatitis and HPB

POINTER

In infected pancreatitis, should drainage be done immediately or postponed?

Key finding

Immediate drainage did not give better results than waiting and required more procedures; some patients who waited recovered with antibiotics alone.

Why was this study done?

When dead pancreatic tissue becomes infected after pancreatitis (infected necrotising pancreatitis), a drain is usually placed. The study asked whether placing the drain as soon as infection is recognised is better than waiting until the dead tissue walls off and draining later.

What did it find?

At six months, a score reflecting the total burden of complications was similar in both groups; immediate drainage was not shown to be better. More procedures were done in the immediate-drainage group (on average about 4.4 vs 2.6). About 4 in 10 patients in the postponed group were treated with antibiotics alone without needing a drain. In longer follow-up, death or major complications were similar in both groups.

What does it mean for me?

In selected stable patients, postponing drainage may lead to fewer procedures. This does not mean waiting is appropriate for patients who are deteriorating or need urgent intervention.

Which patients does it apply to?

Selected patients with infected necrotising pancreatitis who are clinically stable and in whom intervention can be postponed.

Show scientific details

SURG-POINTER

Full scientific title

POINTER

Clinical question

In infected necrotizing pancreatitis, does immediate drainage reduce 6-month complication burden (CCI) versus postponed intervention?

Themes

Timing and strategy

Study design

Pragmatic randomized controlled trial

Primary endpoint

Comprehensive Complication Index (CCI) at 6 months

Primary endpoint met?

Not met

Clinical interpretation

Practice-limiting (negative/neutral landmark)

Population

Infected necrotizing pancreatitis; 104 patients (immediate drainage n=55, postponed n=49).

Intervention

Immediate drainage

Comparator

Postponed intervention

Results

Mean CCI 57 vs 58; difference −1 (95% CI −12 to 10), p=0.90. Superiority of immediate drainage not shown. Mean interventions 4.4 vs 2.6. In the postponed arm, 19/49 (39%) were treated with antibiotics without drainage (17 survived).

Long-term update

In the 2024 long-term follow-up study, 88 of 104 patients were reassessed at a median of 51 months. After the initial 6 months, the composite of death and major complications occurred in 7/47 (15%) with immediate drainage vs 7/41 (17%) with postponed drainage (RR 0.87; 95% CI 0.33–2.28; p=0.78). Over total follow-up, median interventions were 4 with immediate vs 1 with postponed drainage (p=0.001). Among 15 selected patients successfully treated with antibiotics alone in the first 6 months, 14 remained intervention-free long term. This ratio must not be generalized to the entire randomized population.

Sample size

104

Registry

ISRCTN33682933

PMID

34614330

DOI

10.1056/NEJMoa2100826

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Primary: NEJM full-text page + PubMed (PMID 34614330). Long-term: PubMed abstract (Van Veldhuisen Ann Surg 2024, PMID 37450701) — composite outcome of death/major complications and numbers of interventions.

Funding

Fonds NutsOhra and Amsterdam UMC (source materials).

Limits

The 14/15 figure must not be generalized. It does not justify delaying intervention when there is clinical deterioration, failure to control sepsis, or need for urgent intervention.
✓ Physician-approved · 9 Nov 2026Pancreatic and periampullary tumours

DROP

With jaundice from a pancreatic head tumour, is bile duct drainage needed before surgery?

Patients with a serious complication within 120 days (per 100 patients, approximately)
Time point: 120 days (from randomization)Orange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With early surgery, about 35 fewer serious complications per 100 patients.

The difference was statistically significant; routine drainage was linked to more serious complications.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

Cancer of the head of the pancreas can block the bile duct and cause jaundice. The study asked whether routinely placing a stent in the bile duct by endoscopy before surgery (biliary drainage by ERCP, endoscopic retrograde cholangiopancreatography) is needed, compared with going straight to surgery within about a week.

What did it find?

Within 120 days, serious complications occurred in about 39 in 100 patients in the early-surgery group and about 74 in 100 in the preoperative-drainage group.

What does it mean for me?

Under the conditions of this study, routine preoperative stent drainage was associated with more serious complications than early surgery. Bile duct infection (cholangitis), delayed surgery and chemotherapy before surgery are outside the scope of this result. Preoperative drainage with the metal stents used today has been assessed in separate studies.

Which patients does it apply to?

Patients with removable cancer of the head of the pancreas and a blood bilirubin level of 2.3–14.6 mg/dL.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-DROP

Full scientific title

DROP

Clinical question

In resectable pancreatic head cancer, does routine preoperative biliary drainage (ERCP endoprosthesis) increase serious complications within 120 days versus early surgery?

Themes

Timing and strategy · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Serious complications within 120 days from randomization

Clinical interpretation

Practice-limiting (negative/neutral landmark)

Population

Resectable pancreatic head cancer with bilirubin 2.3–14.6 mg/dL. 202 randomized; early surgery 96, drainage 106 (six excluded from analysis).

Intervention

4–6 weeks of preoperative drainage, attempted primarily by ERCP endoprosthesis (stent) placement. Stent type (plastic/metal) is not stated in the abstract; not verified against the full text.

Comparator

Early surgery (within about 1 week)

Results

The trial found more serious complications in the routine preoperative biliary drainage arm than in the early-surgery arm. Serious complications: early surgery 39%, drainage 74%; RR 0.54 favoring early surgery (95% CI 0.41–0.71), p<0.001.

Sample size

202

Registry

ISRCTN31939699

PMID

20071702

DOI

10.1056/NEJMoa0903230

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

NEJM full text + PubMed (PMID 20071702); plastic-stent era. The recorded “primary endpoint met” status was not changed without a review of the hypothesis; the single “Met” badge is hidden on the card. 2026-11-09: the NEJM full text could not be accessed in this round (403); because the abstract describes the stent only as an “endoprosthesis”, the “plastic stent” description is not used as a definite statement in the card texts (the physician should confirm it from the full text).

Limits

Findings do not cover all settings with cholangitis, delayed surgery, neoadjuvant therapy, or modern SEMS use. SEMS-2026 is a separate record; rates must not be presented as a head-to-head superiority test. Funding/COI absent from source pack (left empty). Unresolved source detail (not verified from the primary source; not a scientific conflict): the exact type of stent used could not be verified from an accessible primary source (PubMed abstract: “endoprosthesis”).

Related separate studies

✓ Physician-approved · 9 Nov 2026Pancreatic and periampullary tumours

SEMS-2026 (preoperative metal stent vs early surgery)

With a tumour of the pancreas or nearby area, is a metal stent before surgery an acceptable alternative to early surgery?

Patients with at least one serious problem within 120 days (per 100 patients, approximately)
Time point: 120 daysOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Metal stent: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior. In the period before surgery, serious problems and deaths were more common in the metal-stent group.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

When cancer of the pancreas or nearby area blocks the bile duct, a self-expanding metal stent (SEMS) can be placed by endoscopy before surgery. The study asked whether preoperative drainage with a metal stent increases serious problems unacceptably compared with early surgery.

What did it find?

Within 120 days, at least one serious problem occurred in about 29 in 100 patients in the metal-stent group and about 27 in 100 in the early-surgery group. The prespecified acceptance margin (15 percentage points) was met, meaning the metal stent was not found to be unacceptably worse. However, in the period before surgery serious problems were more common in the metal-stent group (about 15 vs 5 in 100), and before surgery 7 of 144 patients in the metal-stent group and 1 of 140 in the early-surgery group died. Including these preoperative deaths, total death over the whole 120-day follow-up was similar in both groups (11 patients in each group; about 8 in 100).

What does it mean for me?

In selected patients, preoperative drainage with a metal stent may be a reasonable option, but the risks during the wait before surgery should be weighed as well. This does not mean drainage is better or that every patient needs routine drainage; the acceptance margin was also relatively wide.

Which patients does it apply to?

Patients with removable cancer of the pancreas or surrounding (periampullary) area, a blood bilirubin of 5.8 mg/dL or more, in whom surgery first is planned.

Show scientific details

SURG-SEMS-2026

Full scientific title

SEMS-2026 (preoperative metal stent vs early surgery)

Clinical question

In resectable pancreatic/periampullary cancer, is preoperative SEMS drainage non-inferior to early surgery for serious adverse events at 120 days?

Themes

Technology comparison · Timing and strategy

Study design

Non-inferiority randomized controlled trial

Primary endpoint

At least one serious adverse event within 120 days after randomization

Primary endpoint met?

Met

Clinical interpretation

Non-inferiority

Population

Resectable pancreatic/periampullary cancer, bilirubin ≥5.8 mg/dL, resection primarily planned. 11 centers in 9 countries; 284 randomized (SEMS 144, early surgery 140).

Intervention

Preoperative biliary drainage — WallFlex Biliary RX (Boston Scientific); covered or uncovered SEMS per protocol

Comparator

Early surgery

Results

Analysis population: modified ITT (SEMS 138, early surgery 136). SEMS: 40/138 (29.0%). Early surgery: 36/136 (26.5%). Risk difference: 2.5 percentage points. One-sided 95% upper confidence limit: 11.7%. Non-inferiority: p=0.011. Prespecified non-inferiority margin: 15% (full text; upper CL below the margin). Non-inferiority criterion met; this does not imply superiority. Preoperative period (full-text Table 2; order SEMS/early surgery; from randomization to surgery, within the 120-day window): at least one protocol-specified serious adverse event — modified-ITT denominator — 21/138 (15.2%) vs 7/136 (5.1%), difference 10.1 percentage points (exact 95% CI 2.3–17.9); preoperative death — randomized denominator — 7/144 (4.9%) vs 1/140 (0.7%), difference 4.2 (95% CI 0.3–9.2). Operative and postoperative protocol-specified serious adverse events 23/119 (19.3%) vs 30/130 (23.1%). Death — study follow-up was 120 days after randomization (up to day 150 where needed); there are no separate longer-term mortality data. 120-day all-cause mortality (full text, randomized denominator): 11/144 (7.6%) vs 11/140 (7.9%), difference −0.3 (95% CI −7.0 to 6.4). The abstract’s “during follow-up” mortality gives the same 11 vs 11 deaths with modified-ITT denominators: 11/138 (7.9%) vs 11/136 (8.0%); it is not a separate time point. Of the deaths, 8 were preoperative (7 vs 1) and 14 postoperative (4 vs 10); preoperative deaths are included in the 120-day total, not an additional result.

Sample size

284

Registry

NCT01774019

PMID

41951900

DOI

10.1245/s10434-026-19546-9

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Peer-reviewed full text (Europe PMC / PMC13242453) + PubMed. The non-inferiority margin of 15% was verified from the Statistical Analysis / Results sections of the full text. Funding and Disclosure from the full text. 2026-11-09: the denominators for preoperative serious adverse events/death and 120-day death were verified from full-text Table 2 and the “Deaths within 120 Days” section.

Funding

Boston Scientific Corporation. The sponsor took part, together with the academic investigators, in study design, conduct, the central database, monitoring, statistical analysis, interpretation and writing of the article (Trial Design and Oversight / Funding — full text).

Conflicts of interest

Some authors have research, consulting or board relationships with Boston Scientific/Olympus/Cook; J. Peetermans was a Boston Scientific employee during the study; M. Rousseau is a Boston Scientific employee; J.Y.W. Lau is a Boston Scientific consultant. For the full author-level statement, see the article's Disclosure section: https://pmc.ncbi.nlm.nih.gov/articles/PMC13242453/ and https://doi.org/10.1245/s10434-026-19546-9

Limits

DROP used bilirubin 2.3–14.6 mg/dL while this trial’s inclusion threshold was ≥5.8 mg/dL; stent technology and design differ. Rates from the two trials must not be presented as a direct head-to-head superiority test. The discussion notes the 15% margin was wide for pragmatic reasons. Although non-inferiority was met for the primary endpoint, preoperative serious adverse events and preoperative death were more frequent in the SEMS arm; these are secondary/subgroup analyses and must be interpreted separately from the primary result. Supplementary tables were not reviewed in this pack.

Related separate studies

✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

Luijendijk

Does using a mesh to repair an incisional hernia reduce the chance it comes back?

Key finding

According to estimates calculated from follow-up data, the chance of the hernia coming back was clearly lower after mesh repair than after stitches alone, and the difference persisted with longer follow-up.

Why was this study done?

For patients who developed a hernia at the scar of an earlier abdominal operation (incisional hernia), the question was whether repair with stitches alone or with a mesh is better. This is a repair study, not a prevention study.

What did it find?

For hernias repaired for the first time, the estimated chance of the hernia coming back (recurrence) within three years, calculated from follow-up data, was about 43 in 100 with stitches alone and about 24 in 100 with mesh. These figures are calculated estimates, not directly counted numbers of patients. The difference in favour of mesh persisted with longer follow-up.

What does it mean for me?

In incisional hernia repair, an appropriate mesh can lower the chance of the hernia coming back compared with stitches alone. Mesh choice depends on the hernia and the patient.

Which patients does it apply to?

Patients with a midline incisional hernia smaller than 6 cm in width or length. It is not definitively generalizable to other hernia types.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-LUIJENDIJK

Full scientific title

Luijendijk — mesh vs suture for incisional hernia

Clinical question

In midline incisional hernia, does mesh repair reduce 3-year cumulative recurrence versus suture repair?

Themes

Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Hernia recurrence (3-year cumulative follow-up)

Primary endpoint met?

Met

Clinical interpretation

Benefit

Population

200 randomized; 181 eligible (154 primary, 27 first recurrence); midline incisional hernia; defect width or length <6 cm.

Intervention

Mesh repair

Comparator

Suture repair

Results

In primary incisional hernia, 3-year cumulative recurrence: suture 43%, mesh 24% (p=0.02). First recurrence: 58% vs 20% (p=0.10; subgroup difference not shown). Percentages are cumulative life-table estimates and must not be presented as crude observed event rates. No single rate for the combined eligible population (primary + first recurrence) is given in the PubMed abstract; the 46% vs 23% (p=0.005) figure previously on this card could not be verified against the full text and is not used as a result (not verified from the primary source).

Long-term update

Burger 2004 (same randomized cohort; not a separate primary RCT): 126 of 181 eligible patients completed long-term follow-up. Ten-year cumulative recurrence: suture 63%, mesh 32% (p<0.001). Median follow-up among completers 98 months. Complications 17% mesh vs 8% suture (p=0.17); abdominal pain more frequent after suture.

Sample size

200

PMID

10933738

DOI

10.1056/NEJM200008103430603

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract / accessible text (PMID 10933738, 15383785). Funding/COI not verified from the full text. Kaplan–Meier cumulative estimates were not presented as observed crude event rates. 2026-11-09: the patient text was corrected to ‘calculated estimate’; the NEJM full text could not be accessed (403), and the combined 46%/23% value was removed from the results.

Limits

Repair of an existing midline incisional hernia; do not confuse with prophylactic mesh (PRIMA). Period-specific technique/material context; not absolute generalization to every hernia type. Surgeon-experience field absent from the source pack. Funding and author COI unverified (empty).
✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

LEVEL

Does keyhole groin hernia repair give an easier recovery than open repair?

Key finding

Keyhole repair meant less pain in the first weeks and a faster return to daily life; there was no significant difference in long-lasting pain or recurrence.

Why was this study done?

A groin (inguinal) hernia can be repaired by keyhole (laparoscopic) surgery placing a mesh in front of the abdominal lining (totally extraperitoneal repair, TEP) or by open Lichtenstein surgery through the groin. The study compared the two for recovery, pain and the hernia coming back (recurrence).

What did it find?

After TEP, pain in the first weeks was lower and return to daily life was faster. However, problems during the operation occurred in about 6 in 100 patients with TEP and about 2 in 100 with open surgery. No significant difference was found in long-lasting pain (about 25 vs 29 in 100) or recurrence (about 4 vs 3 in 100).

What does it mean for me?

Keyhole repair may give faster recovery for some patients, but long-term results were similar and surgeon experience matters.

Which patients does it apply to?

Patients over 18 having planned groin hernia surgery; first-time, recurrent and two-sided hernias were included. Patients whose hernia extended into the scrotum, and those who had previous prostate surgery, a bikini-line incision (Pfannenstiel incision) or surgery in the area in front of the abdominal lining, were not included. These criteria come from the trial’s registry record and could not be verified against the full publication.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-LEVEL

Full scientific title

LEVEL — TEP vs open Lichtenstein

Clinical question

In inguinal hernia, how does TEP compare with open Lichtenstein for early pain, recovery, and quality of life, and what about recurrence and chronic pain?

Themes

Minimally invasive · Surgical technique · Technology comparison · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Multiple primary outcomes: postoperative pain, hospital stay, time to full recovery, quality of life (not collapsed to one flag)

Study outcomes (by outcome)

  • Early postoperative pain

    Time point: Up to 6 weeks

    Result: Less early pain with TEP (p=0.01); faster return to daily life and work.

    Statistical interpretation: Early pain and recovery favored TEP.

    Endpoint met?: Met

  • Intraoperative adverse events

    Time point: Intraoperative

    Result: TEP 5.8% vs open 1.6% (p<0.004).

    Statistical interpretation: Intraoperative events were higher with TEP.

    Endpoint met?: Met

  • Chronic pain

    Result: 25% vs 29%; no significant difference reported.

    Statistical interpretation: No open superiority shown; this is not the same as equivalence.

    Endpoint met?: Not met

  • Recurrence

    Time point: Mean 49 months follow-up

    Result: 3.8% vs 3.0% (p=0.64).

    Statistical interpretation: No significant recurrence difference shown.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

660 randomized; TEP 336, Lichtenstein 324. Selection criteria could not be verified against the publication’s full text (not in the PubMed abstract). Per the ClinicalTrials.gov record (NCT00788554; trial started 2000, registered 2008 — retrospective registration), inclusion: age >18, elective repair, primary, recurrent or bilateral inguinal hernia, informed consent. Exclusion: scrotal hernia; history of prostatectomy, Pfannenstiel incision or preperitoneal surgery; participation in another trial; pregnancy; communicative or cognitive restrictions. Registry criteria could not be compared with the final published criteria.

Intervention

Total extraperitoneal (TEP) repair

Comparator

Open Lichtenstein repair

Results

TEP had less early pain up to 6 weeks (p=0.01) and faster return to daily life and work. Mean operative time for unilateral repair: TEP 54 min vs open 49 min (p=0.03). Intraoperative adverse events: TEP 5.8% vs open 1.6% (p<0.004). Chronic pain 25% vs 29% (no significant difference reported). Recurrence at mean 49 months: 3.8% vs 3.0% (p=0.64). Postoperative complications 33% vs 33%. TEP is recommended in experienced hands.

Sample size

660

Registry

NCT00788554

PMID

20395851

DOI

10.1097/SLA.0b013e3181d96c32

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract (PMID 20395851, NCT00788554). Funding/COI awaiting full-text verification. 2026-11-09: eligibility criteria were taken from the ClinicalTrials.gov record (retrospectively registered); the Ann Surg full text could not be accessed and was not compared with the published criteria. Source author name checked against PubMed: Langeveld.

Limits

Early pain benefit must be read together with higher intraoperative adverse events. Funding/COI full-text verification pending (left empty). Unresolved source detail (not verified from the primary source; not a scientific conflict): patient selection criteria were taken from the retrospective ClinicalTrials.gov record and could not be verified against the publication’s full text.
Physician-approved (conditional) · 14 Nov 2026Hernia and abdominal wall

Fitzgibbons

Is watchful waiting safe for a groin hernia that causes few symptoms?

Why conditional approval?

Many patients crossed over between groups (watched patients who had surgery, and surgery-group patients who did not); the hypothesis the trial tested could not be verified from the source.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: The hypothesis type (superiority / non-inferiority) and threshold could not be verified; crossover between groups was high in both directions.

Key finding

At two years there was no significant difference in pain between waiting and surgery; however, some men who waited had surgery over time and hernia trapping was rare but did occur.

Why was this study done?

For men with a groin hernia causing few symptoms, the question was whether regular follow-up (watchful waiting) is a safe alternative to immediate surgery.

What did it find?

At two years, pain limiting daily activity occurred in about 5 in 100 men with watchful waiting and about 2 in 100 with surgery; this difference was not statistically significant. About a quarter of the watchful-waiting group crossed over to surgery within two years, and about one in six men assigned to surgery stayed under watchful waiting without an operation. Results were compared according to the group men were originally assigned to. Trapping of the hernia (an emergency) was rare but did occur. With longer follow-up, most men in the watchful-waiting group eventually had surgery.

What does it mean for me?

Some men with few symptoms can choose follow-up instead of immediate surgery, but pain may increase, surgery may still be needed, and rare emergencies can occur.

Which patients does it apply to?

Men whose groin hernia causes no or minimal symptoms. It does not apply to women, to people with significant symptoms or to thigh (femoral) hernias.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-FITZGIBBONS

Full scientific title

Fitzgibbons — watchful waiting vs repair in minimally symptomatic inguinal hernia

Clinical question

In minimally symptomatic men with inguinal hernia, is watchful waiting an appropriate option versus open repair for activity-limiting pain and physical function at 2 years?

Themes

De-escalation · Timing and strategy · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Separate primary endpoints: activity-limiting pain/discomfort at 2 years; change in SF-36 physical component score

Study outcomes (by outcome)

  • Activity-limiting pain/discomfort

    Time point: 2 years

    Result: Watchful waiting 5.1% vs surgery 2.2% (p=0.06; value corrected by a 2006 erratum). Intention-to-treat analysis.

    Statistical interpretation: No statistically significant difference shown. Hypothesis type (superiority/equivalence/non-inferiority) and any prespecified margin not verified from the full text; absence of a difference is not presented as proof of equivalence.

    Endpoint met?: Not met

  • Change in SF-36 physical component score

    Time point: 2 years

    Result: 0.29 vs 0.13 (p=0.79).

    Statistical interpretation: No significant difference shown.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Five centers; men only; asymptomatic or minimally symptomatic inguinal hernia; 720 patients (watchful waiting 364, open repair 356).

Intervention

Watchful waiting

Comparator

Open surgical repair

Results

Primary intention-to-treat analysis (patients analyzed as randomized). Activity-limiting pain: watchful waiting 5.1% vs surgery 2.2% (p=0.06; erratum-corrected value). SF-36 physical score change 0.29 vs 0.13 (p=0.79). Crossover in both directions: 23% assigned to watchful waiting had surgery (most often for increased hernia-related pain); 17% assigned to repair remained under watchful waiting. One acute incarceration at 2 years (0.3%, without strangulation); one further acute incarceration with bowel obstruction at 4 years; rate including follow-up up to 4.5 years 1.8/1000 patient-years. Risk is not zero despite low counts.

Long-term update

Fitzgibbons 2013 (long-term follow-up of the same trial; not a separate primary card): 254 men from the original waiting arm. At up to 11.5 years, Kaplan–Meier estimated cumulative crossover to surgery was 68% (79% if >65 years, 62% if younger). Three patients needed emergency surgery; no deaths reported in this follow-up. The cumulative crossover estimate must not be presented as a crude rate for the entire waiting arm.

Sample size

720

Registry

NCT00263250

PMID

16418463

DOI

10.1001/jama.295.3.285

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the JAMA 2006 primary paper and the long-term paper; ClinicalTrials.gov record. The p=0.06 value follows the JAMA 2006 erratum. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Watchful waiting is supported only for selected asymptomatic or minimally symptomatic men; do not generalize to women, clearly symptomatic hernias, or femoral hernias. Two-way crossover (23% and 17%) dilutes the intention-to-treat comparison; results compare a waiting strategy with a repair strategy. Hypothesis type/prespecified margin could not be verified from the primary source. Funding/COI full-text verification pending (left empty — could not be verified from the primary source).
✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

PRIMA

Does a preventive mesh during abdominal surgery reduce incisional hernia?

Patients who developed an incisional hernia within two years (per 100 patients, approximately)
Time point: 2 yearsOrange dot: a patient who had this outcomeGrey dot: a patient who did not

The difference for mesh on top of the muscle was clearly significant; for mesh behind the muscle it was borderline. Fluid collection under the wound was more common with mesh on top of the muscle.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For patients at high risk of an incisional hernia after abdominal surgery, the study asked whether placing a preventive mesh when closing the incision reduces hernias. This is a prevention study, not repair of an existing hernia. The mesh was placed either on top of the muscle layer (onlay) or behind it (sublay).

What did it find?

At two years, an incisional hernia occurred in about 30 in 100 patients with stitches alone, about 13 in 100 with onlay mesh and about 18 in 100 with sublay mesh. The onlay difference was clearly significant; the sublay difference was borderline. At five years, both mesh positions were associated with fewer hernias than stitches alone. Fluid collection under the wound (seroma) was more common with onlay mesh.

What does it mean for me?

In high-risk patients, preventive mesh can reduce later hernias, although fluid collections may be more common with some positions.

Which patients does it apply to?

Patients over 18 having planned open surgery through the middle of the abdomen who have a widened abdominal aorta (aneurysm) or a body mass index of 27 or more.

Show scientific details

SURG-PRIMA

Full scientific title

PRIMA — prophylactic mesh in high-risk midline laparotomy

Clinical question

In high-risk elective midline laparotomy, does prophylactic onlay or sublay mesh reduce 2-year incisional hernia versus primary suture alone?

Themes

Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Incisional hernia within 2 years

Study outcomes (by outcome)

  • Incisional hernia — onlay vs suture

    Time point: 2 years

    Result: Suture 33/107 (30%), onlay 25/188 (13%); adjusted OR 0.37 (95% CI 0.20–0.69; p=0.0016).

    Statistical interpretation: Fewer hernias at 2 years with onlay.

    Endpoint met?: Met

  • Incisional hernia — sublay vs suture

    Time point: 2 years

    Result: Sublay 34/185 (18%); OR 0.55 (95% CI 0.30–1.00; p=0.05) — borderline; not presented as definite benefit.

    Statistical interpretation: Borderline result; no claim of definite benefit.

    Endpoint met?: Unknown / unverified

  • Seroma

    Result: Onlay 34/188; suture 5/107; sublay 13/185 — more frequent with onlay.

    Statistical interpretation: Mesh benefit is shown together with the seroma disadvantage.

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

11 centers; age ≥18; elective midline laparotomy; AAA or BMI≥27. 498 enrolled, 18 excluded before randomization; 480 randomized: suture 107, onlay 188, sublay 185.

Intervention

Prophylactic onlay or sublay mesh (prevention; not repair of an existing hernia)

Comparator

Primary suture only

Results

At 2 years: suture 33/107 (30%), onlay 25/188 (13%), sublay 34/185 (18%). Onlay vs suture adjusted OR 0.37 (95% CI 0.20–0.69; p=0.0016). Sublay vs suture OR 0.55 (95% CI 0.30–1.00; p=0.05) — borderline. Seroma more frequent with onlay (34/188 vs suture 5/107, sublay 13/185). No significant wound-infection difference reported. Clinical interpretation: at 2 years onlay shows clear benefit versus suture; sublay is borderline; at 5 years both mesh positions associate with lower incisional-hernia risk versus suture. No direct superiority claim between onlay and sublay.

Long-term update

PRIMA 2024 long-term (same cohort; not a separate primary card): 5-year cumulative estimated incisional hernia suture 53.4% (95% CI 40.4–64.8), onlay 24.7% (12.7–38.8), sublay 29.8% (17.9–42.6). Versus suture: onlay HR 0.390 (0.248–0.614; p<0.001), sublay HR 0.485 (0.309–0.761; p=0.002). Cumulative estimates are not crude event rates. Funding (2024): B. Braun — shown separately from 2017 primary funding.

Sample size

480

Registry

NCT00761475

PMID

28641875

DOI

10.1016/S0140-6736(17)31332-6

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed / additional correction (PMID 28641875, 38188275, NCT00761475). 2017 funding: Baxter + B. Braun Surgical SA. 2024: B. Braun separately. Author COI not verified. Five-year results are cumulative estimates. No claim of direct onlay–sublay superiority.

Funding

2017 Lancet primary publication: Baxter and B. Braun Surgical SA. 2024 long-term publication: B. Braun funding separately. Author-level conflicts of interest (both publications) were not verified in this review — COI is left empty on the card.

Limits

High-risk elective midline laparotomy (AAA or BMI≥27). No direct onlay-versus-sublay superiority claim. Surgeon-experience field absent from the source pack. Author COI unverified (empty).
Physician-approved (conditional) · 14 Nov 2026Hernia and abdominal wall

Neumayer VA

Is keyhole mesh repair of a groin hernia as durable as open mesh repair?

Why conditional approval?

This is an older trial (US Veterans Affairs hospitals); the result of keyhole surgery was found to depend on the surgeon’s experience.

Verified source: The primary publication was verified at abstract level; the secondary analysis in PMC was also reviewed.

Conditions kept on this card: Historical VA setting; the laparoscopic result depends on surgeon experience (separate secondary analysis).

Patients whose hernia came back within two years (per 100 patients, approximately)
Time point: 2 yearsOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With the keyhole method, the hernia came back in about 5 more people per 100.

The difference was statistically significant and was seen in first-time hernias; no clear difference was reported for re-repair of hernias that had come back. Keyhole repair meant less early pain and a slightly earlier return to daily activities.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

The aim was to compare groin hernia repair with mesh by keyhole (laparoscopic) surgery versus open surgery with mesh.

What did it find?

At two years, the hernia came back (recurrence) in about 10 in 100 patients after laparoscopic repair and about 5 in 100 after open repair. This difference was seen in first-time (primary) hernias (about 10 vs 4 in 100 patients); for re-repair of hernias that had already come back, no clear difference between the groups was reported. In the laparoscopic group early pain was lower and return to usual activities was about one day sooner.

What does it mean for me?

In this large trial published in 2004, under the conditions of that time, laparoscopic repair was associated with more recurrence. A separate, later-published analysis of the same trial’s data found a higher recurrence risk after laparoscopic repair by less experienced surgeons (fewer than 250 operations); that analysis is not itself a randomized comparison. The result therefore cannot be directly generalized to every laparoscopic method done today by experienced surgeons.

Which patients does it apply to?

Men having groin hernia surgery in American veterans’ (Veterans Affairs) hospitals; the main finding concerns first-time hernias. It does not directly apply to women or to all techniques used in current expert centres.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-NEUMAYER-VA

Full scientific title

Neumayer VA — open mesh vs laparoscopic mesh (inguinal)

Clinical question

In men with inguinal hernia, how does laparoscopic mesh repair compare with open mesh repair for recurrence at 2 years?

Themes

Minimally invasive · Technology comparison · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Hernia recurrence at 2 years

Primary endpoint met?

Met

Clinical interpretation

Mixed

Population

Men at 14 VA centers; 2164 randomized; 1983 operated; 1696 (85.5%) with 2-year follow-up.

Intervention

Laparoscopic mesh repair (laparo-mesh techniques in the trial; not TEP-only)

Comparator

Open mesh repair

Results

In the analysis set, recurrence: laparoscopic 87/862 (10.1%) vs open 41/834 (4.9%); OR 2.2 (95% CI 1.5–3.2). Complications 39.0% lap vs 33.4% open. Early pain and return to usual activities favored laparoscopy (about 1 day earlier). Primary hernias: recurrence 10.1% lap vs 4.0% open; repair of recurrent hernias 10.0% vs 14.1% — no clear difference shown. Subgroups must be presented with correct denominators.

Sample size

2164

PMID

15107485

DOI

10.1056/NEJMoa040093

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the NEJM 2004 primary paper; for surgeon experience, the Ann Surg 2005 secondary analysis of the same trial (PMC full text). Information that could not be verified is stated explicitly under “Limitations”.

Funding

The trial was conducted within the Veterans Affairs Cooperative Studies Program 456 (PubMed record). Detailed funding and sponsor role could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Historical multicenter VA context; open mesh versus varied laparoscopic mesh techniques. Not a generalization to all modern experienced TEP/TAPP. Of 2164 randomized, 1696 (85.5%) had 2-year follow-up. Prespecified primary/recurrent hernia interaction significant (p=0.012); subgroups must not be mixed. Surgeon experience examined only in a separate secondary analysis (Ann Surg 2005, PMID 16135920; abstract level) — observational, not a randomized comparison; full text could not be verified from the primary source. Funding/COI could not be verified from the primary source.

Related separate studies

✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

RIVAL

Is the same keyhole groin hernia repair better when done with a robot than with standard laparoscopy?

Key finding

There was no significant difference in pain or quality of life in the first 30 days; robotic surgery took longer and cost more.

Why was this study done?

A groin hernia can be repaired by keyhole surgery entering the abdomen and placing a mesh behind the abdominal lining (transabdominal preperitoneal repair, TAPP). This small preliminary (pilot) trial asked whether doing the same operation with a robot eases recovery compared with standard laparoscopy.

What did it find?

In the first 30 days there was no significant difference in pain, quality of life, mobility or wound problems. The robotic operation took longer (median about 76 vs 41 minutes) and cost more in this study. At two years, the hernia came back in one patient in each group.

What did longer follow-up show? Two-year results were also similar, but the number of patients is too small to establish long-term superiority of one method or that the two are equal.

What does it mean for me?

For a one-sided groin hernia, the robotic approach showed no clear recovery advantage in this small trial. Because numbers were small, it is also not proven that the two methods are equal in the long term; some patients could not be reached at two-year follow-up.

Which patients does it apply to?

Patients with a one-sided groin hernia, either appearing for the first time or returning after a previous repair, operated on by experienced surgeons. People with hernias on both sides and those with previous open abdominal surgery below the navel were not included.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-RIVAL

Full scientific title

RIVAL — robotic TAPP vs laparoscopic TAPP (inguinal)

Clinical question

In unilateral primary or recurrent inguinal hernia, does robotic TAPP improve early or mid-term outcomes versus laparoscopic TAPP?

Themes

Technology comparison · Minimally invasive

Study design

Pragmatic randomized controlled trial

Primary endpoint

No single prespecified primary endpoint (pilot trial)

Clinical interpretation

Inconclusive

Population

6-center patient-blinded pilot RCT. Per the CONSORT diagram, 113 patients consented to surgery; 11 were excluded (8 withdrew before surgery, 1 had no inguinal hernia found, 2 had an incidental contralateral hernia); 102 were randomized and analyzed (lap TAPP 54, robotic TAPP 48). Completed 30-day follow-up: lap 53, robotic 45. Because some exclusion reasons can only be found during surgery and the text states randomization occurred at the preoperative visit, it could not be determined whether some exclusions occurred after randomization (not verified from the primary source). Unilateral primary or recurrent inguinal hernia; age ≥21, BMI ≤40. Bilateral hernia, prior open abdominal surgery at/below the umbilicus and prior preperitoneal mesh on the same side were excluded. Surgeons had at least 25 cases on each platform.

Intervention

Robotic TAPP

Comparator

Laparoscopic TAPP

Results

Pilot; no single prespecified primary endpoint. At 30 days, no significant differences in pain, quality of life, activity, wound events, or cosmesis. Median operative time 75.5 min robotic vs 40.5 lap; median cost USD 3258 vs 1421 (setting-specific); surgeon frustration (NASA-TLX) 32.7 robotic vs 20.1 lap.

Long-term update

Miller 2023 (RIVAL follow-up; not a separate primary card): 83/102 (81%) at 1 year; 77/102 (75%) at 2 years. Recurrence at 2 years 1 vs 1 (p=1.0); no significant differences in pain, quality of life, or activity. Low event counts and attrition preclude a firm equivalence claim.

Sample size

102

Registry

NCT02816658

PMID

32186683

DOI

10.1001/jamasurg.2020.0034

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

JAMA full article + long-term PubMed abstract (PMID 32186683, 35578051). No single primary endpoint → no single primary endpoint; the “met” badge is hidden. 2026-11-09 second check: verified from the PMC7081145 full text — 113 patients enrolled, 102 analysed after exclusions (lap 54 / robotic 48); population unilateral primary OR recurrent inguinal hernia (bilateral excluded); registration NCT02816658; surgeons ≥25 cases on each platform. The distribution of exclusions before/after randomization and the number randomized per arm were not clear from the text (CONSORT diagram) → not verified from the primary source. 2026-11-09 review of the CONSORT diagram (Figure 1, PMC image): 113 patients consented to surgery → 11 excluded (8 withdrew before surgery, 1 no inguinal hernia found, 2 incidental contralateral hernias) → 102 randomized in the diagram (lap 54 / robotic 48 received the procedure); 1-week follow-up 48/44, 1-month follow-up 53/45. Inconsistency: 3 exclusion reasons are findings that could only be detected during surgery, while the text states that randomization took place at the preoperative assessment; whether these exclusions occurred after randomization could not be determined → not verified from the primary source. The sample size is given as the number randomized in the diagram (102).

Funding

Intuitive Surgical grant IUSI1602MR. According to the article, the sponsor had no role in study design, data analysis or writing decisions.

Conflicts of interest

The JAMA article reports fee/consulting/institutional research relationships of several authors with Intuitive, BD, Medtronic, Ethicon, Gore and others. “No COI” is not stated. Full author-level Disclosure: https://jamanetwork.com/journals/jamasurgery/fullarticle/2763027 — the separate disclosure of the long-term publication can be confirmed from its full text (pending).

Limits

Pilot, small sample; experienced surgeons; unilateral (primary or recurrent) cases, bilateral hernia excluded; no generalization to complex hernias. 113 consented → 11 excluded → 102 randomized/analyzed (timing of exclusions uncertain — not verified from the primary source); long-term follow-up 83/102 at 1 year and 77/102 at 2 years. Absence of a significant difference is not equivalence. Do not pool with ROLAIS/ROGER.
✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

Olavarria

Is robotic abdominal wall (ventral) hernia repair better than standard laparoscopic repair?

Key finding

There was no significant difference in days spent in hospital; robotic surgery took longer and cost more. The difference in recurrence was not statistically certain.

Why was this study done?

A ventral (abdominal wall) hernia can be repaired by keyhole surgery with a mesh placed inside the abdomen (intraperitoneal onlay mesh repair, IPOM). The study asked whether doing this operation with a robot reduces time spent in hospital compared with standard laparoscopy.

What did it find?

There was no significant difference between the groups in days spent in hospital in the 90 days after surgery; this does not prove the two methods are equal. The robotic operation took longer (about 141 vs 77 minutes) and cost more in this study. At two years the hernia came back in about 4 in 100 patients in the robotic group and about 13 in 100 in the laparoscopic group, but this difference was not statistically significant.

What did longer follow-up show? At two years, although there were some favourable signs in the robotic group, the difference in hernia recurrence was not statistically established.

What does it mean for me?

In this study the robotic approach did not shorten hospital time. Although some signals were favourable, it was not proven that the robot is better.

Which patients does it apply to?

Selected patients with a small or medium abdominal wall hernia planned for intraperitoneal mesh repair. It does not apply to complex abdominal wall reconstruction.

Show scientific details

SURG-OLAVARRIA

Full scientific title

Olavarria — robotic vs laparoscopic ventral IPOM

Clinical question

In ventral hernia, does robotic IPOM reduce total hospital days within 90 days versus laparoscopic IPOM?

Themes

Technology comparison · Minimally invasive

Study design

Pragmatic randomized controlled trial

Primary endpoint

Total hospital days within 90 days after surgery

Primary endpoint met?

Not met

Clinical interpretation

Inconclusive

Population

124 randomized; robotic 65, lap 59. Small/medium ventral hernia; intraperitoneal onlay mesh (IPOM); similar technique on different platforms.

Intervention

Robotic ventral IPOM

Comparator

Laparoscopic ventral IPOM

Results

Median hospital days within 90 days: 0 in both arms; relative rate 0.90 (95% CI 0.37–2.19), p=0.82 — primary superiority not shown. Operative time 141 min robotic vs 77 lap. Total health-system cost USD 15,865 vs 12,955 (US context). Two enterotomies in the robotic arm vs 0 lap; not alone proof of a definitive risk difference.

Long-term update

Dhanani 2023 (same-cohort follow-up; not a separate primary card): 101/124 completed (robotic 54/65; lap 47/59). Recurrence 2/54 (4%) robotic vs 6/47 (13%) lap; RR 0.30 (95% CI 0.06–1.39), p=0.12 — not statistically significant; no superiority claim. Reoperation 0 robotic vs 5/47 (11%) lap, p=0.019: secondary, sparse-event, hypothesis-generating finding.

Sample size

124

Registry

NCT03490266

PMID

32665218

DOI

10.1136/bmj.m2457

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Primary full text + 2-year PubMed/linked text (PMID 32665218, 37203558, NCT03490266). The 2023 statement is not presented as the 2020 author statement. 2026-11-09 second check: the 2020 funding and ICMJE conflict-of-interest statement were verified from the PMC7359869 full text; 175 eligible patients were approached, 124 randomized (65/59).

Funding

2020 BMJ full text: supported by an Intuitive Surgical investigator-initiated grant. According to the article, the funder had no role in design, data collection, analysis, interpretation, writing or the decision to submit for publication.

Conflicts of interest

2020 BMJ ICMJE statement: Intuitive Surgical support for the submitted work; SKS — C-SATS consulting, Gore education fees, past research grants from Neosurgical and Medigus; TDW — consulting for CONMED, EndoEvolution, Davol, Livsmed and ColubrisMX advisory board; no other relationships declared. The statement of the 2023 follow-up publication is separate (S.S.: Activ Surgical) and is not presented as the 2020 statement.

Limits

Small/medium ventral IPOM in a selected population; not generalizable to retromuscular robotic repair or complex abdominal-wall reconstruction. Attrition and multiple secondary outcomes. Recurrence difference must not be presented as statistically significant superiority. The trial was designed for a superiority question (sample size: 0 vs 1 day difference, two-sided α 0.05, 80% power); no equivalence or non-inferiority margin was defined. The absence of a statistically significant difference in the primary outcome is not proof that the two methods are equivalent.

Related separate studies

Physician-approved (conditional) · 14 Nov 2026Hernia and abdominal wall

ROLAIS

Does doing the same keyhole groin hernia repair with a robot make recovery easier?

Why conditional approval?

It was a single-centre, open-label trial; its main measure was a blood marker of inflammation (CRP), not an outcome patients experience directly.

Verified source: Verified at the level of the primary publication abstract and publisher statements.

Conditions kept on this card: Single-centre, open-label trial; the primary endpoint is a surrogate measure (CRP); preoperative withdrawals were unbalanced between groups (2 / 9).

Key finding

After robotic surgery, the blood marker of inflammation was lower and fewer early complications were reported; there was no significant difference in long-lasting pain or recurrence.

Why was this study done?

For groin hernia repair by keyhole surgery entering the abdomen and placing a mesh behind the abdominal lining (transabdominal preperitoneal repair, TAPP), the study asked whether the robotic approach changes the body's inflammatory response and early results compared with standard laparoscopy.

What did it find?

In the first days after robotic surgery, a blood marker of inflammation (C-reactive protein, CRP) was lower. Reported early complications occurred in about 23 in 100 patients in the robotic group and about 42 in 100 in the laparoscopic group; same-day discharge was more common after robotic surgery. There was no significant difference in long-lasting pain or the hernia coming back.

What does it mean for me?

In this small single-centre study some early results were better with the robot. A lower blood inflammation marker does not by itself mean fewer complications or better long-term results; these findings do not prove that robotic groin hernia surgery is generally better in the long term.

Which patients does it apply to?

Patients having planned groin hernia surgery at this centre by the same experienced surgeons. Abdominal wall (umbilical or incisional) hernias were not the subject of this study. Detailed selection criteria have not been verified from the published full text.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-ROLAIS

Full scientific title

ROLAIS — robotic TAPP vs laparoscopic TAPP (inflammation / early outcomes)

Clinical question

In inguinal hernia, does robotic TAPP differ from laparoscopic TAPP for early inflammatory response (plasma CRP) and short-term clinical outcomes?

Themes

Technology comparison · Minimally invasive

Study design

Pragmatic randomized controlled trial

Primary endpoint

Plasma CRP levels (surrogate / inflammatory response)

Primary endpoint met?

Met

Clinical interpretation

Mixed

Population

Elective groin (inguinal) hernia repair; not a ventral (abdominal wall) hernia trial. Single-center open-label RCT; 150 randomized (R-TAPP 76, L-TAPP 74); 11 withdrew before surgery; 139 in the ITT analysis (R-TAPP 74, L-TAPP 65) — withdrawals unbalanced between arms (2 vs 9). Same experienced surgeons. Selection criteria not verified from the published full text; per the ClinicalTrials.gov record: ASA 1–3, unilateral/bilateral, recurrent or inguinoscrotal hernia suitable for laparoscopy; exclusions included incarcerated hernia needing emergency surgery, pregnancy, chronic pain needing regular analgesics, active cancer, inflammatory/immunological disease (registry information, not verified in the publication).

Intervention

Robotic TAPP (R-TAPP)

Comparator

Laparoscopic TAPP (L-TAPP)

Results

CRP was 23% and 32% lower on postoperative days 1 and 3 with R-TAPP (p=0.001). Early IL-6 was lower. Operative time was 13.4 min shorter with robotics (p<0.001). Reported complications 23.0% robotic vs 41.5% lap (p=0.029); same-day discharge 95.9% vs 81.5% (p=0.012). No significant differences in chronic pain or recurrence — not proof of long-term superiority.

Sample size

150

Registry

NCT05839587

PMID

40277023

DOI

10.1093/bjs/znaf074

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the BJS 2025 primary paper and the publisher's open-access disclosures; ClinicalTrials.gov NCT05839587. Information that could not be verified is stated explicitly under “Limitations”.

Funding

University Hospital of Southern Denmark, University of Southern Denmark Department of Regional Health Research and Knud og Edith Eriksens Mindefond. The paper states that the sponsors were not involved in the study process.

Conflicts of interest

The authors declared no conflicts of interest (BJS open-access disclosure).

Limits

Single center; open-label; surrogate primary (CRP) — lower CRP is not definitive proof of fewer clinical complications or long-term superiority; complications and discharge are secondary outcomes. Pre-operative withdrawals unbalanced between arms (2 vs 9). Long-term recurrence superiority not established. Published selection criteria: could not be verified from the primary source. Do not mix data with RIVAL/ROGER; explain contradictions without hiding findings.
✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

ROGER

Does robotic groin hernia repair reduce pain after surgery?

Key finding

There was no significant difference in pain one day after surgery; robotic surgery took longer and complications were similar.

Why was this study done?

A groin hernia can be repaired by two keyhole methods: laparoscopic repair in front of the abdominal lining (totally extraperitoneal repair, TEP) and repair entering the abdomen with a mesh placed behind the lining (transabdominal preperitoneal repair, TAPP). This study compared laparoscopic TEP with robotic TAPP for pain on the first day after surgery.

What did it find?

There was no significant difference in pain when coughing one day after surgery. The robotic operation took longer (for a one-sided hernia, on average about 80 vs 64 minutes). Complications occurred in about 10–11 in 100 patients and were similar. Surgeons reported a lower workload with the robotic approach.

What does it mean for me?

Robotic TAPP was not shown to be better for first-day pain. Because the groups differed in both robot/laparoscopy and TEP/TAPP technique, the result cannot be attributed to the robot alone.

Which patients does it apply to?

Patients having planned surgery for a first-time groin hernia at this single centre. Hernias that had come back after a previous repair were not included. Long-term pain and hernia recurrence cannot be judged from this study.

Show scientific details

SURG-ROGER

Full scientific title

ROGER — laparoscopic TEP vs robotic TAPP

Clinical question

In inguinal hernia, does robotic TAPP reduce cough pain on an NRS at 24 hours versus laparoscopic TEP?

Themes

Technology comparison · Surgical technique · Minimally invasive

Study design

Pragmatic randomized controlled trial

Primary endpoint

Cough pain on NRS at 24 hours

Primary endpoint met?

Not met

Clinical interpretation

Inconclusive

Population

Single-center patient- and assessor-blinded RCT; elective primary inguinal hernia (unilateral or bilateral); recurrent hernia, prior open abdominal surgery at/below the umbilicus, ascites, dialysis, pregnancy and emergency surgery excluded. 182 randomized; TEP 91, robotic TAPP 91. Primary analysis: TEP 90, rTAPP 88. Surgeons had to have done ≥30 of each procedure in the past year.

Intervention

Robotic TAPP

Comparator

Laparoscopic TEP

Results

Median cough pain at 24 h (NRS): TEP 5 (IQR 2–7), rTAPP 4 (IQR 2–7), p=0.431 — superiority not shown. Mean operative time for unilateral hernia 64.2 min (TEP) vs 80.3 min (rTAPP; p<0.001). Complications 11% vs 10%, no significant difference. Surgeon NASA-TLX workload TEP 34.0 vs rTAPP 18.4 (p<0.001).

Sample size

182

Registry

NCT05216276

PMID

41511872

DOI

10.1093/bjs/znaf283

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Open-access BJS full text (DOI 10.1093/bjs/znaf283, NCT05216276). 2026-11-09 second check: verified from the PMID 41511872 PubMed primary bibliographic record (same record lists DOI 10.1093/bjs/znaf283, PMC12785885, NCT05216276; Br J Surg 113(1), e-published 24 Dec 2025). Eligibility criteria and surgeon experience were read from the PMC12785885 full text.

Funding

Intuitive Surgical Sàrl provided materials, a research nurse and software/hardware support; the article states that the sponsor had no role in design, conduct, analysis, interpretation or reporting.

Conflicts of interest

The authors declared no conflicts of interest in the article.

Limits

Single center; early pain only as primary outcome; robotic platform and anatomic approach change together; insufficient for long-term chronic pain and recurrence. Per the article, surgeons were more proficient in TEP (the institution’s standard); some rTAPP operations may have been done before completing the learning curve. Sample size was powered for the primary pain outcome and may be insufficient for complication/recurrence differences.
Physician-approved (conditional) · 14 Nov 2026Hernia and abdominal wall

SCUR 1999

Does keyhole groin hernia repair give a faster recovery?

Why conditional approval?

The trial has no single main measure and hernia recurrence was followed for only 1 year; a mesh manufacturer (Ethicon) was among the funders.

Verified source: The full text of the primary publication (Ann Surg 1999) was reviewed.

Conditions kept on this card: The primary aim combines three outcomes (no single primary endpoint); the 1-year recurrence follow-up is short; Ethicon is among the funders.

Key finding

Full recovery and return to work were earlier after keyhole repair; however, complications were reported more often and cost was higher. There was no significant difference in one-year recurrence.

Why was this study done?

For groin hernia, the aim was to compare keyhole repair entering the abdomen with a mesh behind the abdominal lining (transabdominal preperitoneal repair, TAPP) with two different open operations for recovery, return to work, complications and the hernia coming back.

What did it find?

Full recovery took on average about 18 days with TAPP and about 24–26 days with the open methods; return to work was also earlier with TAPP. Cost was higher with TAPP. Within one year, hernia recurrence did not differ significantly between groups. Complications, however, were reported more often in the TAPP group.

What does it mean for me?

Keyhole TAPP repair may allow faster recovery, but it was more expensive and complications were more frequent in this study. The open operations in this study should not be treated as identical to a single modern Lichtenstein technique.

Which patients does it apply to?

Men aged 40–75 with a groin hernia. One-year follow-up is short for judging long-term recurrence.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-SCUR-1999

Full scientific title

SCUR 1999 — TAPP vs open preperitoneal mesh vs conventional open

Clinical question

In men with inguinal hernia, how does laparoscopic TAPP compare with open preperitoneal mesh and conventional open repair for recovery, return to work, recurrence, and cost?

Themes

Minimally invasive · Surgical technique · Technology comparison · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

According to the full text, the primary aim combined three outcomes: time to return to normal physical activity, early (1-year) recurrence rate and complications. There is no single primary endpoint; the sample size was based on 1-year recurrence and time to full recovery, with 200 patients planned per group.

Study outcomes (by outcome)

  • Time to full recovery

    Result: TAPP 18.4 days; open preperitoneal mesh 24.2 days; conventional open 26.4 days; TAPP vs each open arm p<0.001.

    Statistical interpretation: Recovery favored TAPP.

    Endpoint met?: Met

  • Return to work

    Result: TAPP 14.7 days; open preperitoneal mesh 17.7 days (p=0.05); conventional open 17.9 days (p=0.04).

    Statistical interpretation: Return to work favored TAPP.

    Endpoint met?: Met

  • 1-year recurrence (event counts)

    Time point: 1 year

    Result: Conventional open 4; open mesh 11; TAPP 4. Between-group difference not statistically significant. Event counts only; not converted to group percentages.

    Statistical interpretation: No significant difference shown; not equivalence. One-year follow-up is short.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Multicenter; 613 men aged 40–75; 604/613 (98.5%) followed at 1 year.

Intervention

Laparoscopic TAPP preperitoneal mesh

Comparator

Open preperitoneal mesh AND conventional open repair (three arms; open arms are not labeled Lichtenstein)

Results

Recovery and return to work favored TAPP; restriction of physical activity at day 7 was lower with TAPP. Cost was higher with TAPP (operating time and equipment, even after accounting for earlier return to work). Complications were reported more often in the TAPP group, with a different pattern between groups (per-arm counts could not be verified from the primary source). One-year recurrence event counts did not differ significantly. Clinical interpretation candidate: mixed — physician review required.

Sample size

613

PMID

10450737

DOI

10.1097/00000658-199908000-00013

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract and PMC full text of the Ann Surg 1999 primary paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Grants from the County of Älvsborg Research and Development Foundation, the Bröderna Eriksson Research Fund, Ethicon and Astra Pain Control AB (Ann Surg 1999 full text). The funders include device and drug companies.

Conflicts of interest

No separate author conflict-of-interest statement was found in the full text. This does not mean there were no conflicts.

Limits

Historical three-arm design; 1-year recurrence follow-up is short. The primary aim combines three separate outcomes; there is no single primary endpoint. Per-arm randomized numbers, the complication breakdown and surgeon experience were not extracted from the full-text tables in this review. The funders include Ethicon and Astra Pain Control AB. Source level: PMC full text.
Physician-approved (conditional) · 9 Nov 2026Hernia and abdominal wall

MRC 1999

What does keyhole groin hernia repair gain and risk compared with open repair?

Why conditional approval?

This is an older trial; different techniques were combined in the open-surgery group and the mix of keyhole techniques was not verified.

Verified source: Verified at the level of the primary publication abstract and the long-term publication abstract (PubMed); the full text was not reviewed.

Conditions kept on this card: Historical trial; mixed techniques in the open arm; the laparoscopic sub-technique mix and cost figures were not verified; funding, conflicts of interest and surgeon experience were not verified.

Key finding

Keyhole repair gave a faster return to normal life and less persistent groin pain at one year; however, the serious surgical complications and early recurrences occurred in this group.

Why was this study done?

In 26 hospitals in the UK and Ireland, keyhole (laparoscopic) and open groin hernia repair were compared for recovery, pain, complications and the hernia coming back.

What did it find?

Return to normal social activities was faster after laparoscopic repair (median 10 vs 14 days). At one year, persistent groin pain occurred in about 29 in 100 patients after laparoscopic repair and about 37 in 100 after open repair. Complications in the first week were more common after open repair. However, all 3 serious operative complications occurred in the laparoscopic group, and within one year the hernia came back in about 2 in 100 laparoscopic patients and in none after open repair.

What did longer follow-up show? In questionnaire follow-up up to five years, groin numbness was reported less often after keyhole surgery. Groin pain also tended to be less common; some comparisons were statistically borderline. These follow-up data should not be read as a definitive long-term comparison of hernia recurrence in all patients.

What does it mean for me?

Laparoscopic repair may give faster recovery and less persistent pain, but in this historical study there were warning signals for serious complications and recurrence. Results should be read with the techniques and surgeon experience of that time in mind.

Which patients does it apply to?

Patients undergoing groin hernia repair. Detailed selection criteria have not been verified for this card.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-MRC-HERNIA-1999

Full scientific title

MRC 1999 — laparoscopic vs open groin hernia repair

Clinical question

How does laparoscopic inguinal hernia repair compare with open repair for complications, return to social activity, persistent pain, recurrence, and cost?

Themes

Minimally invasive · Technology comparison · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Multiple primary endpoints: complications; return to normal social activity; hernia recurrence; persistent groin pain at 1 year; health-service costs

Study outcomes (by outcome)

  • ≥1 complication in week 1

    Time point: Week 1

    Result: Laparoscopic 108 (29.9%) vs open 155 (43.5%), p<0.001. Percent denominators are patients assessed at that time point; do not recalculate as 108/468.

    Statistical interpretation: Early complication rate higher in the open arm.

    Endpoint met?: Met

  • Serious surgical complications

    Result: 3 serious surgical complications; all in the laparoscopic arm.

    Statistical interpretation: Safety signal; shown alongside patient benefits.

    Endpoint met?: Met

  • Return to normal social activity

    Result: Median 10 days (IQR 7–21) vs 14 days (IQR 7–28), p=0.004.

    Statistical interpretation: Recovery favored laparoscopy.

    Endpoint met?: Met

  • Persistent groin pain at 1 year

    Time point: 1 year

    Result: 28.7% vs 36.7%, p=0.018.

    Statistical interpretation: Persistent pain lower after laparoscopy.

    Endpoint met?: Met

  • 1-year recurrence

    Time point: 1 year

    Result: Laparoscopic 7 (1.9%) vs open 0, p=0.017. Historical finding; not generalizable to all contemporary expert laparoscopic repairs.

    Statistical interpretation: Recurrence signal in the laparoscopic arm; read in a safety context.

    Endpoint met?: Met

  • Health-service costs

    Result: Cost figures not verified from the PubMed abstract; not added (not verified from the primary source).

    Statistical interpretation: Cost field left empty of invented numbers.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

26 UK/Ireland hospitals; 928 randomized; laparoscopic 468, open 460 (433 tension-free mesh in the open arm). The open arm is not entirely Lichtenstein.

Intervention

Laparoscopic repair (subgroup technique mix not coded as TEP-only or TAPP-only without source verification)

Comparator

Open repair (mixed techniques; mostly tension-free mesh)

Results

Patient benefit: faster return to social activity and less persistent pain at 1 year; week-1 complication rate higher after open repair. Safety: 3 serious surgical complications (all lap) and 1-year recurrence 7 (1.9%) lap vs 0 open. Clinical interpretation candidate: mixed.

Long-term update

Grant 2004 (SURG-GRANT-2004; not a separate primary card): 750/928 (80.8%) returned ≥1 postal questionnaire between years 2–5. Groin pain was less often observed after laparoscopy but some comparisons were statistically borderline; groin numbness about half as frequent (p<0.001). No significant differences for testicular pain or thigh numbness. One-year recurrence must not be restated as five-year recurrence.

Sample size

928

PMID

10421299

DOI

10.1016/S0140-6736(98)10010-7

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract (PMID 10421299 + Grant 15515112). Open arm used mixed techniques (433/460 tension-free mesh); not coded as Lichtenstein only / TEP-TAPP only. Cost figures not in the abstract. Funding/conflicts of interest/surgeon experience not verified from the primary source.

Limits

Historical MRC; open arm mixed; lap subgroup technique mix unverified. Cost figures absent from abstract. Funding/conflicts of interest/surgeon experience not verified from the primary source. Source level: PubMed abstract (+ Grant abstract long-term).
✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

Elmessiry 2020

Is keyhole repair more comfortable than open repair for groin hernias on both sides?

Key finding

Keyhole surgery took longer; early pain, hospital stay and return to daily life favoured keyhole repair. No significant difference in recurrence was shown.

Why was this study done?

For uncomplicated hernias appearing for the first time in both groins, the aim was to compare keyhole repair entering the abdomen with a mesh behind the abdominal lining (transabdominal preperitoneal repair, TAPP) with two different open operations.

What did it find?

TAPP took longer; however, early pain, hospital stay and return to work/daily activity favoured TAPP. Long-lasting pain and mesh sensation were reported less and satisfaction was higher with TAPP. At three years no significant difference in hernia recurrence was shown; this does not mean they are equal.

What does it mean for me?

For hernias on both sides, keyhole TAPP may be associated with faster recovery and less pain. As this was a single-centre study, results should not be generalized to every patient or every technique.

Which patients does it apply to?

Patients aged 20–80 with first-time, painless, uncomplicated hernias in both groins. Because people with groin pain before surgery were not included, the pain results cannot be directly applied to them. People with recurrent hernias, very large or complicated hernias, severe other illness, or previous surgery below the navel were not included.

Show scientific details

SURG-ELMESSIRY-2020

Full scientific title

Elmessiry 2020 — bilateral primary inguinal: TAPP vs open PP vs Lichtenstein

Clinical question

In uncomplicated primary bilateral inguinal hernia, how does bilateral TAPP compare with open preperitoneal single mesh and bilateral Lichtenstein for early perioperative outcomes and 3-year recurrence/QoL?

Themes

Minimally invasive · Surgical technique · Technology comparison

Study design

Pragmatic randomized controlled trial

Primary endpoint

Early perioperative outcomes (multiple components: time, complications, stay, pain, return to work/activity) — not collapsed to a single yes/no label

Study outcomes (by outcome)

  • Early pain, hospital stay, return to activity/work

    Result: TAPP took longer but was reported as advantageous versus open arms for early pain, stay, and return to activity/work.

    Statistical interpretation: Early components favored TAPP; no single overall-superiority label.

    Endpoint met?: Met

  • Chronic pain, mesh sensation, satisfaction

    Result: Lower chronic pain and mesh sensation and higher satisfaction reported (TAPP).

    Statistical interpretation: Patient-reported components favored TAPP.

    Endpoint met?: Met

  • 3-year recurrence

    Time point: 3 years

    Result: No statistically significant difference shown; not an equivalence claim. Lost to follow-up: TAPP 6, open PP 7, Lichtenstein 9.

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Single center; 180 patients (60/60/60); painless, uncomplicated primary bilateral inguinal hernia; age 20–80. Exclusions include recurrence, giant scrotal/complicated cases, ASA III–IV, prior infraumbilical surgery and groin pain from any other cause; per the article, no patient had preoperative pain.

Intervention

Laparoscopic bilateral TAPP (two separate meshes, tack fixation)

Comparator

Open preperitoneal single mesh AND bilateral open Lichtenstein (three arms; not a TEP comparison)

Results

TAPP longer; early pain/stay/activity and chronic pain/mesh sensation/satisfaction reported favoring TAPP. No significant 3-year recurrence difference (not equivalence). Attrition 6/7/9.

Sample size

180

Registry

NCT04437784

PMID

33024556

DOI

10.1016/j.amsu.2020.08.055

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Open PMC full text (PMID 33024556, NCT04437784). This is not a TEP vs TAPP study. Table items were not reconstructed without a source. Funding/conflicts of interest are author statements (not an independent audit). 2026-11-09 second check: verified from the PMC7527655 full text — inclusion criterion “painless uncomplicated primary bilateral inguinal hernias”, groin pain from another cause as an exclusion criterion; according to the article, no patient had preoperative pain.

Funding

Author statement in the full text: Funding none to declare (not an independent audit).

Conflicts of interest

Author statement in the full text: competing interest none to declare (not an independent audit).

Limits

Single center, selected population (painless, uncomplicated primary bilateral hernia only — not generalizable to patients with preoperative pain, especially for chronic-pain outcomes), three arms, two experienced surgeons, attrition, power/rare-event limits. Source level: open full text (selected findings).
✓ Physician-approved · 9 Nov 2026Hernia and abdominal wall

Koch 2006

Does the mesh need to be fixed in place in keyhole groin hernia repair?

Key finding

In this small 40-patient study, hospital stay was shorter when the mesh was not fixed; no recurrence was seen in either group during follow-up.

Why was this study done?

In keyhole groin hernia repair in front of the abdominal lining (totally extraperitoneal repair, TEP), the question was whether the mesh needs to be fixed with small metal staples (tacks).

What did it find?

In this small study of 40 patients, without fixation the hospital stay was shorter (about 8 vs 16 hours), less pain medication was needed in the recovery unit and difficulty passing urine was less common. There was no significant difference in pain or return to normal activity. No hernia recurrence was seen in either group during follow-up.

What does it mean for me?

For some patients, not fixing the mesh may ease short-term recovery. However, because the study was very small, it is not proven that no fixation is equally safe for every hernia type.

Which patients does it apply to?

Patients having groin hernia surgery with the TEP method. It does not apply to large or complex hernias.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-KOCH-2006

Full scientific title

Koch 2006 — mesh fixation vs no fixation in TEP

Clinical question

In TEP repair, how does tack fixation of mesh compare with no fixation for early recovery, pain, and recurrence?

Themes

Surgical technique · Minimally invasive

Study design

Pragmatic randomized controlled trial

Primary endpoint

Abstract reports multiple aims; single prespecified primary left unknown without verification

Study outcomes (by outcome)

  • Hospital stay and PACU analgesia / urinary retention

    Result: No fixation: stay 8.3 vs 16.0 hours (p=0.01); less PACU narcotic use; less urinary retention.

    Statistical interpretation: Some short-term recovery outcomes favored no fixation.

    Endpoint met?: Met

  • Pain and return to usual activity

    Result: No statistically significant difference found.

    Statistical interpretation: No significant difference shown.

    Endpoint met?: Not met

  • Recurrence

    Time point: Follow-up 6–30 months, median 19 months

    Result: No recurrence in either arm. Not equivalence or non-inferiority evidence (N=40).

    Statistical interpretation: Small sample; absence of recurrence is not proof of safety equivalence.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Single-blind randomized; 40 patients: TEP + tack 20, TEP without fixation 20.

Intervention

TEP mesh without fixation

Comparator

TEP with tack fixation

Results

Small (N=40) trial. No fixation favored short-term stay/analgesia/retention; no significant pain/activity difference; no recurrence — not equivalence.

Sample size

40

PMID

17575757

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Open PMC full text + PubMed abstract (PMID 17575757). DOI could not be verified in PubMed → left empty. Funding/conflicts of interest left empty (not invented in the absence of an explicit statement).

Limits

40 patients; follow-up 6–30 months (median 19); not generalizable to large direct/complex hernias; not sole basis for all fixation guidance. DOI empty. Funding/conflicts of interest not verified from the primary source. Source: open full text.
Physician-approved (conditional) · 9 Nov 2026Hernia and abdominal wall

Bringman 2006

Does a lightweight mesh cause fewer complaints than a standard mesh in groin hernia repair?

Why conditional approval?

It covers only men with a first, one-sided hernia repaired by the open (Lichtenstein) method; a single main measure could not be determined from the abstract.

Verified source: Verified only at the level of the primary publication abstract (PubMed); the full text was not reviewed.

Conditions kept on this card: Men with primary, unilateral hernia and open Lichtenstein repair only; not generalizable to TEP; a single primary endpoint could not be determined from the abstract; funding, conflicts of interest and surgeon experience were not verified.

Key finding

At three years, recurrence was similar in both groups; some groin complaints and mesh sensation were less common with lightweight mesh.

Why was this study done?

In open Lichtenstein groin hernia surgery, the study asked whether a lighter, large-pore mesh reduces pain and foreign-body sensation compared with a standard mesh.

What did it find?

At three years the hernia came back in about 4 in 100 patients in both groups; there was no significant difference. In the lightweight mesh group, pain on examination, some groin complaints and mesh sensation were less common.

What does it mean for me?

In open Lichtenstein surgery, lightweight mesh may reduce some pain and foreign-body complaints. This result cannot be transferred directly to keyhole (laparoscopic) operations.

Which patients does it apply to?

Men having open Lichtenstein surgery for a first-time, one-sided groin hernia.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-BRINGMAN-2006

Full scientific title

Bringman 2006 — lightweight vs standard mesh in Lichtenstein

Clinical question

In primary unilateral inguinal hernia repaired by open Lichtenstein, how does lightweight mesh compare with standard polypropylene for 3-year recurrence and pain/mesh sensation?

Themes

Surgical technique

Study design

Pragmatic randomized controlled trial

Primary endpoint

Single primary endpoint not clear from abstract → unknown; recurrence and pain components shown separately

Study outcomes (by outcome)

  • 3-year recurrence (examined cohort)

    Time point: ~3 years

    Result: Standard 9/243 (3.7%) vs lightweight 9/251 (3.6%); no significant difference reported. Examined: standard 243/294, lightweight 251/296.

    Statistical interpretation: No significant recurrence difference; not an equivalence claim.

    Endpoint met?: Not met

  • Pain, groin symptoms, mesh sensation

    Result: In the lightweight group, examination pain, pain when rising from lying, some groin symptoms, and mesh sensation were less.

    Statistical interpretation: Pain/mesh sensation favored lightweight mesh (open Lichtenstein context).

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Six centers; 600 men randomized; primary unilateral inguinal hernia; open Lichtenstein. 590 underwent surgery.

Intervention

Lightweight (large-pore, partly absorbable) mesh — Lichtenstein

Comparator

Standard polypropylene mesh — Lichtenstein

Results

3-year recurrence similar (~3.7% vs 3.6%); some pain/mesh-sensation findings lower with lightweight. Clinical reading: mixed (open pathway).

Sample size

600

PMID

16862613

DOI

10.1002/bjs.5403

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract (PMID 16862613). Not generalisable to TEP. Funding/conflicts of interest/surgeon experience/primary full text not verified from the primary source.

Limits

Men, primary, unilateral; mesh material + open pathway together. Funding/conflicts of interest/surgeon experience not verified from the primary source. Source level: PubMed abstract.
Physician-approved (conditional) · 9 Nov 2026Hernia and abdominal wall

TULP 2016

Does a lightweight mesh reduce pain after keyhole groin hernia repair?

Why conditional approval?

It compared only two specific mesh products (Ultrapro, Prolene) in men with a first, one-sided hernia; the trial’s main measure was not verified from the full text.

Verified source: Verified at the level of the 2016 primary publication abstract and the 2018 long-term publication abstract (PubMed); the full text was not reviewed.

Conditions kept on this card: Specific mesh products (Ultrapro / Prolene); men with primary, unilateral hernia only; a single primary endpoint and the protocol were not verified from the full text; not generalizable to open Lichtenstein findings (Bringman); funding and conflicts of interest were not verified.

Key finding

Lightweight mesh did not reduce pain; hernias came back more often in the lightweight mesh group.

Why was this study done?

In keyhole groin hernia repair in front of the abdominal lining (totally extraperitoneal repair, TEP), the study asked whether a lightweight mesh (Ultrapro) reduces pain compared with a heavyweight mesh (Prolene).

What did it find?

The lightweight mesh did not reduce pain: at one year, significant pain occurred in about 3 in 100 patients with the lightweight mesh and about 1 in 100 with the heavyweight mesh. At two years the hernia came back in about 3 in 100 with the lightweight mesh and about 1 in 100 with the heavyweight mesh. At five years recurrence remained higher with the lightweight mesh (about 4 vs 1 in 100).

What did longer follow-up show? At five years, hernias still came back more often in the lightweight mesh group; the difference was especially notable in hernias coming through a weakness in the back wall of the groin canal (direct hernias). This result relates to specific mesh products and selected patients.

What does it mean for me?

In this study the lightweight mesh used in TEP did not give the expected benefit and was associated with more recurrence. The result relates to specific mesh products and does not prove the same for every lightweight mesh.

Which patients does it apply to?

Men having TEP surgery for a first-time, one-sided groin hernia that can be pushed back in.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-TULP-2016

Full scientific title

TULP 2016 — lightweight (Ultrapro) vs heavyweight (Prolene) mesh in TEP

Clinical question

In primary unilateral reducible inguinal hernia repaired by TEP, how does Ultrapro lightweight mesh compare with Prolene heavyweight mesh for pain and recurrence?

Themes

Surgical technique · Minimally invasive · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

2016 reports pain/recurrence; single primary left unknown without protocol/full-text verification

Study outcomes (by outcome)

  • Clinically relevant pain (NRS 4–10)

    Time point: 1 year and 2 years

    Result: 1y: lightweight 2.9% vs heavyweight 0.7% (p=0.01); difference persisted at 2y (p=0.03). Lightweight did not reduce pain.

    Statistical interpretation: Pain findings against lightweight mesh.

    Endpoint met?: Met

  • 2-year recurrence

    Time point: 2 years

    Result: Heavyweight 4 (0.8%), lightweight 13 (2.7%), p=0.03. Arm order: lightweight=Ultrapro, heavyweight=Prolene.

    Statistical interpretation: Recurrence against lightweight mesh.

    Endpoint met?: Met

  • Foreign-body sensation and quality of life

    Result: No difference found (abstract).

    Statistical interpretation: No significant difference shown.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Double-blind RCT; 950 men; primary, reducible, unilateral inguinal hernia; TEP.

Intervention

TEP + Ultrapro lightweight (partly absorbable) mesh

Comparator

TEP + Prolene heavyweight mesh

Results

At 2y: more clinically relevant pain and more recurrence with lightweight mesh; no foreign-body/QoL difference. Clinical reading: mixed / lightweight caution (TEP).

Long-term update

Roos 2018 (TULP 5y; not a separate primary card): recurrence data from 790/950 (83.2%). Five-year total recurrence: lightweight 3.8% vs heavyweight 1.1% (p=0.01); difference notable especially for primary direct hernia (p=0.003). Shown separately from 2016 two-year results.

Sample size

950

PMID

26779980

DOI

10.1097/SLA.0000000000001579

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract (2016) + 2018 long-term abstract. The Bringman open Lichtenstein findings are not carried over to this card. Funding/conflicts of interest/protocol not verified from the primary source.

Limits

Specific mesh products (Ultrapro/Prolene); men, primary unilateral. Funding/conflicts of interest/protocol not verified from the primary source. Not generalized from Bringman. Source: PubMed abstract (+ 2018 long-term abstract).
✓ Physician-approved · 9 Nov 2026Gallbladder

ACDC

For gallbladder inflammation, should surgery be done right away or after the inflammation settles?

Patients with a related complication within 75 days (per 100 patients, approximately)
Time point: 75 daysOrange dot: a patient who had this outcomeGrey dot: a patient who did not

A difference in favour of early surgery was shown; total hospital stay was also shorter.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For patients with gallbladder inflammation (acute cholecystitis), the study asked whether keyhole removal of the gallbladder is better within the first 24 hours after admission, or after antibiotics first with surgery 1–6 weeks later.

What did it find?

Within 75 days, related complications occurred in about 12 in 100 patients with early surgery and about 34 in 100 with delayed surgery. Total hospital stay was shorter with early surgery (on average about 5 vs 10 days) and study-period hospital costs were lower. No significant difference was shown in conversion to open surgery or death.

What does it mean for me?

In patients fit for surgery, removing the gallbladder within the first 24 hours was associated with fewer complications than antibiotics first and surgery later. Timing is still decided according to the patient’s condition and surgical risk.

Which patients does it apply to?

Patients with acute cholecystitis who are fit for surgery. It does not apply to patients too ill to tolerate surgery.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-ACDC

Full scientific title

ACDC — early versus delayed laparoscopic cholecystectomy

Clinical question

In acute cholecystitis, does laparoscopic cholecystectomy within 24 hours of hospital admission reduce 75-day clinical morbidity versus antibiotics first and elective laparoscopic cholecystectomy after 7–45 days?

Themes

Timing and strategy · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Predefined relevant morbidity within 75 days

Study outcomes (by outcome)

  • Predefined relevant morbidity

    Time point: 75 days

    Result: ILC (early) 11.8% vs DLC (delayed) 34.4% — final peer-reviewed Ann Surg 2013 PubMed abstract.

    Statistical interpretation: Primary endpoint favored early surgery; not converted into a blanket early-surgery recommendation for every patient.

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Acute cholecystitis; per the PubMed abstract ILC n=304, DLC n=314 (618 in total). The abstract does not state whether these are randomized or analysed patients; not resolved because the Ann Surg full text was not accessible (the 644 in the ClinicalTrials.gov record is the estimated/planned enrolment). Early surgery was defined as within 24 hours of hospital admission, not from symptom onset. At least 48 hours of moxifloxacin for infection was reported in both arms.

Intervention

Immediate laparoscopic cholecystectomy (ILC) — ≤24 hours from admission

Comparator

Initial antibiotics and delayed elective laparoscopic cholecystectomy (DLC) — 7–45 days

Results

Primary 75-day relevant morbidity ILC 11.8% vs DLC 34.4% (final Ann Surg 2013 abstract; the abstract gives no p value or confidence interval). Mean total stay 5.4 vs 10.0 days (p<0.001). Mean total hospital cost €2919 vs €4262 (p<0.001; period-specific costs; not presented as current Türkiye costs). No significant difference reported for conversion to open surgery or mortality; absolute rates not invented because the abstract did not provide detailed rates.

Sample size

618

Registry

NCT00447304

PMID

24022431

DOI

10.1097/SLA.0b013e3182a1599b

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract (PMID 24022431). Full-text methods / funding / COI were not checked. Verification level: abstract (not full text). 2026-11-09: meeting-abstract values whose source could not be found in any accessible record were removed from the card; the randomized/analysed split of the 618 patients was not settled.

Limits

Does not show that every high-risk patient unfit for surgery needs immediate operation. Timing depends on clinical status, infection severity, operative risk, and local capability. Funding/COI/surgeon experience not verified from full text (empty). Source level: PubMed abstract. Unresolved source detail (not verified from the primary source; not a scientific conflict): whether the 618 patients (304/314) are the randomized or the analysed population could not be confirmed from accessible sources.

Related separate studies

✓ Physician-approved · 9 Nov 2026Gallbladder · Pancreatitis and HPB

PONCHO

After mild gallstone pancreatitis, should the gallbladder be removed during the same hospital stay?

Death or readmission for a gallstone problem within six months (per 100 patients, approximately)
Time point: 6 months (from randomization)Orange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With surgery during the same admission, death or readmission for gallstone-related problems occurred in about 12 fewer people per 100.

The difference was statistically significant.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For patients who had mild pancreatitis caused by gallstones (mild biliary pancreatitis), the study asked whether the gallbladder should be removed during the same hospital stay or about one month later.

What did it find?

Within six months, death or readmission for a new gallstone-related problem occurred in about 5 in 100 patients operated on during the same stay and about 17 in 100 in the group whose surgery was delayed by a month. Each group had one bile leak and one bleed; no deaths resulted. A separate economic evaluation suggested same-stay surgery may also be cost-advantageous; this does not transfer directly to other health systems.

What does it mean for me?

In patients recovered from mild pancreatitis and fit for surgery, removing the gallbladder during the same stay can reduce recurrent gallstone problems.

Which patients does it apply to?

Patients recovered from mild biliary pancreatitis, with an inflammation marker (C-reactive protein, CRP) below 100 mg/L, no need for strong painkillers, eating normally and expected to go home within 48 hours. Under the American Society of Anesthesiologists (ASA) classification, patients over 75 with severe systemic disease (ASA III), patients with severe systemic disease that is a constant threat to life (ASA IV), and those with chronic pancreatitis or heavy alcohol use were not included; it does not apply to severe pancreatitis.

Show scientific details

SURG-PONCHO

Full scientific title

PONCHO — same-admission cholecystectomy for mild biliary pancreatitis

Clinical question

In selected patients recovering from mild gallstone pancreatitis, does same-admission cholecystectomy reduce the 6-month composite of recurrent biliary events/death versus delaying surgery about 25–30 days after discharge?

Themes

Timing and strategy · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Death or readmission for gallstone-related complications at 6 months (composite)

Study outcomes (by outcome)

  • Death or readmission for gallstone-related complications

    Time point: 6 months

    Result: Same-admission 6/128 (5%) vs interval 23/136 (17%); RR 0.28 (95% CI 0.12–0.66); p=0.002. Analyzed n=264 (randomized 266).

    Statistical interpretation: Superiority for same-admission timing; not generalizable to severe necrotizing pancreatitis.

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

23 Dutch centres; mild gallstone pancreatitis; 266 randomized (same-admission 129, interval 137); 264 analyzed (128/136). Eligibility: recovered; CRP <100 mg/L; no opioid analgesia need; tolerating normal diet; discharge expected within 48 hours. Excluded ASA III and age >75, all ASA IV, chronic pancreatitis, and ongoing heavy alcohol use. Not generalizable to severe necrotizing pancreatitis.

Intervention

Same-admission laparoscopic cholecystectomy — within 3 days after randomization

Comparator

Interval cholecystectomy — 25–30 days after randomization following discharge

Results

Primary composite (6-month death or readmission for gallstone-related complications): same-admission 6/128 (5%) vs interval 23/136 (17%); RR 0.28 (95% CI 0.12–0.66); p=0.002. Safety: one bile leak and one postoperative bleed in each arm (four serious treatment-related adverse events total); no deaths from these — not absolute safety or equal surgical-risk proof.

Sample size

266

Registry

ISRCTN72764151

PMID

26460661

DOI

10.1016/S0140-6736(15)00274-3

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract (PMID 26460661). BJS 2016 (PMID 27517163) is a 6-month cost-effectiveness analysis of the same randomized group; a separate follow-up publication — not a long-term update. Full-text COI not checked. Verification level: abstract.

Funding

Dutch Digestive Disease Foundation (Lancet PubMed abstract).

Limits

Does not justify same-admission surgery for severe necrotizing pancreatitis, unrecovered patients, or all high operative-risk patients. The BJS 2016 economic analysis is not long-term clinical follow-up; costs must not be generalized directly to different health systems. Author COI from the Lancet full text was not verified in this pack (empty). Surgeon experience unverified. Source level: PubMed abstract (+ BJS 2016 abstract, secondary publication).

Related separate studies

Physician-approved (conditional) · 9 Nov 2026Appendicitis

APPAC

Can uncomplicated appendicitis be treated with antibiotics alone, without surgery?

Why conditional approval?

The main analysis could not show that antibiotics are “not worse” than surgery, and the full text of the 2015 primary paper could not be accessed.

Verified source: Verified at the level of the 2015 primary publication abstract (the JAMA 2015 full text could not be accessed); the full text of the 10-year follow-up publication (PMC) was reviewed.

Conditions kept on this card: Non-inferiority of antibiotics was not shown; age 18–60, CT diagnosis, appendicolith excluded; the surgery arm was open appendectomy; the antibiotic regimen is historical; details of exclusion criteria, funding and conflicts of interest were not verified.

Key finding

Antibiotics were not shown to stay within the prespecified acceptance margin. Separately: most patients who started with antibiotics avoided surgery in the first year, but over ten years about 44 in 100 eventually had surgery.

Why was this study done?

For appendicitis shown to be uncomplicated on a computed tomography (CT) scan, the question was whether antibiotics can be enough instead of surgery.

What did it find?

In the surgery group, the operation was completed successfully in almost all 100 of 100 patients. In the antibiotic group, about 27 in 100 still had surgery within the first year and about 73 did not. Antibiotics were not shown to stay within the prespecified acceptance margin; this also does not prove antibiotics are worse. Over ten years, about 44 in 100 patients who started with antibiotics eventually had surgery.

What did longer follow-up show? Over ten years, about 44% of patients who started with antibiotics eventually had their appendix removed. The true recurrence rate confirmed by examining the removed tissue (about 38%) is considered separately from this figure.

What does it mean for me?

For some selected adults, antibiotics may be an alternative to immediate surgery, but recurrence and later need for surgery over the years are real possibilities. Because the surgery group had open surgery, results cannot be compared directly with today’s keyhole surgery.

Which patients does it apply to?

Adults aged 18–60 with uncomplicated appendicitis confirmed on CT. People with a stone in the appendix (appendicolith), perforation, abscess or suspected tumour were not included.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-APPAC

Full scientific title

APPAC — antibiotics versus open appendectomy for uncomplicated appendicitis

Clinical question

In CT-confirmed uncomplicated acute appendicitis (appendicolith excluded), is antibiotic therapy noninferior to open appendectomy for predefined treatment success within the noninferiority margin?

Themes

De-escalation · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Surgery arm: successful completion of appendectomy; antibiotics arm: avoiding surgery at initial discharge and no recurrence in year 1 (noninferiority margin 24%)

Study outcomes (by outcome)

  • Treatment success (noninferiority)

    Time point: 1 year (antibiotics-arm definition)

    Result: Surgical success 272/273 (99.6%). Antibiotics arm: appendectomy within 1 year 70/256 (27.3%); 186/256 (72.7%) without appendectomy in year 1. ITT treatment-efficacy difference −27.0% (95% CI −31.6 to ∞; p=0.89). Noninferiority (24% margin) was not shown — this does not prove antibiotics are inferior.

    Statistical interpretation: primary endpoint not met; noninferiority not demonstrated.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

6 Finnish hospitals; CT-confirmed uncomplicated acute appendicitis; ages 18–60; N=530 (open appendectomy 273, antibiotics 257). Appendicolith, perforation, abscess, and suspected tumor were excluded.

Intervention

Antibiotics: 3 days IV ertapenem then 7 days oral levofloxacin + metronidazole (historical protocol; not an automatic current regimen)

Comparator

Open appendectomy (surgery arm)

Results

Primary noninferiority (24% margin) was not shown. Surgical success 99.6%; antibiotics arm 1-year appendectomy 27.3% / without appendectomy 72.7%. Analysis denominators may differ from randomized N. Do not present as ‘antibiotics noninferior to surgery’.

Long-term update

2018 (PMID 30264120; PubMed abstract): antibiotics arm 5-year cumulative appendicitis recurrence 39.1% (100/256; 95% CI 33.1–45.3) — historical definition; not merged with 2026 histopathology-confirmed true recurrence as one endpoint. 2026 (PMID 41563747; PMC full text): not a new randomization. Antibiotics-arm recurrence assessment 253/257 (98.4%). Histopathology-confirmed true recurrence 37.8% (87/230; 95% CI 31.6–44.1). 10-year cumulative appendectomy 44.3% (112/253; 95% CI 38.2–50.4) — these two percentages are not the same. Complication metrics (27.4% surgery / 8.5% antibiotics) must not claim superiority over contemporary laparoscopy without noting the open surgery arm and follow-up definitions.

Sample size

530

Registry

NCT01022567

PMID

26080338

DOI

10.1001/jama.2015.6154

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

APPAC 2015 JAMA full text (PMID 26080338); 2018 PubMed abstract (PMID 30264120); 2026 PMC full text (PMID 41563747). Publication-level verification; funding/conflicts of interest not verified from the primary source. Final check (2026-11-09): the JAMA 2015 full text could not be accessed in this round; the primary endpoint, N (273/257), NI difference (−27.0%; 95% CI −31.6 to ∞; p=0.89; margin 24%) and year-1 appendectomy 70/256 were re-verified against the primary abstract. Details of the exclusion criteria (appendicolith, perforation, abscess, suspected tumour) could not be verified from the abstract in this round. 10-year PMC full text (source level 2).

Limits

Ages 18–60; CT; appendicolith excluded; open appendectomy surgery arm; historical antibiotic regimen. Not for acute deterioration/peritonitis/sepsis. Do not mix 2018 and 2026 recurrence definitions. Patient selection/primary endpoint/technique differ from CODA; not framed as contradiction. Funding/conflicts of interest not verified from the primary source.
Physician-approved (conditional) · 14 Nov 2026Appendicitis

CODA

Does antibiotic treatment for appendicitis worsen short-term health compared with surgery?

Why conditional approval?

The 40%/46%/49% figures on this card are the share of patients who eventually had their appendix removed, not recurrence rates; the full texts of the papers could not be read.

Verified source: Verified at the level of the primary publication abstract and an ACS 2021 bulletin; the NEJM full texts could not be read.

Conditions kept on this card: The 40%/46%/49% values are cumulative appendectomy rates, not recurrence; author conflict-of-interest statements were not verified.

Key finding

General health at 30 days was not unacceptably worse with antibiotics, but a substantial share of patients given antibiotics later had surgery anyway.

Why was this study done?

The study asked whether antibiotic treatment for appendicitis makes a patient’s short-term general health unacceptably worse compared with surgery (appendectomy).

What did it find?

At 30 days, patient-reported general health was similar and antibiotics stayed within the prespecified acceptance margin, meaning they were not unacceptably worse for 30-day general health. Of every 100 patients given antibiotics, about 29 still had surgery within 90 days. Patients with a stone in the appendix (appendicolith) on imaging were a separate risk group: among those given antibiotics, about 41 in 100 with a stone and about 25 in 100 without had surgery in the first 3 months. In patients with a stone, complications were clearly more frequent with antibiotics than with surgery; in those without a stone there was no clear difference. With longer follow-up, of every 100 patients who started with antibiotics, about 40 in total had surgery within one year and about 46 within two years. These numbers are the total of operations for any reason; they do not only reflect appendicitis coming back.

What did longer follow-up show? Of patients who started with antibiotics, about 40% in total had surgery within one year and about 46% within two years. This is the total of operations for any reason; it is not the same as the rate of appendicitis coming back. The three-to-four-year rate (about 49%) is based on follow-up of fewer patients.

What does it mean for me?

Antibiotics may be an alternative to immediate surgery for some patients, but this result does not show avoidance of surgery, complete recovery or equal recurrence risk. Imaging findings and the patient’s condition matter.

Which patients does it apply to?

Adults diagnosed with appendicitis at 25 centres in the USA; patients with an appendicolith were also included.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-CODA

Full scientific title

CODA — antibiotics versus appendectomy for appendicitis (EQ-5D)

Clinical question

In selected adults, is antibiotic therapy noninferior to appendectomy for 30-day EQ-5D general health status (margin 0.05)?

Themes

De-escalation · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

30-day EQ-5D general health status (noninferiority margin 0.05 points)

Study outcomes (by outcome)

  • EQ-5D general health status

    Time point: Day 30

    Result: Antibiotics noninferior: mean EQ-5D 0.92 vs 0.91; difference 0.01 (95% CI −0.001 to 0.03). This endpoint does not show avoidance of surgery, complete recovery, or equal recurrence.

    Statistical interpretation: Primary endpoint met: non-inferiority shown for 30-day EQ-5D.

    Endpoint met?: Met

  • SecondaryAppendectomy by 90 days (secondary)

    Time point: 90 days

    Result: Antibiotics arm 29% appendectomy; 41% with appendicolith vs 25% without. NSQIP complications 8.1 vs 3.5 per 100 patients; the increase was largely attributed to the appendicolith subgroup. Serious adverse events 4.0 vs 3.0 per 100 — not presented alone as proven superiority or harm.

    Statistical interpretation: Appendicolith subgroup excess risk is not omitted; not conflated with the primary EQ-5D result.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

25 US centers; N=1552 adults (776 antibiotics, 776 surgery); 414 patients with appendicolith included. Patient selection differs from APPAC.

Intervention

Antibiotics protocol for 10 days

Comparator

Appendectomy (96% of the surgery arm laparoscopic)

Results

30-day EQ-5D noninferior (0.92 vs 0.91; difference 0.01). Antibiotics-arm appendectomy by 90 days 29% (appendicolith 41% / without 25%). NSQIP 8.1 vs 3.5 per 100; serious adverse events 4.0 vs 3.0 per 100.

Long-term update

2021 longer-term research letter (N Engl J Med 2021;385(25):2395-2397; PMID 34694761; same cohort, not a separate RCT): cumulative appendectomy in the antibiotics group 40% at 1 year, 46% at 2 years; limited longer-term data 49% at years 3–4. Rates verified against the American College of Surgeons press release (25 Oct 2021); the letter full text and its confidence intervals could not be verified in this review. These are cumulative appendectomy rates for any reason, not appendicitis recurrence rates. The year 3–4 value rests on restricted follow-up. No direct superiority comparison with APPAC 10-year data.

Sample size

1552

Registry

NCT02800785

PMID

33017106

DOI

10.1056/NEJMoa2014320

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed records of the NEJM 2020 primary paper and the NEJM 2021 long-term letter; for the long-term rates, the American College of Surgeons press release of 25 October 2021; ClinicalTrials.gov record. The NEJM 2020 and 2021 full texts could not be read in this review. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Patient-Centered Outcomes Research Institute (PCORI). Funding information does not replace the authors' conflict-of-interest disclosures.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Noninferiority applies only to 30-day EQ-5D. Appendicolith excess risk is not hidden. Open-label; patient-reported outcome; crossover possible. Different selection/primary/technique from APPAC — results are not pooled. Author COI list could not be verified from the primary source. Longer-term rates (40% / 46% / 49%) are cumulative appendectomy, not recurrence rates; confidence intervals from the 2021 letter could not be verified and are not shown on the card.
Physician-approved (conditional) · 9 Nov 2026Diverticulitis

AVOD

Are antibiotics really needed for uncomplicated diverticulitis?

Why conditional approval?

Finding no difference between treatment with and without antibiotics does not mean they are equivalent; the full text of the primary paper could not be accessed.

Verified source: Verified at the level of the primary publication abstract (PubMed); the primary full text could not be accessed.

Conditions kept on this card: “no difference” is not equivalence; not generalizable to complicated diverticulitis (abscess, free perforation, fistula, peritonitis, sepsis); the full protocol endpoint definition, exclusion criteria, antibiotic regimen, funding and conflicts of interest were not verified.

Key finding

No significant difference was found in complications or recurrence between observation without antibiotics and antibiotics; this does not prove they are equal.

Why was this study done?

For uncomplicated left-sided attacks of inflammation of colon pouches (diverticulitis) confirmed by computed tomography (CT), the question was whether antibiotics are really necessary.

What did it find?

Complications such as perforation or abscess occurred in about 2 in 100 patients without antibiotics and about 1 in 100 with antibiotics; the difference was not significant. A recurrent attack needing hospital admission within the first year occurred in about 16 in 100 in both groups. At eleven years, recurrence was about 31 in 100 in both groups.

What did longer follow-up show? At a median of about eleven years, recurrent diverticulitis was about 31% in both groups. These long-term rates must not be mixed with the trial’s first-year results.

What does it mean for me?

For uncomplicated diverticulitis, close monitoring without antibiotics may be an option for some suitable patients. No significant difference does not prove the two approaches are equal.

Which patients does it apply to?

Patients in Sweden and Iceland with uncomplicated left-sided colonic diverticulitis confirmed by CT. It does not apply to complicated diverticulitis such as abscess, perforation or fistula.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-AVOD

Full scientific title

AVOD — antibiotics versus no antibiotics for uncomplicated diverticulitis

Clinical question

In CT-confirmed uncomplicated acute left-sided diverticulitis, how does supportive care without antibiotics compare with antibiotic therapy for complications and recurrence?

Themes

De-escalation · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Recovery without complications after 12 months of follow-up (wording from the primary abstract; full protocol definition not verified from the primary source)

Study outcomes (by outcome)

  • Complications (perforation or abscess)

    Time point: Primary publication period (abstract)

    Result: No antibiotics 6/309 (1.9%) vs antibiotics 3/314 (1.0%); p=0.302. Low event counts; interpret power/CIs cautiously.

    Statistical interpretation: No significant difference shown; not proof of equivalence.

    Endpoint met?: Unknown / unverified

  • Recurrence requiring readmission in year 1

    Time point: 1 year

    Result: About 16%; no significant between-group difference (p=0.881). Not conflated with 2019 11-year rates.

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Sweden and Iceland, 11 surgical departments; CT-proven uncomplicated acute left-sided colonic diverticulitis; 623 randomized (antibiotics 314, no antibiotics 309). Not automatically generalized to abscess, free perforation, or fistula.

Intervention

Supportive care without antibiotics

Comparator

Antibiotic therapy

Results

Abstract: complications 1.9% vs 1.0% (p=0.302); median stay 3 days each arm; ~16% year-1 readmission recurrence (p=0.881). No significant difference; no equivalence claim. Exact primary protocol wording → whether the primary endpoint was met is unclear.

Long-term update

Isacson 2019 (PMID 31386199; same cohort; not a separate RCT): 556/623 (89.2%) followed at median 11 years. No antibiotics / antibiotics: recurrence 31.3% / 31.3% (p=0.986); complications 4.4% / 5.0% (p=0.737); diverticulitis surgery 6.2% / 7.1% (p=0.719). Not conflated with 2012 one-year endpoints. EQ-5D response-rate differences limit QoL interpretation.

Sample size

623

Registry

NCT01008488

PMID

22290281

DOI

10.1002/bjs.8688

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstract (PMID 22290281 primary; PMID 31386199 long-term). Full-text primary definition/funding/conflicts of interest not verified from the primary source. Verification level: abstract. Check (2026-11-09): the primary endpoint was aligned with the wording in the primary abstract, "recovery without complications after 12 months of follow-up"; 623 patients (antibiotics 314 / no antibiotics 309), complications 6/309 vs 3/314 p=0.302, recurrence 16% p=0.881 re-verified against the abstract. Primary full text not accessible (source level 2).

Limits

Not generalized to complicated diverticulitis (abscess, free perforation, fistula, peritonitis, sepsis). ‘No difference’ ≠ equivalence. Primary protocol wording, exclusions/immunosuppression, antibiotic regimen, funding/conflicts of interest not verified from the primary source from full text. Source level: PubMed abstract. Not conflated with DIABOLO / APPAC / 03C-C surgical cards.
Physician-approved (conditional) · 14 Nov 2026Diverticulitis

DIABOLO

Does observation without antibiotics delay recovery from a first diverticulitis attack?

Why conditional approval?

Whether the “not worse” criterion for observation without antibiotics was met, and the number of patients in each group, could not be verified from the source.

Verified source: Verified at the level of the primary publication abstract and the protocol publication in PMC.

Conditions kept on this card: Whether the non-inferiority criterion was met and the per-arm numbers (262/266) could not be verified.

Key finding

Recovery time did not differ significantly and hospital stay was shorter with observation; this does not prove the two treatments are equal.

Why was this study done?

For a first attack of uncomplicated left-sided diverticulitis (inflammation of colon pouches) confirmed by computed tomography (CT), the study asked whether close observation without antibiotics delays recovery compared with antibiotic treatment.

What did it find?

Median time to full recovery was 14 days with observation and 12 days with antibiotics; there was no significant difference. At six months, most outcomes such as complications, recurrence, surgery and death also showed no significant difference. Hospital stay was shorter in the observation group. At about two years, recurrence was about 15 in 100 in both groups.

What did longer follow-up show? At about two years, the rate of recurrent diverticulitis was similar in both groups (about 15%). This does not prove the two treatments are equal.

What does it mean for me?

For suitable patients with a first uncomplicated diverticulitis attack, observation without antibiotics may be an option. This does not mean the two treatments are proven to be equal.

Which patients does it apply to?

Patients at 22 Dutch centres with a first left-sided diverticulitis attack confirmed as uncomplicated on CT. It does not apply to widespread inflammation of the abdominal lining, unstable patients or all abscesses.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-DIABOLO

Full scientific title

DIABOLO — observation versus antibiotics for first uncomplicated diverticulitis

Clinical question

In a first episode of CT-proven left-sided acute diverticulitis (protocol Hinchey 1a and selected 1b), how does observation versus antibiotics affect time to full recovery at six months?

Themes

De-escalation · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Time to full recovery during 6 months of follow-up. Protocol (Ünlü 2010, PMID 20646266) non-inferiority design: recovery in the observation strategy must not exceed antibiotics by more than 5 days; tested with a one-sided 95% CI of the Cox HR; 264 patients per arm, 85% power. Whether the primary paper met this NI criterion could not be verified from the primary source.

Study outcomes (by outcome)

  • Time to full recovery

    Time point: 6-month follow-up

    Result: Observation median 14 days (IQR 6–35); antibiotics median 12 days (IQR 7–30). Recovery HR 0.91; one-sided 95% CI lower bound 0.78; p=0.151. Formal NI result not verified in this review → whether the primary endpoint was met cannot be determined. No equivalence/superiority claim.

    Statistical interpretation: No significant difference shown; formal NI incomplete in pack.

    Endpoint met?: Unknown / unverified

  • Secondary6-month secondary clinical events (abstract)

    Time point: 6 months

    Result: Observation vs antibiotics: complicated diverticulitis 3.8% vs 2.6% (p=0.377); ongoing 7.3% vs 4.1% (p=0.183); recurrence 3.4% vs 3.0% (p=0.494); sigmoid resection 3.8% vs 2.3% (p=0.323); readmission 17.6% vs 12.0% (p=0.148); mortality 1.1% vs 0.4% (p=0.432). Median stay 2 vs 3 days (p=0.006; discharge protocols affect interpretation).

    Statistical interpretation: Most secondaries without significant difference; not equivalence proof.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

22 Dutch centres; first CT-proven left-sided acute diverticulitis; 528 randomized (observation 262, antibiotics 266). Protocol Hinchey 1a–1b; selected 1b localized small pericolic abscess scope is not coded as ‘all abscesses in/out’ without full text. Not generalized to clinical instability, generalized peritonitis, or severe immunosuppression.

Intervention

Observation (close monitoring without routine antibiotics)

Comparator

Antibiotic therapy

Results

Primary: recovery time 14 vs 12 days; HR 0.91; p=0.151; formal NI not verified in this review. Most secondaries without significant difference. Do not claim equivalence proven. 24-month recurrence rates are given separately in the 2018 long-term paper.

Long-term update

van Dijk 2018 (PMID 29700480; same cohort; not a separate RCT): complete-case; observation / antibiotics recurrence 15.4% / 14.9% (p=0.885); complicated diverticulitis 4.8% / 3.3% (p=0.403); sigmoid resection 9.0% / 5.0% (p=0.085). No significant difference; underpowered to exclude small differences — not equivalence proof. Not conflated with primary 6-month recovery time.

Sample size

528

Registry

NCT01111253

PMID

27686365

DOI

10.1002/bjs.10309

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the BJS 2017 primary paper and the 24-month follow-up paper; BMC Surg 2010 protocol paper (PMC full text). The per-arm patient numbers (262 / 266) are in the primary paper's full text, which could not be read in this review. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

In the protocol paper (BMC Surg 2010) the authors declared no competing interests. The statement of the primary BJS 2017 paper could not be verified from the primary source.

Limits

Hinchey 1b / all-abscess generalization forbidden (full text). Protocol NI design (5-day margin, one-sided 95% CI of Cox HR) is documented; however, the HR threshold matching the margin and whether the primary paper met the NI criterion are not verified → whether the primary endpoint was met cannot be determined. ‘No difference’ ≠ equivalence. Funding/COI, antibiotic regimen, immunosuppression could not be verified from the primary source. Source level: PubMed abstract. Not conflated with AVOD / APPAC / LADIES–SCANDIV–DIRECT–LASER.
Physician-approved (conditional) · 9 Nov 2026Diverticulitis

LADIES–LOLA

In perforated diverticulitis, is washing out the abdomen better than removing the bowel?

Why conditional approval?

The trial was stopped early and covers only purulent peritonitis; a 1-patient discrepancy in the main analysis could not be verified.

Verified source: The 12-month results were verified from the primary publication abstract and the 36-month results from the full text of the follow-up publication (PMC).

Conditions kept on this card: Purulent peritonitis only; stopped early; lavage not being superior is not equivalence; the 36-month results are secondary and based on 77/88 patients; the 1-patient gap in the primary analysis (88/87), conflicts of interest and surgeon experience were not verified.

Patients with a major complication or death within one year (per 100 patients, approximately)
Time point: 1 yearOrange dot: a patient who had this outcomeGrey dot: a patient who did not

The difference was not significant; washout was not shown to be better. The trial was stopped early because of more adverse events with washout.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

When inflamed colon pouches (diverticulitis) perforate and pus spreads in the abdomen (purulent peritonitis), the study asked whether washing out the abdomen by keyhole surgery alone (laparoscopic lavage) is better than removing the diseased bowel segment (sigmoid resection).

What did it find?

At one year, major complications or death occurred in about 67 in 100 patients after lavage and about 60 in 100 after resection; lavage was not shown to be better. The trial was stopped early because of increased adverse events in the lavage group. At three years, the lavage group had fewer re-operations and fewer people living with a stoma; however, drains through the skin were needed more often and about half of lavage patients later had bowel surgery anyway.

What did longer follow-up show? At three years, the washout group had fewer re-operations (including stoma closure) and fewer people living with a stoma. On the other hand, drains placed through the skin were more common after washout, and about half of all washout patients (17/38) later had bowel surgery anyway. Overall complication and death rates did not differ. Because the trial stopped early, these results are not definitive.

What does it mean for me?

For this type of perforated diverticulitis, lavage alone was not better than bowel surgery. This does not mean the two methods are equal.

Which patients does it apply to?

Patients with perforated diverticulitis and pus (not stool) in the abdomen. People with stool in the abdomen, over 85, with unstable circulation or on high-dose steroids were not included.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-LADIES-LOLA

Full scientific title

LADIES–LOLA — lavage versus sigmoid resection for purulent peritonitis

Clinical question

In perforated diverticulitis with purulent peritonitis, does laparoscopic peritoneal lavage reduce the 12-month composite of major morbidity/mortality versus sigmoid resection?

Themes

Minimally invasive · Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Composite of major morbidity and mortality within 12 months

Study outcomes (by outcome)

  • Major morbidity + mortality (composite)

    Time point: 12 months

    Result: Lavage 30/45 (67%) vs resection 25/42 (60%); OR 1.28 (95% CI 0.54–3.03); p=0.58. Lavage superiority not shown. DSMB stopped early for excess events in the lavage arm.

    Statistical interpretation: Superiority not shown; not equivalence proof. Early stopping is stated in interpretation.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Perforated diverticulitis with purulent peritonitis. Fecal peritonitis, age >85, hemodynamic instability, and high-dose steroids were excluded. 90 randomized; 2 excluded for protocol violation (88). Primary analysis: lavage 45, sigmoid resection 42 (total 87); the 1-patient gap between 88 and 87 could not be verified from the abstract (not verified from the primary source). The 36-month analysis uses a separate denominator (77/88). Shares LADIES program registry NCT01317485 with DIVA but is a separate randomized question — denominators are not transferred.

Intervention

Laparoscopic peritoneal lavage

Comparator

Sigmoid resection

Results

12-month composite: lavage 67% vs resection 60%; OR 1.28 (0.54–3.03); p=0.58. Lavage superiority not shown. Early stopping (DSMB).

Long-term update

Hoek 2022 (PMID 35606544, PMC9485102 full text; same LOLA cohort, not a separate RCT; LOLA arm randomized lavage:Hartmann:primary anastomosis 2:1:1, resection arms pooled). At 36 months 77/88 patients were assessed (sigmoidectomy 39, lavage 38); all rates use these denominators. Favouring lavage: ≥1 reoperation (including stoma reversal) 17/38 (45%) vs 27/39 (69%), p=0.039; patients alive with a stoma at 36 months (Kaplan–Meier cumulative proportion, deaths censored) 11% vs 17%, log-rank p=0.0268. Against lavage: percutaneous drainage 9/38 (24%) vs 2/39 (5%), p=0.025; 17/38 (45%) lavage patients later underwent sigmoidectomy (4 for sigmoid cancer). No difference: cumulative morbidity 84% vs 72% (p=0.272), mortality 16% vs 18% (p=1.000), overall reinterventions 21/38 vs 27/39 (p=0.244). Reoperation counts include stoma reversals, so they do not show intervention burden on their own. Not conflated with the 12-month primary composite; because the trial stopped early, these secondary results are also underpowered.

Sample size

90

Registry

NCT01317485

PMID

26209030

DOI

10.1016/S0140-6736(15)61168-0

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

LOLA primary PubMed/abstract (PMID 26209030); 36-month PMID 35606544. COI/surgeon experience not verified from the primary source. Faecal peritonitis is not included in this card. Check (2026-11-09): 36-month results verified against PMC9485102 full-text Tables 2–4 (77/88; reoperation 17/38 vs 27/39 p=0.039; alive with stoma KM 11% vs 17% log-rank p=0.0268; percutaneous drainage 9/38 vs 2/39 p=0.025; sigmoidectomy after lavage 17/38; morbidity p=0.272; mortality p=1.000; total reinterventions p=0.244). Primary 12-month results from the primary abstract (source level 2). In the primary analysis 45+42=87 (88 after 2 exclusions); the 1-patient difference is not verified from the primary source.

Funding

Netherlands Organisation for Health Research and Development (primary abstract). Full author-level COI statement not verified in this review.

Limits

Purulent peritonitis only; fecal peritonitis excluded. Early stopping. No lavage superiority ≠ equivalence. DIVA denominators not transferred. 36-month results are secondary, in 77/88 patients of an early-stopped trial; the stoma rate is a KM cumulative proportion (not a raw count). The resection arm pools Hartmann and primary anastomosis. COI/surgeon experience not verified from the primary source.
Physician-approved (conditional) · 9 Nov 2026Diverticulitis

LADIES–DIVA

In surgery for perforated diverticulitis, can the bowel ends be joined straight away?

Why conditional approval?

It covers only selected, stable Hinchey III–IV patients with normal immunity; no advantage in death rates was shown.

Verified source: The primary publication and the 36-month publication were verified at abstract level; the primary full text could not be accessed.

Conditions kept on this card: Selected, stable, immunocompetent Hinchey III–IV patients only; no mortality advantage; the 36-month data are retrospective after month 12 and abstract-level; the full exclusion list, conflicts of interest and surgeon experience were not verified.

Patients living without a stoma at one year (per 100 patients, estimated)
Time point: 1 yearTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Absolute difference: With joining the bowel ends, about 23 more people per 100 were living without a stoma (estimated).

The difference was statistically significant. There was no significant difference in death or complications; this does not mean better survival.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

When inflamed colon pouches (diverticulitis) perforate with pus or stool in the abdomen, the study asked whether joining the bowel ends immediately after removing the diseased segment (primary anastomosis) is better than Hartmann’s procedure, where the ends are not joined and a stoma (an opening of the bowel on the abdominal wall) is made.

What did it find?

At one year, about 95 in 100 patients in the joining group and about 72 in 100 in the Hartmann group were living without a stoma. There was no significant difference in complications (about 39 vs 44 in 100) or death (about 6 vs 3 in 100). In the joining group, a temporary protective stoma could be made if needed.

What did longer follow-up show? At three years (119 of 130 patients), being stoma-free was again more common in the joining group (about 92 vs 81 in 100). Overall complication and death rates did not differ; hernia around the stoma was more common after Hartmann’s procedure. This three-year rate is not the same measure as the one-year result, and data after month 12 were collected retrospectively.

What does it mean for me?

In suitably selected patients, joining the bowel ends can increase the chance of being stoma-free at one year. This does not mean better survival.

Which patients does it apply to?

Selected patients with perforated diverticulitis and widespread inflammation of the abdominal lining who are circulation-stable, not immunosuppressed and under 85. It does not apply to severely ill or unstable patients.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-LADIES-DIVA

Full scientific title

LADIES–DIVA — Hartmann versus primary anastomosis in Hinchey III–IV

Clinical question

In selected stable, immunocompetent Hinchey III or IV perforated diverticulitis, does primary anastomosis (with diverting stoma when needed) improve 12-month stoma-free survival versus Hartmann’s procedure?

Themes

Surgical technique · Organ preservation · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

12-month stoma-free survival

Study outcomes (by outcome)

  • Stoma-free survival

    Time point: 12 months

    Result: Primary anastomosis 94.6% (95% CI 88.7–100) vs Hartmann 71.7% (60.1–83.3); HR 2.79 (95% CI 1.86–4.18); log-rank p<0.0001. Stoma-free survival benefit is not mortality superiority.

    Statistical interpretation: Primary endpoint favored; no mortality-superiority claim.

    Endpoint met?: Met

  • Post-index morbidity / mortality

    Result: Morbidity PA 25/64 (39%) vs Hartmann 29/66 (44%), p=0.60; mortality PA 4/64 (6%) vs Hartmann 2/66 (3%), p=0.44. Diverting stoma optional in PA; 17/64 (27%) had no stoma.

    Statistical interpretation: No significant superiority shown for mortality/morbidity.

    Endpoint met?: Not met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Hinchey III or IV; selected hemodynamically stable, immunocompetent patients <85 years. 133 randomized (Hartmann 68, PA 65); mITT Hartmann 66, PA 64. Not generalized to unstable/severe patients. Not conflated with LOLA’s purulent lavage question.

Intervention

Primary anastomosis (diverting stoma when needed)

Comparator

Hartmann’s procedure

Results

12-month stoma-free survival PA 94.6% vs Hartmann 71.7%; HR 2.79; p<0.0001. Morbidity/mortality differences not significant. Stoma-free survival ≠ mortality superiority.

Long-term update

Edomskis 2022 (PMID 35026462; same DIVA; not a separate RCT): 119/130 mITT followed (PA 57, Hartmann 62); first 12 months prospective, months 12–36 retrospective. 36-month stoma-free rates PA 92%, Hartmann 81% (HR 2.326; 95% CI 1.538–3.517; log-rank p<0.0001; abstract level). Stoma reversal PA 31/40 (78%) vs Hartmann 45/61 (74%), p=0.814. Cumulative morbidity PA 21/57 (36%) vs Hartmann 30/62 (48%), p=0.266; mortality 6/57 vs 7/62 (11% / 11%), p=1.000. Parastomal hernia more frequent after Hartmann: 10/62 (16%) vs 1/57 (2%), p=0.009. Total hospital stay PA 14 vs Hartmann 17 days, p=0.025. These figures differ in definition/time point from 12-month stoma-free survival estimates; not presented as the same metric.

Sample size

133

Registry

NCT01317485

PMID

31178342

DOI

10.1016/S2468-1253(19)30174-8

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

DIVA primary PubMed (PMID 31178342); 36-month PMID 35026462 (abstract level). COI/surgeon experience not verified from the primary source. 12-month stoma-free survival ≠ 36-month stoma-free rate. Final check (2026-11-09): re-verified against the primary and 36-month abstracts; 36-month secondary endpoints (stoma reversal, morbidity, mortality, parastomal hernia, hospital stay) added from the abstract. Primary full text not accessible (source level 2).

Funding

Netherlands Organisation for Health Research and Development. Full author-level COI statement not verified in this review.

Limits

Selected stable/immunocompetent Hinchey III–IV only. No mortality superiority. 36-month stoma-free rate ≠ 12-month stoma-free survival. Not conflated with LOLA. 36-month data are retrospective after month 12 and abstract-level. Selection (stable/immunocompetent) is stated in the abstract interpretation; full exclusion list not verified from the primary source. COI/surgeon experience not verified from the primary source.
Physician-approved (conditional) · 9 Nov 2026Diverticulitis

SCANDIV

In perforated diverticulitis, does washing out the abdomen reduce serious complications?

Why conditional approval?

Washout (lavage) not being better does not mean the two methods are equivalent; in faecal peritonitis, bowel resection was required by the protocol.

Verified source: The primary results were verified from the publication abstract and the accessible results section, and the 5-year results from the full text of the follow-up publication (PMC).

Conditions kept on this card: No superiority is not equivalence; resection was protocol-mandated for fecal peritonitis; the 199 randomized patients differ from the long-term analysis population; funding and conflicts of interest were not verified.

Patients with a serious complication within 90 days (per 100 patients, approximately)
Time point: 90 daysOrange dot: a patient who had this outcomeGrey dot: a patient who did not

The difference was not significant; washout was not shown to be better. After washout, re-operation and new attacks were more common, and living with a stoma was less common.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In perforated diverticulitis (inflammation of colon pouches), the study asked whether washing out the abdomen by keyhole surgery (laparoscopic lavage) reduces serious complications compared with removing the diseased bowel (resection).

What did it find?

At 90 days, serious complications occurred in about 31 in 100 patients after lavage and about 26 in 100 after resection; lavage was not shown to be better. In patients without stool in the abdomen, re-operation was more common after lavage (about 20 vs 6 in 100). In four lavage patients, a diagnosis of bowel cancer was delayed. At five years, recurrent diverticulitis and unplanned operations were more common after lavage, while living with a stoma was less common.

What did longer follow-up show? Over five years, lavage was linked with more recurrent diverticulitis and unplanned procedures, while living with a stoma may be less likely. These risks are weighed together.

What does it mean for me?

Lavage was not better than resection at reducing serious complications; the two have different risks and benefits. The choice depends on the patient’s condition and findings at surgery.

Which patients does it apply to?

Patients in Sweden and Norway with free air in the abdomen on a computed tomography (CT) scan who had emergency surgery for suspected perforated diverticulitis. Those with stool in the abdomen had resection per protocol.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-SCANDIV

Full scientific title

SCANDIV — lavage versus resection for perforated diverticulitis

Clinical question

In suspected perforated diverticulitis (free air on CT; need for emergency surgery), does laparoscopic lavage reduce 90-day serious complications versus resection?

Themes

Minimally invasive · Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Serious complications at 90 days (Clavien–Dindo >IIIa)

Study outcomes (by outcome)

  • Serious complications (Clavien–Dindo >IIIa)

    Time point: 90 days

    Result: Lavage 31/101 (30.7%) vs resection 25/96 (26.0%); absolute difference +4.7 points (95% CI −7.9 to +17.0); p=0.53. Superiority not shown; not equivalence proof.

    Statistical interpretation: primary endpoint not met; no superiority.

    Endpoint met?: Not met

  • Reoperation in non–fecal peritonitis subgroup

    Result: Lavage 15/74 (20.3%) vs resection 4/70 (5.7%), p=0.01. Protocol mandated resection for fecal peritonitis. Delayed detection of four sigmoid malignancies after lavage is noted in the paper.

    Statistical interpretation: Reoperation signal against lavage (selected subgroup).

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

21 centres (Sweden/Norway); 199 randomized (lavage 101, resection 98; primary analysis resection 96). Suspected perforated diverticulitis, free air on CT, need for emergency surgery. Protocol-mandated resection for fecal peritonitis. Independent of LADIES (NCT01047462).

Intervention

Laparoscopic lavage

Comparator

Resection (protocol-mandated for fecal peritonitis)

Results

90-day serious complications 30.7% vs 26.0%; p=0.53 — no superiority. Reoperation against lavage in non-fecal subgroup. Delayed malignancy detection noted.

Long-term update

Azhar 2021 (PMID 33355658; same cohort; not a separate RCT): true purulent-perforation subpopulation lavage 73, resection 69 (not conflated with 199 randomized). Serious complications 36% vs 35% (p=0.92); diverticulitis recurrence 21% vs 4% (p=0.004); stoma prevalence among survivors 4/50 (8%) vs 17/52 (33%) (p=0.002); unplanned reoperation 19/73 (26%) vs 8/69 (12%) (p=0.03); total secondary operations including stoma reversal 36% vs 35% (p=0.92); 21 lavage patients (30%) later underwent sigmoid resection; overall mortality 32% vs 25% (p=0.36). The 5-year primary endpoint is severe complications; lower stoma prevalence alone is not treatment superiority. Trade-offs among clinical risks; no one-sided superiority claim.

Sample size

199

Registry

NCT01047462

PMID

26441181

DOI

10.1001/jama.2015.12076

Physician approval

Conditionally approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

SCANDIV primary PubMed + accessible results section (PMID 26441181); 5-year PMID 33355658. Funding/conflicts of interest not verified from the primary source. 199 randomized ≠ 142 in the long-term analysis population. Final check (2026-11-09): 5-year results verified against PMC7758831 full-text Table 2 (stoma 4/50 vs 17/52; unplanned reoperation 19/73 vs 8/69; total secondary operations p=0.92; sigmoid resection after lavage 30%; mortality p=0.36). Primary 2015 results from the abstract (source level 2).

Limits

No superiority ≠ equivalence. Protocol resection for fecal peritonitis. 199 ≠ 5-year analysis N. Funding/conflicts of interest not verified from the primary source. No blind synthesis with LOLA. Clinical advice not generalized beyond the population.
Physician-approved (conditional) · 14 Nov 2026Diverticulitis

DIRECT

Does planned surgery improve quality of life in recurrent diverticulitis?

Why conditional approval?

This is a small trial (109 patients) that was stopped early; 46% of patients in the non-surgical group eventually had surgery.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Small sample (109); early stopping; 46% of patients in the conservative arm eventually had surgery.

Key finding

Planned surgery significantly improved quality of life and the difference persisted at five years, but some operated patients had a leak at the bowel join.

Why was this study done?

For patients with recurring diverticulitis (inflammation of colon pouches) or ongoing complaints after an attack, the study asked whether planned removal of the last part of the colon (elective sigmoid resection) improves quality of life compared with non-surgical treatment.

What did it find?

At six months, digestive quality-of-life scores were significantly higher in the surgery group. Among patients who had surgery, about 15 in 100 had a leak at the bowel join. The trial was stopped early because recruitment was difficult. At five years, quality of life was still better in the surgery group; about half of those who started without surgery eventually had an operation.

What did longer follow-up show? At five years, quality of life tended to remain higher in the surgery group; about half of those who started with non-surgical treatment may eventually have had surgery. This long-term result is considered separately from the main six-month result.

What does it mean for me?

For selected patients whose complaints affect quality of life, planned surgery can improve it, but it carries important risks such as leaks and re-intervention. It does not mean surgery is needed after every attack.

Which patients does it apply to?

Patients with recurrent left-sided diverticulitis or persistent complaints after an attack.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-DIRECT

Full scientific title

DIRECT — elective sigmoid resection versus conservative care for recurrent/persistent diverticulitis

Clinical question

In recurrent left-sided diverticulitis or ongoing complaints after an episode, does elective sigmoid resection improve 6-month GIQLI versus conservative treatment?

Themes

Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Gastrointestinal Quality of Life Index (GIQLI) at 6 months

Study outcomes (by outcome)

  • GIQLI

    Time point: 6 months

    Result: Surgery 114.4 vs conservative 100.4; mean difference 14.2 (95% CI 7.2–21.1); p<0.0001. Trial stopped early because of recruitment difficulty.

    Statistical interpretation: 6-month QoL benefit shown; early stopping is stated.

    Endpoint met?: Met

  • Anastomotic leak (patients who underwent surgery)

    Result: Anastomotic leak in 7 of 47 patients who underwent surgery (15%). This percentage is not recalculated on the 53 randomized denominator.

    Statistical interpretation: Safety risk is not omitted; not conflated with randomized N.

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Recurrent left-sided diverticulitis or ongoing complaints after a prior episode. 109 randomized: sigmoid resection 53, conservative 56. Not pooled with LASER or LADIES–SCANDIV acute peritonitis populations.

Intervention

Elective sigmoid resection

Comparator

Conservative treatment

Results

6-month GIQLI favored surgery (difference 14.2; p<0.0001). Early stopping. Anastomotic leak 7/47 (15%) among those operated. Does not mean routine surgery after every attack.

Long-term update

5-year DIRECT follow-up (PMID 30247329; same cohort; not a separate RCT): mean GIQLI surgery 118.2 vs conservative 108.5; difference 9.7 (95% CI 1.7–17.7). Among those assigned to conservative care, 26 (46%) eventually required surgery. Follow-up reports anastomotic leak 8 (11%) and reintervention 11 (15%) among operated patients — different denominator/follow-up from primary 7/47 (15%); not mixed. Not presented as the same metric as 6-month primary GIQLI.

Sample size

109

Registry

NTR1478

PMID

28404008

DOI

10.1016/S2468-1253(16)30109-1

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the Lancet Gastroenterol Hepatol 2017 primary paper, the 5-year follow-up paper and the economic analysis. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Netherlands Organisation for Health Research and Development (primary paper abstract).

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Open-label; early stopping; crossover; selected population. Not pooled with LASER. Distinct from LADIES/SCANDIV acute peritonitis. COI/technique/surgeon experience could not be verified from the primary source. Source level: PubMed abstract (+ follow-up abstract).
✓ Physician-approved · 14 Nov 2026Diverticulitis

LASER

Does planned surgery improve quality of life in recurrent or painful diverticulitis?

Key finding

Planned surgery improved quality of life and reduced new attacks in the first six months; at four years the quality-of-life difference was no longer significant.

Why was this study done?

For patients with recurring, complicated or persistently painful diverticulitis (inflammation of colon pouches), the study asked whether planned removal of the last part of the colon (elective sigmoid resection) improves quality of life compared with non-surgical care.

What did it find?

At six months, improvement in quality of life was significantly greater in the surgery group. About 10 in 100 operated patients had a major complication. Within six months, a new attack occurred in about 5 in 100 after surgery and about 31 in 100 without surgery. At four years the quality-of-life difference was no longer statistically significant; about a third of those who started without surgery had an operation within four years.

What did longer follow-up show? At two and four years, new attacks may be less common in the surgery group; at four years the quality-of-life difference was not statistically significant. Some of those who started with non-surgical treatment may eventually have had surgery.

What does it mean for me?

In selected patients, planned surgery may improve quality of life and reduce attacks in the first months, but it carries complication risks and a long-term quality-of-life difference was not established.

Which patients does it apply to?

Patients at 6 Finnish centres with recurrent, complicated or persistently painful diverticulitis.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-LASER

Full scientific title

LASER — elective resection versus conservative care for recurrent/complicated/persistent painful diverticulitis

Clinical question

In recurrent, complicated, or persistently painful diverticulitis, does elective sigmoid resection improve the 6-month change in GIQLI versus conservative care?

Themes

Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Difference in GIQLI change from baseline to 6 months (surgery − conservative)

Study outcomes (by outcome)

  • Difference in GIQLI change

    Time point: 6 months

    Result: Surgery–conservative change difference 11.96 points (95% CI 3.72–20.19); p=0.005. Primary analysis n=72 (37/35). Not conflated with 90 randomized or absolute 6-month GIQLI level.

    Statistical interpretation: 6-month QoL-change benefit shown; not conflated with 4-year neutral QoL.

    Endpoint met?: Met

  • 6-month major complications / new attacks (clinical ITT)

    Time point: 6 months

    Result: Clinical ITT 85 (41/44). Major complications surgery 4/41 (10%). New diverticulitis attack surgery 2/41 (5%) vs conservative 12/44 (31%).

    Statistical interpretation: Safety and recurrence signals kept separate from primary GIQLI change.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

6 Finnish centres; recurrent, complicated, or persistently painful diverticulitis. 90 randomized (45/45). Primary 6-month GIQLI analysis 72 (37/35); clinical ITT 85 (41/44). Not conflated with DIRECT or LADIES–SCANDIV.

Intervention

Elective sigmoid resection

Comparator

Conservative care

Results

6-month GIQLI change difference +11.96 (p=0.005) — not absolute level. Major complications 10% surgery. New attacks 5% vs 31% (clinical ITT). Do not summarize as long-term surgical superiority.

Long-term update

2y (PMID 37074706): GIQLI assessed in 70 patients; ITT 2-year mean GIQLI reported similar between groups (per-protocol separate). Diverticulitis recurrence 25/41 (61%) conservative vs 4/37 (11%) surgery (these follow-up denominators). 4y 2025 (PMID 40202724): mean GIQLI 115.3 vs 109.8; difference 5.54 (95% CI −2.98 to 14.06) — no statistically significant difference. Conservative arm: 14/44 (32%) had sigmoid resection by 4 years. Primary 6-month QoL benefit and 4-year neutral QoL are shown as separate time points; no ‘long-term superiority’ claim. New-attack measurement definitions may differ from the primary report.

Sample size

90

Registry

NCT02174926

PMID

33206182

DOI

10.1001/jamasurg.2020.5151

Physician approval

Approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts and PMC full texts of the JAMA Surg 2020 primary paper and the 2- and 4-year follow-up papers. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Vatsatautien Tutkimussäätiö Foundation, Mary and Georg Ehrnrooth’s Foundation, Martti I. Turunen Foundation, The Finnish Medical Foundation and Helsinki University Hospital research funds. According to the paper, the funders had no role in design, analysis or writing (publication statement).

Conflicts of interest

Author disclosures (JAMA Surg full texts): some authors reported foundation research grants; in the follow-up papers one author reported personal fees from Johnson & Johnson outside the submitted work.

Limits

Keep 90 ≠ 72 ≠ 85 ≠ 70 denominators distinct. 6-month benefit ≠ significant 4-year QoL difference. Not pooled with DIRECT; distinct from LADIES/SCANDIV acute peritonitis. Funding and author conflict-of-interest statements are taken from the JAMA Surg full texts (PMC).
Physician-approved (conditional) · 14 Nov 2026Perioperative safety

STOP-IT

Once an abdominal infection is controlled, is a short course of antibiotics enough?

Why conditional approval?

The result on a shorter antibiotic course applies only to patients whose source of infection was adequately controlled by surgery or drainage; the trial did not reach its planned size.

Verified source: The full text of the primary publication (NEJM 2015) was reviewed.

Conditions kept on this card: The result applies only to patients with adequate source control; the sample size planned for equivalence was not reached; the content of the 2018 erratum could not be read.

Wound infection, recurrent infection or death within 30 days (per 100 patients, approximately)
Time point: 30 daysOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Applies only to patients whose infection source has been adequately controlled by surgery or a drain. Results were similar, but because the trial did not reach its planned size, the two approaches were not formally shown to be equal.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

After the source of an abdominal infection has been controlled by surgery or a drain, the study asked whether antibiotics should be given for a short fixed time (about 4 days) or longer, until fever and other signs resolve.

What did it find?

Within 30 days, wound infection, recurrent abdominal infection or death occurred in about 22 in 100 patients with short treatment and about 22 in 100 with longer treatment. Median antibiotic duration was 4 vs 8 days. Because the trial did not reach its planned size, the two approaches have not been formally shown to be equivalent.

What does it mean for me?

This result applies only to patients whose infection source has been adequately controlled by surgery or a drain; in these patients a shorter antibiotic course may be adequate. Whether source control is adequate is decided by the surgical team. This does not mean antibiotics must always stop on day four.

Which patients does it apply to?

Patients with complicated abdominal infection whose source has been adequately controlled by surgery or drainage. Treatment is reassessed if source control is inadequate, the patient deteriorates or the immune system is suppressed.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-STOP-IT

Full scientific title

STOP-IT — short-course versus conventional antibiotics after source-controlled intra-abdominal infection

Clinical question

After adequate source control for complicated intra-abdominal infection, how does a fixed 4±1-day antibiotic course compare with treatment continued until 2 days after physiologic resolution (maximum 10 days) for the 30-day composite of surgical-site infection / recurrent IAI / death?

Themes

Perioperative safety · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

30-day composite of surgical-site infection or recurrent intra-abdominal infection or death

Study outcomes (by outcome)

  • Composite: SSI / recurrent IAI / death

    Time point: 30 days after source control

    Result: Short course 56/257 (21.8%) vs control 58/260 (22.3%); absolute difference −0.5 percentage points (95% CI −7.0 to +8.0); p=0.92. Prespecified equivalence sample size was not reached — no claim of formal equivalence/NI success.

    Statistical interpretation: No significant difference was found in the primary endpoint; because the trial did not reach its planned size, equivalence cannot be claimed.

    Endpoint met?: Unknown / unverified

  • Antibiotic duration

    Result: Median 4 days (IQR 4–5) vs 8 days (IQR 5–10); duration difference −4 days, p<0.001.

    Statistical interpretation: Duration shortened; not proof of clinical-event equivalence.

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Complicated intra-abdominal infection after adequate anatomic source control (surgery or percutaneous). 518 patients were randomized (experimental 258, control 260). One patient withdrew consent after randomization; the primary analysis was intention-to-treat in the 517 patients who consented and were randomized (experimental 257, control 260). Not automatically generalized to failed source control, persistent infection, immunosuppression, or clinical deterioration. Not conflated with AVOD/DIABOLO (whether to start antibiotics).

Intervention

Fixed 4±1 calendar days of antibiotics after source control

Comparator

Antibiotics until 2 days after resolution of fever, leukocytosis, and ileus (maximum 10 days)

Results

In patients with adequate source control, the 30-day composite outcome was 21.8% vs 22.3% (56/257 vs 58/260); difference −0.5 points (95% CI −7.0 to +8.0); p=0.92. Median antibiotic duration 4 vs 8 days (difference −4.0 days; −4.7 to −3.3). In the control arm, antibiotics continued until 2 days after resolution of physiological abnormalities (temperature <38.0 °C, white-cell count <11,000/mm³ and able to meet more than half of nutritional needs orally/enterally), maximum 10 days. Protocol-duration adherence 211/258 (81.8%) in the experimental arm and 189/260 (72.7%) in the control arm. Among patients treated per protocol (a prespecified subgroup), the primary outcome was 17.5% vs 17.5% (37/211 vs 33/189; difference 0.1 points, 95% CI −7.4 to 7.6; P=0.72). The equivalence design required 505 patients per group; enrolment stopped before this target. There is no formal equivalence or non-inferiority claim.

Sample size

518

Registry

NCT00657566

PMID

25992746

DOI

10.1056/NEJMoa1411162

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract and PMC full text of the NEJM 2015 primary paper (PMC4469182); ClinicalTrials.gov NCT00657566. The content of the 2018 erratum could not be read; the primary numbers were not changed. Information that could not be verified is stated explicitly under “Limitations”.

Funding

National Institutes of Health (NIH) (primary paper).

Conflicts of interest

NEJM publication statement: “No other potential conflict of interest relevant to this article was reported.” Detailed author-level disclosures are in the ICMJE forms on the NEJM website and were not read in this review.

Limits

Equivalence sample-size target not met → no formal equivalence/NI. Adequate source control is mandatory. Limited immunosuppression data. Not conflated with AVOD/DIABOLO. Author-level COI details (ICMJE forms) and protocol appendix could not be verified from the primary source; a publication-level COI statement is in the full text. 258 is the randomized experimental arm; 257 is the analysis population after the 1 patient who withdrew consent (inferred from the patient numbers in Table 1 and Table 2; the text does not state this separately). Primary numbers unchanged pending verified erratum impact.
✓ Physician-approved · 9 Nov 2026Pancreatitis and HPB

PANTER

For infected dead pancreatic tissue: a drain first, or open surgery straight away?

Patients with a major complication or death (per 100 patients, approximately)
Time point: From first intervention to 3 months after discharge (per trial protocol)Orange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With the step-up approach, major complications or death occurred in about 29 fewer people per 100.

The difference was statistically significant. Death alone was similar in both groups.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

When dead pancreatic tissue becomes infected after pancreatitis (infected pancreatic necrosis), it was often removed directly by major open surgery (open necrosectomy). The study asked whether a step-up approach, placing a drain first and removing dead tissue through small incisions only if needed, is better.

What did it find?

Major complications or death occurred in about 40 in 100 patients with the step-up approach and about 69 in 100 with open surgery. About one third of step-up patients never needed surgery to remove dead tissue. Death alone was similar in both groups (about 19 vs 16 in 100). In long-term follow-up, incisional hernia and pancreatic insufficiency were less common after step-up.

What does it mean for me?

In infected pancreatic necrosis, major open surgery is not always needed; a step-up approach can reduce complications. A survival advantage was not shown. Decisions are made with an experienced multidisciplinary team.

Which patients does it apply to?

Patients with suspected or confirmed infected pancreatic necrosis who need intervention. It does not apply to non-infected necrosis or other emergencies.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-PANTER

Full scientific title

PANTER — step-up versus open necrosectomy for infected necrotizing pancreatitis

Clinical question

In suspected or confirmed infected pancreatic/peripancreatic necrosis requiring intervention, does a drainage-first step-up strategy (with VARD if needed) reduce major complications or death versus primary open necrosectomy?

Themes

Minimally invasive · Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Major complication or death (composite; study definition)

Study outcomes (by outcome)

  • Major complication or death

    Time point: From the first intervention after randomization to 3 months after discharge — per the published trial protocol (PMID 16606471); not verified in the final NEJM 2010 publication (not verified from the primary source)

    Result: Step-up 17/43 (40%) vs open necrosectomy 31/45 (69%); RR 0.57 (95% CI 0.38–0.87); p=0.006.

    Statistical interpretation: Composite reduced with step-up; primary endpoint met. Not a mortality-superiority claim.

    Endpoint met?: Met

  • Death (alone)

    Result: 19% vs 16% (step-up / open); p=0.70 — no significant mortality difference.

    Statistical interpretation: Mortality superiority not shown.

    Endpoint met?: Not met

  • Drainage alone (step-up arm)

    Result: 35% of the step-up arm were treated with drainage alone without surgical necrosectomy.

    Statistical interpretation: Secondary finding; part of strategy success.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Suspected or confirmed infected pancreatic/peripancreatic necrosis requiring intervention. Multicenter Netherlands; 88 randomized (step-up 43, open necrosectomy 45). Not generalized to sterile necrosis or separate emergencies needing immediate intervention. Not conflated with POINTER (drainage timing).

Intervention

Step-up: primarily percutaneous (rarely endoscopic) drainage; video-assisted retroperitoneal debridement (VARD) if needed

Comparator

Primary open necrosectomy

Results

Primary composite: step-up 17/43 (40%) vs open 31/45 (69%); RR 0.57 (95% CI 0.38–0.87); p=0.006. Drainage alone in 35% of step-up. New-onset multiorgan failure 12% vs 40%. Death 19% vs 16% (p=0.70) — no mortality superiority. Long-term results are in a separate field.

Long-term update

Hollemans 2019 (PMID 30391468; same cohort; not a separate RCT; PubMed abstract): mean follow-up 86±11 months. Of 88 originally randomized, 73 survivors after the index admission were the evaluation target — 73 ≠ primary N=88. Composite death/major complication from index admission through long-term: step-up 44% vs open 73% (p=0.005). Incisional hernia 23% vs 53%; exocrine insufficiency 29% vs 56%; endocrine insufficiency 40% vs 64%. Additional drainage 11% vs 13% (p=0.99); pancreatic surgery 11% vs 5% (p=0.43). Do not merge these percentages with the 2010 primary composite rates. Long-term funding/COI and full-text follow-up scope not verified from the primary source.

Sample size

88

Registry

ISRCTN13975868

PMID

20410514

DOI

10.1056/NEJMoa0908821

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Primary NEJM full text + PubMed (PMID 20410514). Long-term: PubMed abstract (PMID 30391468). Person-level COI not verified from the primary source. Long-term funding/conflicts of interest not verified from the primary source. Primary endpoint assessment period (2026-11-09 check): the NEJM 2010 PubMed abstract does not state the period; the NEJM full text could not be accessed in this check (HTTP 403). The period was taken only from the published protocol (Besselink, BMC Surg 2006; PMID 16606471; PMC1508161): complications occurring “from the first intervention after randomization until three months after discharge”; primary endpoint follow-up is completed 3 months after discharge and secondary endpoints after 6 months. Not verified in the final publication → not verified from the primary source.

Funding

Dutch Organization for Health Research and Development (grant 945-06-910) — primary full text. Funding of the long-term article not verified from the primary source.

Conflicts of interest

Primary author disclosure forms are available; person-level COI details were not verified in this review (not verified from the primary source). No inference of ‘no COI for all authors’ is made. Long-term funding/conflicts of interest not verified from the primary source.

Limits

Strategy comparison ≠ standalone VARD superiority. No generalization to sterile necrosis / separate emergencies. No mortality superiority. Not merged with POINTER (timing). 2019 long-term: 73 survivors ≠ 88 randomized. Primary person-level COI and long-term funding/conflicts of interest not verified from the primary source. Long-term source level: PubMed abstract. Unresolved source detail (not verified from the primary source; not a scientific conflict): the assessment period of the primary endpoint is taken only from the published trial protocol (from the first intervention after randomization to 3 months after discharge); not verified in the final NEJM 2010 publication.

Related separate studies

Physician-approved (conditional) · 14 Nov 2026Pancreatic and periampullary tumours

LEOPARD-2

Does keyhole pancreatic head surgery give a faster recovery?

Why conditional approval?

The trial was stopped early for safety reasons; deaths within 90 days were higher in the keyhole (laparoscopic) surgery group.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Stopped early for safety reasons; 90-day mortality was higher in the laparoscopic arm; the sponsor's role was not verified.

Key finding

The keyhole method did not give a faster recovery; the trial was stopped early because of a safety concern over deaths in the keyhole group.

Why was this study done?

Whipple surgery (pancreatoduodenectomy) for diseases of the head of the pancreas and surrounding area is a complex operation. The study asked whether doing it by keyhole (laparoscopic) surgery gives faster recovery than open surgery.

What did it find?

Return to daily functioning was not faster with keyhole surgery (median about 10 vs 8 days). Among 99 operated patients, complication-related death within 90 days occurred in 5 of 50 in the keyhole group and 1 of 49 in the open group. This difference was not statistically conclusive, but the trial was stopped early because of this safety concern. There was no significant difference in major complications or pancreatic leak.

What does it mean for me?

Small incisions alone do not mean safer surgery or faster recovery. This trial did not show keyhole Whipple to be better and raised a safety signal. The choice depends on the patient, team experience and centre.

Which patients does it apply to?

Adults planned for Whipple surgery without blood-vessel involvement, at 4 Dutch centres doing at least 20 Whipple operations a year. It does not apply to robotic Whipple or to operations on the tail of the pancreas.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-LEOPARD-2

Full scientific title

LEOPARD-2 — laparoscopic versus open pancreatoduodenectomy (Whipple)

Clinical question

In candidates for pancreatoduodenectomy without vascular involvement, does laparoscopic PD shorten time to functional recovery versus open PD, and is it safe?

Themes

Minimally invasive · Surgical technique · Technology comparison

Study design

Pragmatic randomized controlled trial

Primary endpoint

Time to functional recovery (phase 3; oral analgesia, independent mobility, >50% calorie intake, no IV fluids, no infection signs)

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • PrimaryTime to functional recovery (phase 3 primary)

    Result: Median 10 days (95% CI 5–15) LPD vs 8 days (95% CI 7–9) OPD; log-rank p=0.80. Faster recovery not demonstrated.

    Statistical interpretation: Phase-3 superiority not shown → primary endpoint not met. p>0.05 is not equivalence.

    Endpoint met?: Not met

  • 90-day complication-related mortality (operated patients)

    Result: LPD 5/50 (10%) vs OPD 1/49 (2%); RR 4.90 (95% CI 0.59–40.44); p=0.20. Numerical safety signal; not statistically conclusive mortality difference. DSMB stopped the trial early for this imbalance.

    Statistical interpretation: p=0.20 is not proof of safety; early stopping is clinically important. Also not proof that laparoscopy is definitively more lethal in all centres.

    Endpoint met?: Unknown / unverified

  • Clavien–Dindo ≥III major complications

    Result: 25/50 (50%) LPD vs 19/49 (39%) OPD; RR 1.29 (95% CI 0.82–2.02); p=0.26.

    Statistical interpretation: No significant difference; no equivalence claim.

    Endpoint met?: Unknown / unverified

  • Grade B/C postoperative pancreatic fistula

    Result: 14/50 (28%) vs 12/49 (24%); RR 1.14 (95% CI 0.59–2.22); p=0.69.

    Statistical interpretation: No significant difference; no equivalence claim.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Netherlands; 4 centres (≥20 PD/year). Adult PD candidates (benign/premalignant/malignant) without signs of vascular involvement. 105 randomized; 99 operated (LPD 50, OPD 49); 6 not operated — excluded from surgical-outcome denominators per protocol. 99 is not an unqualified ITT denominator. Phase-2 interim (42 randomized; 3/20 vs 0/20 deaths) must not be merged with the final cohort. Not conflated with LEOPARD distal pancreatectomy or robotic Whipple.

Intervention

Laparoscopic pancreatoduodenectomy (LPD)

Comparator

Open pancreatoduodenectomy (OPD)

Results

Early stopping: DSMB terminated the trial early for complication-related mortality imbalance. Among 99 operated: 90-day complication-related death 5/50 vs 1/49 (RR 4.90; p=0.20) — numerical safety signal, not conclusive mortality inferiority/superiority. Functional recovery 10 vs 8 days (p=0.80) — no faster recovery. C-D ≥III 25/50 vs 19/49; POPF B/C 14/50 vs 12/49. Phase-2 interim mortality (3/20 vs 0/20) must not be merged with the final cohort.

Sample size

105

Registry

NTR5689

PMID

30685489

DOI

10.1016/S2468-1253(19)30004-4

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the Lancet Gastroenterol Hepatol 2019 primary paper and the university summary page. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Investigator-initiated study grant from Johnson & Johnson Medical Limited (primary paper abstract). The sponsor's role in design, analysis and writing could not be verified from the primary source.

Conflicts of interest

Author-level conflict-of-interest disclosures could not be verified from the primary source. Funding information does not mean there were no conflicts.

Limits

DSMB early stopping must remain visible. p=0.20 is not proof of safety; also not proof of definitive excess mortality everywhere. 105 randomized ≠ 99 operated. Phase 2 ≠ final cohort. ≠ LEOPARD distal ≠ robotic. Learning-curve completion not proven. Protocol stopping rules / full COI could not be verified from the primary source. No verified long-term follow-up of the same cohort; the card gives no long-term update.
✓ Physician-approved · 9 Nov 2026Liver · Colon and rectal cancer (incl. its spread)

Oslo-COMET

Is keyhole liver surgery safer for bowel cancer that has spread to the liver?

Patients with a complication in the 30 days after surgery (per 100 patients, approximately)
Time point: 30 daysOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With keyhole surgery, about 12 fewer complications per 100 patients.

The difference was statistically significant. There was no significant difference in five-year survival; this is not proof of equal cancer control.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In patients whose bowel cancer has spread to the liver and can be removed surgically, the study asked whether keyhole (laparoscopic) surgery that spares as much liver tissue as possible reduces complications compared with open surgery.

What did it find?

In the first 30 days, complications occurred in about 19 in 100 patients after keyhole surgery and about 31 in 100 after open surgery; the difference was significant. Postoperative hospital stay was shorter after keyhole surgery (about 53 vs 96 hours; roughly 2 vs 4 days). At five years, about 54 vs 55 in 100 patients were alive; there was no significant difference.

What does it mean for me?

In suitable patients at experienced centres, keyhole surgery may mean fewer complications and a shorter stay. No five-year difference does not prove the two methods are equal for cancer control; the trial was not designed to show that.

Which patients does it apply to?

Selected patients at a single Norwegian centre with liver metastases removable while sparing liver tissue. Operations removing half of the liver, needing blood vessel or bile duct reconstruction, or combined with tissue burning (ablation) were not included.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-OSLO-COMET

Full scientific title

Oslo-COMET — laparoscopic versus open parenchyma-sparing resection for colorectal liver metastases

Clinical question

In resectable colorectal liver metastases, does parenchyma-sparing laparoscopic resection reduce 30-day Accordion grade ≥2 complications versus open resection?

Themes

Minimally invasive · Surgical technique · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Accordion grade ≥2 postoperative complications within 30 days

Study outcomes (by outcome)

  • Accordion ≥2 complications (30 days)

    Result: Laparoscopy 19% vs open 31%; difference 12 percentage points (95% CI 1.67–21.8); p=0.021 — favours laparoscopy.

    Statistical interpretation: Primary superiority met → primary endpoint met. This field refers only to 30-day complications; not oncologic superiority.

    Endpoint met?: Met

  • Hospital stay

    Result: 53 vs 96 hours; p<0.001.

    Statistical interpretation: Secondary; not the primary endpoint.

    Endpoint met?: Met

  • 90-day mortality

    Result: Laparoscopy 0 vs open 1 — do not claim a mortality advantage.

    Statistical interpretation: No mortality-superiority claim.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Oslo University Hospital; single-centre; assessor-blinded. Resectable colorectal liver metastases. 280 randomized: laparoscopic parenchyma-sparing 133, open 147. Formal hemihepatectomy, vascular/biliary reconstruction, and resection combined with ablation were protocol exclusions — not equivalent to all major/complex hepatectomies. Not conflated with TRANSMET (transplantation).

Intervention

Laparoscopic parenchyma-sparing liver resection

Comparator

Open parenchyma-sparing liver resection

Results

Primary: Accordion ≥2 19% vs 31% (p=0.021). Hospital stay 53 vs 96 hours (p<0.001). 90-day deaths 0 vs 1. Economic analysis is not primary; cross-country cost extrapolation forbidden. Long-term OS/RFS in a separate field; no equivalence/NI claim.

Long-term update

Aghayan 2021 (PMID 33197213; same 280 randomized; not a separate RCT; median follow-up 70 months). Five-year overall survival 54% vs 55%; HR 0.93 (95% CI 0.67–1.30); p=0.67. Five-year recurrence-free survival 30% vs 36%; HR 1.09 (95% CI 0.80–1.49); p=0.57. No statistical difference ≠ proven equivalence or non-inferiority; the trial was not designed for long-term NI/power (~10-point 5-year OS differences may not be excluded). 2022 Ann Oncol 411P conference abstract must not be coded as a peer-reviewed long-term primary. Long-term funding (abstract): South-Eastern Norway Regional Health Authority — not auto-copied as 2018 primary funding.

Sample size

280

Registry

NCT01516710

PMID

28657937

DOI

10.1097/SLA.0000000000002353

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed primary/long-term + NCT + protocol PMC. Primary 2018 funding/conflicts of interest and surgeon experience not verified from the primary source. 2022 conference abstract is a secondary source.

Funding

2021 long-term abstract: South-Eastern Norway Regional Health Authority. Detailed primary 2018 funding/COI and sponsor role → not verified from the primary source (long-term funding was not presented as primary funding).

Conflicts of interest

Person-level COI (2018 and 2021) not fully verified in this review → not verified from the primary source.

Limits

Selected parenchyma-sparing resections; hemihepatectomy/reconstruction/ablation combinations excluded. Single centre. Surgeon experience not verified from the primary source. Primary complication benefit ≠ oncologic equivalence/NI. 2022 conference ≠ peer-reviewed long-term. ≠ TRANSMET. Primary funding/conflicts of interest not verified from the primary source.
Physician-approved (conditional) · 14 Nov 2026Liver · Colon and rectal cancer (incl. its spread)

TRANSMET

Can a liver transplant extend life when bowel cancer has spread only to the liver and cannot be removed?

Why conditional approval?

The trial was done in a highly selected group of patients; the full-text tables and organ-allocation conditions were not verified.

Verified source: Verified at the level of the primary publication abstract and the feasibility publication in PMC.

Conditions kept on this card: Highly selected patient group; full-text tables and organ-allocation conditions were not verified.

Patients alive at five years (per 100 patients, estimated)
Time point: 5 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Absolute difference: With transplant + chemotherapy, about 44 more people per 100 were alive (estimated; uncertainty is wide in this 94-patient trial).

In 94 very carefully selected patients. Transplantation is major surgery and requires lifelong immune-suppressing medication.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In some colon or rectal cancers, the disease spreads only to the liver but cannot be completely cleared by removing part of the liver. The study asked whether liver transplantation plus drug treatment (chemotherapy) gives longer survival than chemotherapy alone.

What did it find?

In 94 very carefully selected patients, about 57 in 100 were alive at five years in the planned-transplant group versus about 13 in 100 with chemotherapy alone; the transplant strategy was shown to be superior. Serious side effects were common in both groups.

What does it mean for me?

Transplantation is not suitable for every patient with liver metastases. It is a major operation and requires lifelong immune-suppressing medication. Eligibility depends on organ availability and the decision of an experienced transplant team and expert board.

Which patients does it apply to?

Patients aged 18–65 in good general health, with the bowel tumour already removed, disease confined to the liver and not removable by surgery, without a particular gene change (BRAF mutation), and responding to chemotherapy.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-TRANSMET

Full scientific title

TRANSMET — liver transplantation plus chemotherapy versus chemotherapy alone for permanently unresectable CRLM

Clinical question

In selected patients with permanently unresectable liver-only colorectal metastases, does liver transplantation plus chemotherapy improve 5-year overall survival versus chemotherapy alone?

Themes

Timing and strategy · Organ preservation · Evidence that changed practice

Study design

Pragmatic randomized controlled trial

Primary endpoint

Five-year overall survival (ITT)

Study outcomes (by outcome)

  • 5-year overall survival — ITT

    Result: Transplantation + CT 56.6% (95% CI 43.2–74.1) vs CT alone 12.6% (95% CI 5.2–30.1); HR 0.37 (95% CI 0.21–0.65); p=0.0003. Primary randomized analysis.

    Statistical interpretation: ITT survival benefit → primary endpoint met. Do not substitute PP for ITT.

    Endpoint met?: Met

  • 5-year overall survival — per-protocol (separate)

    Result: PP 73.3% (95% CI 59.6–90.0) vs 9.3% (95% CI 3.2–26.8). Labeled per-protocol; not ITT.

    Statistical interpretation: Additional analysis; not conflated with ITT.

    Endpoint met?: Met

  • Serious adverse events — as-treated safety

    Result: Serious adverse events in 32/40 (80%) transplanted vs 45/54 (83%) chemotherapy-only (69 events — persons ≠ event count). Similar percentages ≠ safety equivalence. Three retransplantations; one postoperative multiorgan-failure death.

    Statistical interpretation: As-treated safety denominator; not ITT/PP. No equivalent-safety claim.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

20 European tertiary centres; open-label RCT. Inclusion: ages 18–65; ECOG 0–1; primary colorectal tumour resected; permanently unresectable liver metastases; no extrahepatic disease; BRAF-wild-type; responsive systemic therapy; ≥3 months and ≤3 lines of chemotherapy; independent multidisciplinary board approval. ITT 94 (47/47); PP 74 (36/38); did not receive assigned treatment TX 11 / CT 9. As-treated safety: transplanted 40, CT-only 54 — not mixed with ITT/PP. Not auto-generalized to Turkish organ allocation or all CRLM patients. ≠ Oslo-COMET (resection technique).

Intervention

Liver transplantation plus chemotherapy strategy

Comparator

Chemotherapy alone

Results

Median follow-up 59.3 months (IQR 42.4–60.2). ITT 5-year OS 56.6% vs 12.6%; HR 0.37; p=0.0003. PP 73.3% vs 9.3% (separate). As-treated safety 32/40 vs 45/54 serious AEs — no equivalent-safety claim. Feasibility interim: some selected patients did not reach transplant due to waiting/progression; do not conflate 38 with as-treated 40 (includes cross-over).

Sample size

94

Registry

NCT02597348

PMID

39306468

DOI

10.1016/S0140-6736(24)01595-2

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the Lancet 2024 primary paper; eClinicalMedicine 2024 feasibility paper (PMC full text). Information that could not be verified is stated explicitly under “Limitations”.

Funding

French National Cancer Institute and Assistance Publique–Hôpitaux de Paris (Lancet 2024). The eClinicalMedicine paper also reports AP-HP support.

Conflicts of interest

Lancet 2024 statement: “We declare no competing interests.” This is the paper's own declaration, not an independent audit of the authors' activities.

Limits

ITT ≠ PP ≠ as-treated. Similar serious-AE percentages ≠ safety equivalence. Organ allocation / immunosuppression / recurrence / centre experience constrain interpretation. No generalization to Turkish allocation or all CRLM. ≠ Oslo-COMET. Full-text appendices, CEA threshold, graft/centre detail could not be verified from the primary source. No long-term update on the card.
Physician-approved (conditional) · 14 Nov 2026Colorectal

COLOR II

Is keyhole surgery for rectal cancer as safe as open surgery for cancer coming back?

Why conditional approval?

The trial’s “not worse” (non-inferiority) margin could not be confirmed from the primary publication’s abstract.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: The non-inferiority margin (5 points) could not be confirmed from the primary-source abstract; author conflict-of-interest statements were not verified.

Patients whose cancer came back in the operated area within three years (per 100 patients, approximately)
Time point: 3 yearsOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Keyhole surgery: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For rectal cancer surgery, the study asked whether keyhole (laparoscopic) surgery is unacceptably worse than open surgery for cancer coming back in the operated area within three years.

What did it find?

At three years, cancer returned in or near the operated area in about 5 in 100 patients in both groups. Keyhole surgery stayed within the prespecified acceptance margin, meaning it was not meaningfully worse than open surgery. Keyhole surgery had less blood loss and a shorter hospital stay but took longer.

What does it mean for me?

Keyhole surgery can be an option for suitable patients treated by experienced teams. This does not mean keyhole surgery is superior.

Which patients does it apply to?

Patients with a single rectal cancer within 15 cm of the anal opening and no spread to distant organs. It does not directly apply to advanced tumours invading nearby organs or where the surgical margin is threatened.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-COLOR-II

Full scientific title

COLOR II — laparoscopic versus open surgery for rectal cancer

Clinical question

In selected rectal cancer, is laparoscopic resection non-inferior to open resection for three-year locoregional recurrence within a prespecified non-inferiority margin?

Themes

Minimally invasive · Surgical technique · Technology comparison · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Three-year locoregional recurrence (non-inferiority; margin 5 absolute percentage points)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Primary3-year locoregional recurrence — primary NI

    Result: Lap 5.0% vs open 5.0%; difference 0 percentage points; two-sided 90% CI −2.6 to +2.6. Upper bound +2.6 < +5 → prespecified primary NI criterion met. Analysis set 1044 (699/345); 59 exclusions after 1103 randomized stated separately.

    Statistical interpretation: The “primary endpoint met” status represents only this locoregional-recurrence NI result. No separate superiority or NI claim for DFS/OS.

    Endpoint met?: Met

  • Secondary3-year DFS (secondary; no separate NI/superiority)

    Result: 74.8% vs 70.8%; difference 4.0 percentage points (95% CI −1.9 to +9.9). Not the primary NI criterion.

    Statistical interpretation: Not presented as superiority or proven separate NI.

    Endpoint met?: Not applicable

  • Secondary3-year OS (secondary; no separate NI/superiority)

    Result: 86.7% vs 83.6%; difference 3.1 percentage points (95% CI −1.6 to +7.8). Not the primary NI criterion.

    Statistical interpretation: Not presented as superiority or proven separate NI.

    Endpoint met?: Not applicable

  • Secondary2013 short-term — pathology / perioperative (secondary publication)

    Result: Median blood loss 200 / 400 mL; operative time 240 / 188 min; bowel function 2 / 3 days; stay 8 / 9 days. Mesorectal integrity 589/666 (88%) vs 303/331 (92%); p=0.250. CRM <2 mm positivity 56/588 (10%) vs 30/300 (10%); p=0.850 — CRM threshold not auto-equated with ALaCaRT/Z6051. 28-day morbidity 278/697 (40%) vs 128/345 (37%); mortality 8/699 (1%) vs 6/345 (2%); no significant difference ≠ equivalence. Conversion ~16%. Pathology ≠ 3-year recurrence.

    Statistical interpretation: Short-term secondary outcomes of the same RCT; not a separate primary endpoint / long-term follow-up.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

International multicentre (30 centres, 8 countries), open-label phase III NI RCT — rectum only; ≠ COLOR colon. 1103 randomized (739 lap / 364 open); primary analysis 1044 (699 / 345); 59 post-randomization exclusions. Eligible analysis population: solitary rectal adenocarcinoma ≤15 cm from anal verge; no distant metastases; advanced local invasion / threatened CRM exclusions per primary publication. PME for upper rectum, appropriate TME principles otherwise — not identical TME in every case. 3-year follow-up denominators (locoregional recurrence 771/74%; DFS 923/89%; OS 903/87%) must not be conflated with 1044. No automatic generalization to T4 / threatened CRM / contemporary TNT populations.

Intervention

Laparoscopic rectal resection (TME/PME per protocol)

Comparator

Open rectal resection

Results

Primary (3-year locoregional recurrence): 5.0% / 5.0%; difference 0; 90% CI −2.6…+2.6 → NI met. DFS 74.8% / 70.8% and OS 86.7% / 83.6% are not separate superiority/NI claims. 2013 short-term: blood loss/stay favour laparoscopy, operative time favours open; CRM 10%/10% (<2 mm); mesorectum 88%/92%; morbidity/mortality without significant difference ≠ equivalence. 2019 Ann Surg bowel obstruction/hernia analysis is secondary; does not generate 5-year OS.

Sample size

1103

Registry

NCT00297791

PMID

25830422

DOI

10.1056/NEJMoa1414882

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed records of the NEJM 2015 primary paper, the Lancet Oncol 2013 short-term paper and the 2019 paper. The non-inferiority margin (5 absolute points) could not be confirmed from the primary-source abstract. Information that could not be verified is stated explicitly under “Limitations”.

Funding

NEJM 2015: Ethicon Endo-Surgery Europe and others. Lancet Oncol 2013: Ethicon Endo-Surgery Europe, Swedish Cancer Foundation, West Gothia Region, Sahlgrenska University Hospital. According to the NEJM statement, Ethicon had no role in design, data collection, analysis or writing (publication statement; not independent verification).

Conflicts of interest

Author-level conflict-of-interest disclosures and supplementary material could not be verified from the primary source.

Limits

1103 randomized ≠ 1044 analysed; 59 exclusions and follow-up attrition (recurrence 771, DFS 923, OS 903) remain for physician review. CRM <2 mm ≠ 1 mm thresholds in other trials. Primary NI (3-year recurrence) ≠ ALaCaRT/Z6051 composite pathologic NI; COLOR II’s positive NI does not invalidate those pathologic findings. ≠ COLOR colon. No T4/threatened-CRM/TNT generalization. 2013 and 2019 are secondary; no verified 5-year oncologic long-term follow-up. Full COI could not be verified from the primary source (2019 PMID 29746337 identified). NI margin (5 absolute points) could not be read from the primary-source abstract: could not be verified from the primary source. ALaCaRT and Z6051 are shown as related studies on separate cards; results are not pooled.
Physician-approved (conditional) · 14 Nov 2026Colorectal

ALaCaRT

Can keyhole surgery for rectal cancer remove the tumour as completely as open surgery?

Why conditional approval?

It could not be shown that keyhole surgery is “not worse” than open surgery; this does not mean keyhole surgery is worse.

Verified source: The primary publication was verified at abstract level; the JAMA full text could not be read.

Conditions kept on this card: Non-inferiority was not shown (composite pathological endpoint); this does not mean that laparoscopy is worse.

Key finding

Keyhole surgery could not be shown to stay within the acceptance margin for completeness of the removed tissue; this does not prove it is worse.

Why was this study done?

For rectal cancer, the study asked whether keyhole (laparoscopic) surgery is unacceptably worse than open surgery at producing ‘clean and complete’ removed tissue. ‘Clean and complete’ means the fatty tissue around the tumour is removed intact and there are no cancer cells at the cut edges.

What did it find?

This combined measure was achieved in about 82 in 100 patients after keyhole surgery and about 89 in 100 after open surgery. Keyhole surgery could not be shown to stay within the prespecified acceptance margin. This does not mean keyhole surgery is proven worse; only the statistical threshold was not reached.

What did longer follow-up show? At two years, cancer returning in the same area (about 5% and 3%), disease-free survival (80% and 82%) and overall survival (94% and 93%) were similar between groups, and the differences were not statistically significant. However, the trial was not large enough to prove the results are ‘the same’; small but important differences cannot be fully ruled out.

What does it mean for me?

This measure is based on laboratory examination of tissue; it is not directly cancer recurrence or survival. The type of operation is decided together, based on tumour location and stage, body build and the team’s experience.

Which patients does it apply to?

Patients in Australia and New Zealand with rectal cancer less than 15 cm from the anal opening that had not grown into nearby organs (stage T1–T3).

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-ALACART

Full scientific title

ALaCaRT — laparoscopic versus open surgery for rectal cancer (pathologic endpoint)

Clinical question

In T1–T3 rectal cancer, is laparoscopic-assisted resection non-inferior to open resection for composite ‘successful resection’ (complete TME + CRM ≥1 mm + distal margin ≥1 mm) within a −8% margin?

Themes

Minimally invasive · Surgical technique · Technology comparison

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Composite successful resection: complete TME + CRM ≥1 mm + distal margin ≥1 mm (non-inferiority; margin −8%)

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • PrimaryComposite successful resection — primary NI (JAMA 2015)

    Result: Lap 194 (82%) vs open 208 (89%); risk difference −7.0% (95% CI −12.4% to ∞); P=.38 for non-inferiority. Lower bound −12.4% falls below the −8% margin → NI not shown.

    Statistical interpretation: “Primary endpoint not met” means only ‘NI not demonstrated’. Laparoscopy being pathologically or clinically worse (inferiority) was not tested / not proven.

    Endpoint met?: Not met

  • SecondaryComposite components (secondary; no separate NI/superiority)

    Result: Clear CRM: 93% vs 97%; difference −3.7% (95% CI −7.6% to 0.1%); P=.06. Clear distal margin: 99% vs 99%; −0.4% (95% CI −1.8% to 1.0%); P=.67. Complete TME: 87% vs 92%; −5.4% (95% CI −10.9% to 0.2%); P=.06. Conversion to open 9%.

    Statistical interpretation: No statistically significant component difference ≠ equivalence; CIs do not exclude clinically relevant differences favouring open.

    Endpoint met?: Not applicable

  • Secondary2-year DFS and local recurrence (long-term; Ann Surg 2019; secondary)

    Result: 450 patients without baseline metastases (225/225); median follow-up 3.2 years. 2-year local recurrence (cumulative incidence): lap 5.4% vs open 3.1%; difference 2.3% (95% CI −1.5% to 6.1%); HR 1.7 (95% CI 0.74–3.9). 2-year DFS: lap 80% vs open 82%; difference (laparoscopic − open) −2.0 percentage points, favouring open (95% CI −9.3 to +5.4; the abstract prints an unsigned ‘2.0%’, the CI midpoint is consistent with −2.0); HR 1.17 (95% CI 0.81–1.68), P=.41; adjusted HR 1.07 (95% CI 0.7–1.6). 2-year OS: 94% vs 93%; difference 0.9% (95% CI −3.6% to 5.4%).

    Statistical interpretation: No significant difference ≠ equivalence: the authors note the CIs include potentially important differences favouring open resection and that longer follow-up is needed. DFS difference −2.0 points (lap − open); the CI includes zero and does not exclude a difference of up to 9.3 points favouring open. CI verified only against the PubMed/Europe PMC abstract; Ann Surg full text requires subscription → full-text confirmation could not be verified from the primary source.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

24 sites, 26 surgeons in Australia and New Zealand; phase III NI RCT (March 2010–November 2014). T1–T3 rectal adenocarcinoma <15 cm from the anal verge; 475 randomized (238 lap / 237 open). Long-term analysis: 450 patients without baseline metastases (225/225); mean age 64, 34% women. No automatic generalization to T4 / contemporary TNT populations.

Intervention

Laparoscopic-assisted rectal resection

Comparator

Open rectal resection

Results

Primary: successful resection 82% vs 89%; difference −7.0% (95% CI −12.4%…∞); P=.38 → NI not shown. This is not evidence that laparoscopy is worse. Components: CRM 93%/97%, distal 99%/99%, complete TME 87%/92% — no significant difference ≠ equivalence. Conversion 9%. Long-term (Ann Surg 2019; secondary): 2-year local recurrence 5.4%/3.1%; 2-year DFS 80%/82%; 2-year OS 94%/93% — no significant difference, but CIs do not exclude important differences favouring open.

Long-term update

Stevenson, Ann Surg 2019;269(4):596-602 (PMID 30247332): 450 patients, median follow-up 3.2 years (0.1–5.4). 2-year local recurrence lap 5.4% vs open 3.1% (difference 2.3%; 95% CI −1.5%…6.1%; HR 1.7, 95% CI 0.74–3.9). 2-year DFS 80% vs 82% (difference lap − open −2.0 points, 95% CI −9.3…+5.4; HR 1.17, 95% CI 0.81–1.68; P=.41; adjusted HR 1.07). 2-year OS 94% vs 93%. Authors: no significant difference; CIs include potentially important differences favouring open; longer follow-up needed. No equivalence claim.

Sample size

475

Registry

ACTRN12609000663257

PMID

26441180

DOI

10.1001/jama.2015.12009

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the JAMA 2015 primary paper and the Ann Surg 2019 long-term paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

The PubMed publication type indicates non-government research support; the funding body and sponsor role could not be verified from the primary source (JAMA full text not accessible).

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Primary endpoint is a pathologic surrogate, not recurrence/survival. NI not shown ≠ inferiority. Long-term analysis (450 patients, median 3.2 years) is secondary and not powered for equivalence; no significant difference ≠ equivalence. 2-year DFS difference −2.0 points (lap − open); CI verified from the abstract only, not the full text. CRM ≥1 mm ≠ COLOR II <2 mm; ALaCaRT ≠ Z6051 (population: T1–T3 <15 cm vs stage II–III ≤12 cm after neoadjuvant therapy) — results are not pooled. Funding/COI, safety/complications, neoadjuvant proportion, surgeon credentialing could not be verified from the primary source. No T4/TNT/robotic generalization.
Physician-approved (conditional) · 14 Nov 2026Colorectal

ACOSOG Z6051

After pre-operative treatment for rectal cancer, can keyhole surgery remove the tumour completely?

Why conditional approval?

It could not be shown that keyhole surgery is “not worse”; an inconsistency in one confidence interval and in the definition of local recurrence also awaits physician review.

Verified source: The full texts of the primary publications (JAMA 2015 and Ann Surg 2019) were reviewed.

Conditions kept on this card: Non-inferiority was not shown; an inconsistency in the confidence interval for serious complications and in the definition of local recurrence awaits physician review.

Key finding

Keyhole surgery could not be shown to stay within the acceptance margin for completeness of the removed tissue; this does not prove it is worse.

Why was this study done?

In patients with stage II–III rectal cancer who had received radiation and/or drug treatment before surgery, the study asked whether keyhole (laparoscopic) surgery is unacceptably worse than open surgery at producing ‘clean and complete’ removed tissue.

What did it find?

This combined measure was achieved in about 82 in 100 patients after keyhole surgery and about 87 in 100 after open surgery. Keyhole surgery could not be shown to stay within the prespecified acceptance margin; this does not mean it is proven worse. Keyhole surgery took about 45 minutes longer on average; serious complications and hospital stay were similar.

What did longer follow-up show? With about four years of follow-up, two-year disease-free survival was 79.5% after laparoscopy and 83.2% after open surgery; over the whole follow-up, cancer returning in the operated area or nearby region was about 5% in both groups (this is over the full follow-up, not at two years). The differences were not statistically significant. However, the trial was not designed to prove these results are ‘the same’.

What does it mean for me?

This measure is based on laboratory examination of tissue; it is not directly cancer recurrence or survival. The type of operation is decided together, based on tumour location and stage, body build and the team’s experience.

Which patients does it apply to?

Adults in the USA and Canada with stage II–III rectal cancer (not invading nearby organs) within 12 cm of the anal opening, who had completed treatment before surgery and had a body mass index of 34 or less.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-Z6051

Full scientific title

ACOSOG Z6051 — laparoscopic versus open surgery for stage II–III rectal cancer (pathologic endpoint)

Clinical question

In stage II–III rectal cancer after neoadjuvant therapy, is laparoscopic-assisted resection non-inferior to open resection for composite ‘successful resection’ (CRM >1 mm + tumour-free distal margin + complete TME) within a 6% margin?

Themes

Minimally invasive · Surgical technique · Technology comparison

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Composite successful resection: CRM >1 mm + tumour-free distal margin + complete TME (non-inferiority; margin 6%)

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • PrimaryComposite successful resection — primary NI (JAMA 2015)

    Result: Lap 81.7% (95% CI 76.8–86.6) vs open 86.9% (95% CI 82.5–91.4); difference −5.3% (one-sided 95% CI −10.8% to ∞); P=.41 for non-inferiority. Lower bound −10.8% falls below the −6% margin → NI not shown. Analysed 462 (240/222).

    Statistical interpretation: “Primary endpoint not met” means only ‘NI not demonstrated’. Inferiority of laparoscopy was not tested / not proven.

    Endpoint met?: Not met

  • SecondaryPathology components and perioperative / safety (secondary)

    Result: Negative CRM 87.9% vs 92.3% (P=.11); negative distal margin >98% in both arms (P=.91); TME complete 77% + nearly complete 16.5% (all 462 cases). Conversion 11.3%. Operative time 266.2 vs 220.6 min (difference 45.5; 95% CI 27.7–63.4; P<.001). Length of stay 7.3 vs 7.0 days (difference 0.3; 95% CI −0.6…1.1). 30-day readmission 3.3% vs 4.1% (difference −0.7%; 95% CI −4.2%…2.7%). Severe complications (Clavien-Dindo 3–5) 22.5% vs 22.1% (difference 0.4%; published 95% CI −4.2%…2.7% — identical to the readmission CI; possible typesetting error: could not be verified from the primary source). Grade 5: 2 (0.8%) vs 2 (0.9%).

    Statistical interpretation: No significant difference ≠ equivalence. Pathology components and safety results do not substitute for the primary NI result.

    Endpoint met?: Not applicable

  • Secondary2-year DFS; local and regional (locoregional) recurrence over follow-up — not a 2-year estimate (long-term; Ann Surg 2019; secondary)

    Result: Analysed 462 (240/222); median follow-up 47.9 months. 2-year DFS (Kaplan-Meier): lap 79.5% (95% CI 74.4–84.9) vs open 83.2% (95% CI 78.3–88.3); no P value given for the difference, text states ‘no statistical difference’. Local+regional recurrence 4.6% vs 4.5% — crude rate over follow-up, not a 2-year estimate. Text separately reports ‘LR’ 2.1% vs 1.8% (log-rank P=.86); definition inconsistent with Methods → pending physician review. Distant recurrence 14.6% vs 16.7% (crude).

    Statistical interpretation: No significant difference ≠ equivalence: the publication states the trial was not powered for the secondary endpoint. A 2-year local recurrence estimate could not be verified in the publication.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

35 institutions in the USA and Canada; multicentre NI RCT (October 2008–September 2013). Rectal adenocarcinoma ≤12 cm from the anal verge, clinical stage II–III (T4 excluded); completed neoadjuvant therapy (mostly fluorouracil-based chemoradiotherapy: lap 95.0% / open 91.2%); age ≥18, BMI ≤34, ECOG <3. 486 randomized; analysed 462 (240 lap / 222 open). LAR 76.7%, APR 23.3%. No automatic generalization to high BMI / T4 / contemporary TNT populations.

Intervention

Laparoscopic-assisted rectal resection (per protocol)

Comparator

Open rectal resection

Results

Primary: successful resection 81.7% vs 86.9%; difference −5.3% (one-sided 95% CI −10.8%…∞); P=.41 → NI not shown. This is not evidence that laparoscopy is worse. Negative CRM 87.9%/92.3%; distal >98% in both arms. Conversion 11.3%; operative time 45.5 min longer with laparoscopy; no significant difference in stay, 30-day readmission or severe complications ≠ equivalence. Long-term (Ann Surg 2019; secondary): 2-year DFS 79.5%/83.2%; local and regional (locoregional) recurrence over a median 47.9-month follow-up 4.6%/4.5% (crude rate; not a 2-year cumulative estimate); distant recurrence 14.6%/16.7% — no significant difference, no equivalence claim.

Long-term update

Fleshman, Ann Surg 2019;269(4):589-595 (PMID 30080730): 462 patients, median follow-up 47.9 months. 2-year DFS lap 79.5% (95% CI 74.4–84.9) vs open 83.2% (95% CI 78.3–88.3). Local and regional (locoregional) recurrence over follow-up 4.6% vs 4.5%; distant recurrence 14.6% vs 16.7% (crude rates; not 2-year cumulative estimates). Unsuccessful resection associated with worse DFS (HR 1.87; 95% CI 1.21–2.91) — an observational association across the cohort, not an arm comparison. Publication: no significant difference; not powered for the secondary endpoint. No equivalence claim.

Sample size

486

Registry

NCT00726622

PMID

26441179

DOI

10.1001/jama.2015.10529

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts and PMC full texts of the JAMA 2015 primary paper and the Ann Surg 2019 long-term paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

National Cancer Institute grant (ACOSOG). Support from the American Society of Colon and Rectal Surgeons for the quality-of-life comparison and from SAGES for instrumentation assessment; Covidien gave unrestricted support to the ACOSOG infrastructure. According to the paper, the sponsors had no role in design, analysis or writing (publication statement).

Conflicts of interest

JAMA 2015 full text: some authors reported relationships with device and drug companies (e.g. Covidien research support, Ethicon EndoSurgery personal fees, consultancies and honoraria). The statement of the Ann Surg 2019 paper could not be verified from the primary source.

Limits

Primary endpoint is a pathologic surrogate, not recurrence/survival. NI not shown ≠ inferiority. 486 randomized ≠ 462 analysed. Long-term results are secondary and unpowered; no significant difference ≠ equivalence. No 2-year local recurrence estimate in the publication; 4.6%/4.5% is a crude local+regional rate; ‘LR 2.1%/1.8%’ definition inconsistency. Severe complication CI possibly a typesetting error. BMI ≤34 and T4 excluded → limited generalizability. CRM >1 mm ≠ COLOR II <2 mm; Z6051 ≠ ALaCaRT — results are not pooled. Credentialing details and 2019 COI could not be verified from the primary source.
✓ Physician-approved · 14 Nov 2026Colorectal

ROLARR

Does robotic rectal cancer surgery reduce the need to switch to open surgery?

Patients switched to open surgery during the operation (per 100 patients, approximately)
Time point: During surgeryOrange dot: a patient who had this outcomeGrey dot: a patient who did not

The difference was not statistically significant; robotic superiority was not shown. Robotic surgery took longer and cost more.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For rectal cancer surgery, the study asked whether robot-assisted keyhole surgery reduces the need to switch to open surgery during the operation compared with standard keyhole (laparoscopic) surgery.

What did it find?

Switching to open surgery occurred in about 8 in 100 patients with the robot and about 12 in 100 with laparoscopy; the difference was not statistically significant, so robotic superiority was not shown. There were no significant differences in complications, clean removal of the tumour, or urinary and sexual function in the first 6 months. Robotic surgery took longer and cost more.

What does it mean for me?

This study did not show that the robot is better for every patient, nor did it prove the two methods are the same. The method is chosen according to the surgical team’s experience and the patient’s situation.

Which patients does it apply to?

Patients at 29 centres in 10 countries with rectal cancer suitable for curative surgery.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-ROLARR

Full scientific title

ROLARR — robotic-assisted versus conventional laparoscopic surgery for rectal cancer

Clinical question

In rectal cancer resection, does robotic-assisted surgery reduce the risk of conversion to open laparotomy compared with conventional laparoscopy (superiority)?

Themes

Minimally invasive · Technology comparison · Surgical technique

Study design

Multicentre randomized controlled trial

Primary endpoint

Conversion to open laparotomy (superiority hypothesis)

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • PrimaryConversion — primary

    Result: Robotic 19/236 (8.1%) vs laparoscopic 28/230 (12.2%); unadjusted risk difference 4.1% (95% CI −1.4% to 9.6%); adjusted OR 0.61 (95% CI 0.31–1.21); P=.16. Overall conversion 10.1% — far below the 25% assumed in the power calculation.

    Statistical interpretation: Superiority not shown. This does not prove equivalence; the low observed conversion rate left the trial underpowered relative to its assumptions.

    Endpoint met?: Not met

  • SecondaryPathologic quality (secondary)

    Result: CRM positivity 5.1% (12/235) vs 6.3% (14/224); difference 1.1% (95% CI −3.1% to 5.4%); adjusted OR 0.78 (0.35–1.76); P=.56. Mesorectal plane 76.4% vs 77.6%; adjusted OR 0.94 (0.56–1.57).

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • SecondarySafety (secondary)

    Result: Intraoperative complications 15.3% vs 14.8%; 30-day complications 33.1% vs 31.7%; 30-day mortality 2 vs 2 (0.8% vs 0.9%). Mean operative time 298.5 vs 261.0 min. Per-patient cost difference £980 (robot acquisition/maintenance excluded).

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • SecondaryFunctional outcomes — 6 months (secondary)

    Result: IPSS adjusted difference 0.74 (95% CI −0.59 to 2.07); P=.27. IIEF difference 0.80 (−4.10 to 5.70); P=.75. FSFI difference 1.23 (−3.54 to 6.00); P=.60. Low questionnaire completion (men 56.6%, women 35.8%).

    Statistical interpretation: Functional results are interpreted cautiously given missing data.

    Endpoint met?: Not applicable

  • Local recurrence / DFS / OS

    Result: Primary publication reports outcomes to 6 months; long-term oncologic outcomes not verified in this review.

    Statistical interpretation: No oncologic equivalence or superiority claim.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

29 sites, 10 countries, 40 surgeons; January 2011–September 2014. Rectal adenocarcinoma suitable for curative resection; 471 randomized (237 robotic / 234 laparoscopic); primary analysis 466 (236/230). Mean age 64.9; 67.9% men. High/low anterior resection or APR.

Intervention

Robotic-assisted rectal resection

Comparator

Conventional laparoscopic rectal resection

Results

Primary: conversion 8.1% vs 12.2%; P=.16 → robotic superiority not shown. CRM 5.1%/6.3%; no significant difference in mesorectal plane, complications, 30-day mortality, or 6-month bladder/sexual function (≠ equivalence). Robotic surgery was longer and costlier. The conversion difference in men (OR 0.46) is a subgroup finding only (interaction P=.09). Practice impact: no evidence of routine robotic superiority; technique choice depends on expertise, cost and patient factors.

Sample size

471

Registry

ISRCTN80500123

PMID

29067426

DOI

10.1001/jama.2017.7219

Physician approval

Approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract and PMC full text of the JAMA 2017 primary paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Efficacy and Mechanism Evaluation (EME) Programme — a Medical Research Council and National Institute for Health Research partnership. According to the paper, the funders had no role in design, analysis or writing (publication statement).

Conflicts of interest

JAMA 2017 full text: several authors reported serving as proctors for Intuitive Surgical, one reported consultancy and one reported travel support.

Limits

Observed conversion (10.1%) far below the assumed 25% → low power. High missing data in functional questionnaires. Variable surgeon robotic experience. No long-term oncologic outcomes in this review: could not be verified from the primary source. ROLARR ≠ REAL: different primary endpoints (conversion vs 3-year locoregional recurrence), populations and expertise levels; results are not pooled.
Physician-approved (conditional) · 14 Nov 2026Colorectal

REAL

Does robotic surgery for mid and low rectal cancer reduce cancer coming back?

Why conditional approval?

Far fewer events occurred than planned and some patients were excluded after randomization, so the results need careful interpretation.

Verified source: The full text of the primary publication (JAMA 2025) was reviewed.

Conditions kept on this card: The number of events was far below the planned number; there were post-randomization exclusions in the modified ITT analysis; the trial was done in a single country.

Key finding

At three years, cancer returned in the operated area significantly less often after robotic surgery; there was no survival difference. Results come from very experienced centres.

Why was this study done?

For mid and low rectal cancer, the study asked whether robot-assisted surgery reduces cancer coming back in the operated area within three years compared with standard keyhole (laparoscopic) surgery.

What did it find?

At three years, cancer returned in or near the operated area in about 2 in 100 patients after robotic surgery and about 4 in 100 after laparoscopy; robotic superiority was shown. The robotic group had fewer complications after surgery and better urinary, sexual and bowel function in the first months. Three-year survival was similar (about 95 vs 93 in 100).

What does it mean for me?

The results come from high-volume centres with very experienced surgeons; the same results may not be seen everywhere. This is a single study and did not show a survival difference. The method is decided together with the surgical team.

Which patients does it apply to?

Patients aged 18–80 at 11 Chinese centres with mid or low rectal cancer that had not grown into nearby organs (stage T1–T3) or spread to distant organs.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-REAL

Full scientific title

REAL — robotic versus laparoscopic surgery for middle and low rectal cancer

Clinical question

In middle and low rectal cancer, does robotic surgery reduce 3-year locoregional recurrence compared with laparoscopic surgery (superiority)?

Themes

Minimally invasive · Technology comparison · Surgical technique

Study design

Multicentre randomized controlled trial (superiority design)

Primary endpoint

3-year locoregional recurrence (superiority)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Primary3-year locoregional recurrence — primary (JAMA 2025)

    Result: Robotic 1.6% (95% CI 0.6–2.6) vs laparoscopic 4.0% (95% CI 2.4–5.6); HR 0.45 (95% CI 0.22–0.92); log-rank P=.03; adjusted HR 0.39 (0.19–0.80). Analysed 1171 (586/585); median follow-up 43.0 months.

    Statistical interpretation: Prespecified superiority hypothesis met. However, observed events (35) were well below the planned ~81, and the mITT analysis excludes 69 post-randomization patients → physician review.

    Endpoint met?: Met

  • SecondaryPathologic quality (short-term publication, secondary)

    Result: CRM positivity 4.0% (22/547) vs 7.2% (39/543); difference −3.2 percentage points (95% CI −6.0 to −0.4); P=.023. Macroscopically complete resection 95.4% vs 91.8%; difference 3.6 (0.8–6.5).

    Statistical interpretation: Secondary endpoint; interpreted in the context of multiple comparisons.

    Endpoint met?: Not applicable

  • SecondaryShort-term safety (secondary)

    Result: 30-day Clavien-Dindo ≥II complications 16.2% vs 23.1%; difference −6.9 (−11.4 to −2.3); P=.003. Intraoperative complications 5.5% vs 8.7%. Conversion to open 1.7% vs 3.9%. APR 16.9% vs 22.7%. Median blood loss 40 vs 50 mL; median stay 7 vs 8 days.

    Statistical interpretation: Secondary endpoints; not the primary hypothesis.

    Endpoint met?: Not applicable

  • Secondary3-year DFS (secondary)

    Result: 87.2% vs 83.4%; HR 0.74 (95% CI 0.56–0.98); log-rank P=.04; adjusted HR 0.67 (0.50–0.89).

    Statistical interpretation: Secondary endpoint; not independent confirmatory evidence.

    Endpoint met?: Not applicable

  • Secondary3-year OS (secondary)

    Result: 94.7% (92.9–96.5) vs 93.0% (90.8–95.2); HR 0.75 (0.50–1.12); P=.16.

    Statistical interpretation: No significant difference ≠ equivalence; longer follow-up may be needed.

    Endpoint met?: Not applicable

  • SecondaryFunctional outcomes (secondary)

    Result: Favouring robotic: urinary (IPSS) at 3, 6 and 12 months; male sexual (IIEF-5) at 3, 6 and 12 months; female sexual (FSFI) at 3 and 6 months; defecation (Wexner) at 3 and 6 months. No difference in chronic pain.

    Statistical interpretation: Patient-reported outcomes in an unblinded design; effect sizes subject to physician review.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

11 centres in 8 provinces of China; July 2016–December 2020. Age 18–80; middle (>5–10 cm) or low (≤5 cm) rectal adenocarcinoma; cT1–T3 N0–N1 or ycT1–T3 Nx; no distant metastasis. 1240 randomized (620/620); mITT 1171 (586/585) — 69 post-randomization exclusions (no surgery, peritoneal metastases, invasion of adjacent organs, etc.; including 10–12 cm tumours). Neoadjuvant RT/CRT 43.3% vs 43.9%.

Intervention

Robotic TME/PME

Comparator

Conventional laparoscopic TME/PME

Results

Primary: 3-year locoregional recurrence 1.6% vs 4.0%; HR 0.45; P=.03 → robotic superiority shown. Secondary: CRM 4.0%/7.2%; Clavien ≥II 16.2%/23.1%; 3-year DFS 87.2%/83.4% (HR 0.74); 3-year OS 94.7%/93.0% (not significant); early urinary/sexual/defecation function favoured robotic. Practice impact: randomized oncologic evidence favouring robotic rectal surgery in experienced high-volume centres; as a single trial, generalizability subject to physician review.

Sample size

1240

Registry

NCT02817126

PMID

40455621

DOI

10.1001/jama.2025.8123

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract and PMC full text of the JAMA 2025 primary paper; PubMed abstract of the Lancet Gastroenterol Hepatol 2022 short-term paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Shenkang Hospital Development Center, Shanghai Municipal Health Commission (SHDC12016104) and Zhongshan Hospital, Fudan University (2016ZSLC18). According to the paper, the funder had no role in the study (publication statement).

Conflicts of interest

JAMA 2025: “None reported.” Lancet Gastroenterol Hepatol 2022: “We declare no competing interests.”

Limits

Fewer events than planned (35 vs ~81); 69 exclusions in mITT. Single country (China), few highly experienced surgeons (one ~30% of cases) → limited generalizability. Perioperative treatment not standardized. Unblinded design (pathology assessment masked). PubMed ‘Equivalence Trial’ tag inconsistent with text (superiority). REAL ≠ ROLARR: different primary endpoint, population and expertise; results are not pooled.
Physician-approved (conditional) · 14 Nov 2026Colorectal

Dutch TME

Does short radiation therapy before rectal cancer surgery reduce cancer coming back?

Why conditional approval?

This is an older trial that studied only short-course (5 × 5 Gy) radiotherapy; surgical-margin and side-effect data were not verified.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Short-course 5×5 Gy radiotherapy; circumferential margin (CRM) and functional/toxicity data were not verified; historical trial.

Key finding

Short radiation before surgery significantly reduced cancer returning in the same area, and the difference persisted at ten years; overall survival did not change.

Why was this study done?

In patients with rectal cancer to be operated on with the standard technique that removes the tumour intact with its surrounding fatty tissue (total mesorectal excision, TME), the study asked whether adding 5 days of short-course radiation therapy (radiotherapy) before surgery reduces cancer coming back in the same area.

What did it find?

At two years, cancer returned in the same area in about 2 in 100 patients who had radiation and about 8 in 100 with surgery alone; the benefit of radiation was shown. Overall survival did not change. Both groups had the same surgical technique; the study measured the effect of radiation, not of the operation.

What did longer follow-up show? At ten years, cancer returned in the same area in 5% of patients who had short-course radiotherapy before surgery and 11% of those who had surgery alone. Overall survival, however, was similar in both groups.

What does it mean for me?

Radiation before surgery can lower the risk of cancer returning in the same area for some patients, but it was not shown to lengthen life and it can affect bowel and sexual function. Whether it is needed depends on tumour location, stage and current imaging.

Which patients does it apply to?

Patients with rectal cancer that can be removed by surgery and has not spread to distant organs. The results are not generalised to other patient groups or to different radiation schedules.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-DUTCH-TME

Full scientific title

Dutch TME — short-course preoperative radiotherapy added to standardized TME

Clinical question

In resectable rectal cancer treated with standardized TME, does short-course preoperative radiotherapy (5×5 Gy) reduce local recurrence?

Themes

Timing and strategy · Evidence that changed practice

Study design

Multicentre randomized controlled trial

Primary endpoint

Local recurrence (randomized comparison: RT + TME vs TME alone)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Primary2-year local recurrence — primary (NEJM 2001)

    Result: Among 1748 patients with macroscopically complete resection: RT + TME 2.4% vs TME alone 8.2%; P<.001.

    Statistical interpretation: This is the effect of radiotherapy; because both arms received TME, it does not measure the effect of the TME technique.

    Endpoint met?: Met

  • 10-year local recurrence (12-year follow-up publication)

    Result: Cumulative incidence 5% (RT + TME) vs 11% (TME alone); P<.0001.

    Statistical interpretation: The local-recurrence benefit of RT persisted long term; again an RT effect.

    Endpoint met?: Not applicable

  • SecondaryOverall survival (secondary)

    Result: 2 years: 82.0% vs 81.8%; P=.84. 10 years: no difference between groups (abstract gives no figures). In TNM stage III with negative CRM, 10-year survival 50% vs 40%; P=.032 (subgroup).

    Statistical interpretation: No overall survival benefit; increased cancer-specific survival in CRM-negative patients was offset by deaths from other causes. Subgroup finding is hypothesis-generating.

    Endpoint met?: Not applicable

  • Pathologic quality / functional outcomes / toxicity

    Result: No numeric data in the abstracts; CRM and TME quality data, RT-related bowel/sexual function and late toxicity not verified from full text in this review.

    Statistical interpretation: RT decisions are not based solely on local-recurrence data without weighing functional costs.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Multicentre RCT; randomization 12 January 1996–31 December 1999 (1:1, stratified by centre and planned operation). Resectable rectal cancer without distant disease. 1861 randomized (924 RT + TME / 937 TME alone); 1805 eligible; local recurrence analysis 1748 with macroscopically complete resection.

Intervention

Short-course preoperative radiotherapy (5 days × 5 Gy) + standardized TME

Comparator

Standardized TME alone

Results

RT effect: 2-year local recurrence 2.4% vs 8.2% (P<.001); 10-year 5% vs 11% (P<.0001). No overall survival difference (2-year 82.0% vs 81.8%). Stage III CRM-negative subgroup 10-year survival 50% vs 40% (subgroup). Surgical technique effect: not randomized — no inference from this trial. Practice impact: with TME as the standard surgical background, the trial showed a local-control benefit of short-course preoperative RT; RT indication is now based on contemporary MRI risk stratification.

Long-term update

van Gijn, Lancet Oncol 2011;12(6):575-82 (PMID 21596621): 12-year follow-up publication (median follow-up: could not be verified from the primary source); 10-year local recurrence 5% vs 11% (P<.0001); no overall survival difference; overall recurrence lower with RT; the apparent increase of RT effect with distance from the anal verge disappeared when CRM-positive patients were excluded. Stage III CRM-negative 10-year survival 50% vs 40% (P=.032).

Sample size

1861

PMID

11547717

DOI

10.1056/NEJMoa010580

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the NEJM 2001 primary paper and the 12-year follow-up paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

12-year follow-up paper: Dutch Cancer Society, Dutch National Health Council, Swedish Cancer Society. Funding and sponsor role of the NEJM 2001 paper could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

The randomized variable is RT; no inference about TME itself. 1990s era: before contemporary MRI-based risk stratification, CRT and TNT. No OS benefit; subgroup findings are hypothesis-generating. Functional cost, toxicity, CRM data, funding/COI and trial registration: could not be verified from the primary source. Not pooled with MRC CR07 etc.
Physician-approved (conditional) · 14 Nov 2026Colorectal

GRECCAR 2

For a small rectal tumour that responds well to treatment, can the rectum be preserved?

Why conditional approval?

Removing only the tumour area (local excision) was not shown to be better; about 35% of patients in that group later needed completion major surgery (TME).

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Superiority of local excision was not shown; about 35% of patients in the local-excision arm needed completion TME.

Key finding

Removing only the tumour area did not give a better overall result, and about a third of patients later needed major surgery; cancer return at five years was similar.

Why was this study done?

For small low rectal tumours that respond well to radiation and drug treatment before surgery (chemoradiotherapy), the study asked whether removing only the tumour area (local excision) gives a better overall result than complete removal of the rectum with its surrounding fatty tissue (total mesorectal excision, TME).

What did it find?

At two years, a combined measure of death, cancer return, complications and side effects was unfavourable in about 56 in 100 patients after local excision and about 48 in 100 after complete removal; local excision was not shown to be better. About 35 in 100 local-excision patients later needed complete removal of the rectum based on the tissue result. At five years, cancer returned in the same area in about 7 in 100 in both groups.

What did longer follow-up show? At five years, cancer came back in the same area in 7% in both groups; disease-free survival was 70% and 72%, overall survival 84% and 82%. The differences were not significant; but because the trial was small, the two approaches cannot be said to be identical.

What does it mean for me?

An organ-preserving approach may be an option for some very carefully selected patients, but major surgery may be needed later. Because the trial was small, the two approaches cannot be called equal. The decision is made with a team of different specialists.

Which patients does it apply to?

Adults at 15 French centres with a low rectal tumour up to 4 cm extending into or just beyond the muscle layer (stage T2–T3), with no more than 2 cm of tumour left after pre-operative treatment.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-GRECCAR-2

Full scientific title

GRECCAR 2 — local excision versus TME in low rectal cancer with good response to chemoradiotherapy

Clinical question

In small T2–T3 low rectal cancer with good clinical response to chemoradiotherapy, is local excision superior to TME for a 2-year composite endpoint (death, recurrence, morbidity, side effects)?

Themes

Organ preservation · De-escalation

Study design

Phase III randomized controlled trial

Primary endpoint

2-year composite: death, recurrence, morbidity and side effects (superiority of local excision)

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • Primary2-year composite — primary (Lancet 2017)

    Result: Adverse composite event: local excision 56% (41/73) vs TME 48% (33/69); OR 1.33 (95% CI 0.62–2.86); p=.43. In the local-excision arm, 26 patients underwent completion TME (for ypT2–3).

    Statistical interpretation: Superiority of local excision not shown. This does not prove local excision is worse or equivalent. The completion-TME rate likely increased morbidity.

    Endpoint met?: Not met

  • Secondary5-year local recurrence (secondary)

    Result: 7% (95% CI 3–16) vs 7% (3–16); adjusted HR 0.71 (0.19–2.58); p=.60.

    Statistical interpretation: Small sample; wide CI — no significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • Secondary5-year DFS (secondary)

    Result: 70% (58–79) vs 72% (60–82); HR 0.87 (0.44–1.72); p=.68.

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • Secondary5-year OS and metastasis (secondary)

    Result: OS 84% (73–91) vs 82% (71–90); HR 0.92 (0.38–2.22); p=.85. Metastatic disease 18% vs 19%. Cancer-specific mortality 7% vs 10%.

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • Morbidity, side effects and functional outcomes

    Result: Abstracts give no numbers for composite components (‘no difference in components’). Bowel, urinary and sexual function data not verified from full text in this review.

    Statistical interpretation: Functional benefit of organ preservation is not presented as proven on this card.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

15 French tertiary centres; open-label phase 3 RCT; 1 March 2007–24 September 2012. Age ≥18, T2–T3 low rectal cancer, maximum 4 cm, good clinical response after neoadjuvant CRT (residual ≤2 cm). 148 randomized; 145 analysed (74 local excision / 71 TME); mITT primary analysis 73/69.

Intervention

Local excision; completion TME if ypT2–3

Comparator

TME

Results

Primary: adverse composite 56% vs 48%; OR 1.33; p=.43 → superiority of local excision not shown. 26 completion TMEs in the local-excision arm. 5 years: local recurrence 7%/7%, DFS 70%/72%, OS 84%/82% — no significant difference (≠ equivalence). Practice impact: authors state local excision can be proposed for selected small, good-responding T2–T3 low rectal tumours, and better selection could reduce the need for completion TME.

Long-term update

Rullier, Lancet Gastroenterol Hepatol 2020;5(5):465-474 (PMID 32043980): median follow-up 60 months; 23 deaths, 5 lost to follow-up. Local recurrence 7% vs 7%; metastasis 18% vs 19%; OS 84% vs 82%; DFS 70% vs 72%; cancer-specific mortality 7% vs 10% — none significantly different. 5-year results are secondary endpoints; the trial was not powered for them.

Sample size

148

Registry

NCT00427375

PMID

28601342

DOI

10.1016/S0140-6736(17)31056-5

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the Lancet 2017 primary paper and the Lancet Gastroenterol Hepatol 2020 5-year paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Lancet 2017: National Cancer Institute of France (INCa), Sanofi, Roche Pharma. 2020: INCa. The sponsors' role could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Small sample (148); 5-year oncologic outcomes are secondary and unpowered — no significant difference ≠ equivalence. High completion-TME rate. Composite components carry different weights. No automatic generalization to contemporary watch-and-wait / TNT strategies. COI, functional data, CRT regimen: could not be verified from the primary source.
Physician-approved (conditional) · 14 Nov 2026Colorectal

COST

Is keyhole surgery for colon cancer as safe as open surgery for cancer coming back?

Why conditional approval?

The trial’s “not worse” margin is not given in the publication abstracts; the conversion-to-open rate and pathology data were not verified.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: The non-inferiority margin is not given in the abstracts; conversion rate and pathology data were not verified.

Key finding

Cancer return and survival were similar in both groups; keyhole surgery meant a shorter hospital stay and less need for pain medication.

Why was this study done?

For colon cancer, the study asked whether keyhole (laparoscopic) surgery is unacceptably worse than open surgery for cancer coming back.

What did it find?

At three years, cancer returned in about 16 in 100 patients after keyhole surgery and about 18 in 100 after open surgery. The authors concluded keyhole surgery was not worse than open surgery; however, the prespecified acceptance margin has not been verified on this card. Keyhole surgery shortened hospital stay by about one day and reduced pain-medication needs. At five years, about 76 vs 75 in 100 patients were alive.

What did longer follow-up show? At five years, disease-free survival (about 69%) and overall survival (about 75–76%) were similar in both groups, and recurrence rates were close.

What does it mean for me?

For colon cancer, keyhole surgery performed by experienced teams is an acceptable alternative to open surgery in suitable patients. These results do not directly apply to rectal cancer.

Which patients does it apply to?

Patients operated on for colon cancer at 48 institutions in the USA and Canada. Detailed eligibility criteria have not been verified on this card.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-COST

Full scientific title

COST — laparoscopically assisted versus open colectomy for colon cancer

Clinical question

In colon cancer, is laparoscopically assisted colectomy non-inferior to open colectomy for time to tumour recurrence?

Themes

Minimally invasive · Technology comparison · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Time to tumour recurrence (non-inferiority; margin: could not be verified from the primary source)

Primary endpoint met?

Unknown / unverified

Study outcomes (by outcome)

  • Primary3-year recurrence — primary (NEJM 2004)

    Result: Laparoscopic 16% vs open 18%; HR 0.86 (95% CI 0.63–1.17); two-sided P=.32.

    Statistical interpretation: Authors report a non-inferiority design and conclusion; no numeric NI margin in the abstracts → whether the primary endpoint was met cannot be determined (to be updated once the margin is verified). Not evidence of superiority or equivalence.

    Endpoint met?: Unknown / unverified

  • SecondaryOverall survival (secondary)

    Result: 3 years: 86% vs 85%; HR 0.91 (0.68–1.21); P=.51. 5 years (Ann Surg 2007): open 74.6% vs laparoscopic 76.4%; P=.93.

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • SecondaryDFS and recurrence — 5 years (secondary)

    Result: 5-year DFS open 68.4% vs laparoscopic 69.2%; P=.94. Overall recurrence open 21.8% vs laparoscopic 19.4%; P=.25. Wound recurrence (first site) 0.5% vs 0.9%; at 3 years wound recurrence <1% in both arms.

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • SecondaryShort-term recovery and safety (secondary)

    Result: Median stay 5 vs 6 days (P<.001); parenteral narcotics 3 vs 4 days (P<.001); oral analgesics 1 vs 2 days (P=.02). Intraoperative complications, 30-day mortality, readmission and reoperation ‘very similar’ (no numbers in abstract). Conversion rate and pathologic quality: could not be verified from the primary source.

    Statistical interpretation: Short-term recovery favoured laparoscopy.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

COST Study Group, 48 institutions, 66 credentialed surgeons. Colon adenocarcinoma; 872 randomized (per-arm numbers and exclusion criteria: could not be verified from the primary source). Rectal cancer outside scope (colon trial).

Intervention

Laparoscopically assisted colectomy

Comparator

Open colectomy

Results

Primary: 3-year recurrence 16% vs 18%; HR 0.86 (0.63–1.17) — authors report NI (margin unverified). 3-year OS 86%/85%; 5-year DFS open 68.4% / lap 69.2%; 5-year OS open 74.6% / lap 76.4%. Stay and analgesic needs favoured laparoscopy. Practice impact: one of the key RCTs supporting adoption of laparoscopic colectomy as an acceptable alternative to open surgery for colon cancer.

Long-term update

Fleshman, Ann Surg 2007;246(4):655-62 (PMID 17893502): median follow-up 7 years; 5-year data for 90% of patients. Overall recurrence open 21.8% vs lap 19.4% (P=.25); 5-year DFS 68.4% vs 69.2% (P=.94); 5-year OS 74.6% vs 76.4% (P=.93). Authors: laparoscopic colectomy is not inferior to open surgery on long-term oncologic endpoints.

Sample size

872

PMID

15141043

DOI

10.1056/NEJMoa032651

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the NEJM 2004 primary paper and the Ann Surg 2007 5-year paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Numeric NI margin, per-arm numbers, conversion rate and pathology data not in abstracts: could not be verified from the primary source. 1990s technology and credentialed surgeons. Colon trial: no generalization to rectum. COST ≠ COLOR ≠ CLASICC: different populations, designs and primary endpoints; results are not pooled.
Physician-approved (conditional) · 14 Nov 2026Colorectal

COLOR

Is keyhole surgery for colon cancer as good as open surgery for disease-free survival?

Why conditional approval?

The prespecified “not worse” limit was not fully met; local-recurrence data were not verified.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: The prespecified non-inferiority limit was not fully met; local-recurrence data were not verified.

Key finding

Three-year disease-free survival was close in both groups, but the statistical acceptance margin was not fully met; recovery was faster after keyhole surgery.

Why was this study done?

For right- or left-sided colon cancer, the study asked whether keyhole (laparoscopic) surgery is unacceptably worse than open surgery for three-year disease-free survival.

What did it find?

At three years, about 74 in 100 patients after keyhole surgery and about 76 in 100 after open surgery were alive without disease. The difference was small, but because the prespecified acceptance margin was not fully met, it could not be definitively shown that keyhole surgery is ‘not worse’. This does not mean keyhole surgery is worse. Keyhole surgery had less blood loss, faster recovery and a shorter hospital stay.

What does it mean for me?

For colon cancer, keyhole surgery is widely used by experienced teams in suitable patients; this study alone does not provide definitive proof of equivalence.

Which patients does it apply to?

Patients at 29 European hospitals with a single tumour in the right or left colon and a body mass index of 30 or less. Tumours in the middle (transverse) colon and rectal cancer were not included; COLOR II is a separate rectal study.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-COLOR

Full scientific title

COLOR — laparoscopic versus open surgery for colon cancer (≠ COLOR II)

Clinical question

In right- or left-sided colon cancer, is laparoscopic surgery non-inferior to open surgery for 3-year disease-free survival within a 7% margin?

Themes

Minimally invasive · Technology comparison · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

3-year disease-free survival (non-inferiority; margin 7%)

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • Primary3-year DFS — primary NI (Lancet Oncol 2009)

    Result: Laparoscopic 74.2% (95% CI 70.4–78.0) vs open 76.2% (72.6–79.8); log-rank p=.70. Difference 2.0% favouring open (95% CI −3.2 to 7.2). Upper CI bound 7.2% > prespecified 7% margin → NI not formally demonstrated.

    Statistical interpretation: NI not shown ≠ laparoscopy worse. The authors judged the difference ‘small and clinically acceptable’; this is the authors’ interpretation, not formal NI evidence.

    Endpoint met?: Not met

  • Secondary3-year OS (secondary)

    Result: Laparoscopic 81.8% (78.4–85.1) vs open 84.2% (81.1–87.3); log-rank p=.45; difference 2.4% (−2.1 to 7.0).

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • Short-term — pathology and recovery (Lancet Oncol 2005)

    Result: Median blood loss 100 vs 175 mL (p<.0001); operation ~30 min longer (p<.0001); conversion 91 patients (17%). Lymph-node count and resected bowel length similar. Earlier bowel function, less analgesia, shorter stay (favouring laparoscopy). 28-day morbidity and mortality similar. Positive margin rates ‘similar’ in the long-term publication.

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • SecondaryLocal and port-site recurrence (secondary)

    Result: Listed as secondary endpoints; no numbers in abstract: could not be verified from the primary source.

    Statistical interpretation: Not verified.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

29 European hospitals; 7 March 1997–6 March 2003; unblinded. Solitary right- or left-sided colon cancer, BMI ≤30; transverse colon and rectum excluded. 1248 randomized (627 lap / 621 open); 172 post-randomization exclusions (mainly distant metastases or benign disease) → 1076 analysed (534 / 542). Median follow-up 53 months.

Intervention

Laparoscopic colon resection

Comparator

Open colon resection

Results

Primary: 3-year DFS 74.2% vs 76.2%; difference 2.0% favouring open (95% CI −3.2…7.2); upper bound narrowly exceeded the 7% margin → NI not formally demonstrated. 3-year OS 81.8% vs 84.2% (not significant). Short-term recovery favoured laparoscopy; conversion 17%. Practice impact: together with COST and CLASICC, contributed to wider adoption of laparoscopic colon surgery; COLOR alone does not provide formal NI evidence.

Sample size

1248

Registry

NCT00387842

PMID

19071061

DOI

10.1016/S1470-2045(08)70310-3

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the Lancet Oncol 2009 primary paper and the Lancet Oncol 2005 short-term paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Upper CI bound exceeded the NI margin → NI not formally shown; the authors’ ‘clinically acceptable’ judgement is not formal evidence. 172 post-randomization exclusions. BMI ≤30, transverse colon excluded. Local recurrence data, funding/COI: could not be verified from the primary source. COLOR ≠ COLOR II (colon vs rectum); not pooled with COST/CLASICC.
Physician-approved (conditional) · 14 Nov 2026Colorectal

MRC CLASICC

Is keyhole surgery for colon and rectal cancer as safe as open surgery in the long term?

Why conditional approval?

The hypothesis and margin the trial tested are not stated in the abstracts; colon and rectal results should be interpreted separately.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: The formal hypothesis and margin are not given in the abstracts; colon and rectal results should be interpreted separately.

Key finding

At three, five and ten years there was no significant difference in survival or cancer return; the trial was done in the early years of keyhole surgery.

Why was this study done?

For colon or rectal cancer, the study examined the short- and long-term results of keyhole (laparoscopic) surgery compared with open surgery.

What did it find?

For colon cancer, early results were similar. For rectal cancer, keyhole operations from the front had a higher rate of cancer cells at the surgical margin around the tumour; however, in later years this did not lead to a significant rise in cancer returning. About 29 in 100 keyhole operations were switched to open surgery, and these patients had worse outcomes. At three, five and ten years, there was no significant difference in survival or cancer return.

What did longer follow-up show? With long-term follow-up, no significant difference was found between keyhole and open surgery in survival or cancer recurrence.

What does it mean for me?

The trial was done in the early years of keyhole surgery; techniques and experience have since improved. No significant difference does not prove the two methods are equal. Decisions for colon and rectal cancer are considered separately.

Which patients does it apply to?

Patients at 27 UK centres operated on for colon or rectal cancer. Detailed eligibility criteria have not been verified on this card.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-CLASICC

Full scientific title

MRC CLASICC — laparoscopically assisted versus conventional (open) surgery for colorectal cancer

Clinical question

In colon and rectal cancer, how does laparoscopically assisted resection compare with open resection for short-term pathologic/clinical endpoints and long-term oncologic outcomes?

Themes

Minimally invasive · Technology comparison · Evidence that changed practice

Study design

Multicentre randomized controlled trial

Primary endpoint

Short-term: CRM and longitudinal margin positivity, proportion of Dukes C2, in-hospital mortality; long-term: 3-year OS, DFS, local recurrence

Study outcomes (by outcome)

  • Short-term pathology and mortality (Lancet 2005)

    Result: Dukes C2: open 18 (7%) vs lap 34 (6%); difference −0.3% (95% CI −3.9 to 3.4); p=.89. In-hospital mortality: open 13 (5%) vs lap 21 (4%); difference −0.9% (−3.9 to 2.2); p=.57. CRM and longitudinal margin positivity similar — except after laparoscopic anterior resection (rectum), where CRM positivity was higher (numbers: could not be verified from the primary source).

    Statistical interpretation: Formal hypothesis (superiority/NI) and margin not in abstracts → whether the primary endpoint was met cannot be determined. Colon and rectal results are interpreted separately.

    Endpoint met?: Unknown / unverified

  • 3-year OS, DFS and local recurrence (JCO 2007)

    Result: OS difference 1.8% (95% CI −5.2 to 8.8; P=.55); DFS difference −1.4% (−9.5 to 6.7; P=.70); local recurrence difference −0.8% (−5.7 to 4.2; P=.76). Direction of differences (which arm minus which) not determinable from the abstract: could not be verified from the primary source. The higher CRM positivity after laparoscopic anterior resection did not translate into more local recurrence. No difference in quality of life.

    Statistical interpretation: No significant difference ≠ equivalence; CIs do not exclude clinically relevant differences.

    Endpoint met?: Not applicable

  • 5- and 10-year follow-up (BJS 2010, 2013)

    Result: 5 years: no difference in OS, DFS, local or distant recurrence (no numbers in abstract); port-site/wound recurrence 2.4% in the laparoscopic arm. Long-term (median follow-up 62.9 months): median OS open 78.3 vs lap 82.7 months (P=.780); median DFS 89.5 vs 77.0 months (P=.589).

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • Conversion to open (non-randomized comparison)

    Result: 143 patients (29%) in the laparoscopic arm were converted; converted patients had more complications. At 5 years, conversion was associated with worse OS (not DFS). In colon cancer long term: conversion OS HR 2.28 (1.47–3.53), DFS HR 2.20 (1.31–3.67).

    Statistical interpretation: Conversion is not a randomized variable; the association is not proof of causation (selection bias).

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

27 UK centres; July 1996–July 2002. Colon and rectal cancer; 2:1 randomization favouring laparoscopy — 794 randomized (526 lap / 268 open). Colon/rectum split: could not be verified from the primary source. Treatment received: 484 lap, 253 open; 6 had no surgery; 23 had missing surgical data.

Intervention

Laparoscopically assisted colorectal resection

Comparator

Conventional (open) resection

Results

Short-term: Dukes C2 7%/6%, in-hospital mortality 5%/4% — no difference; CRM positivity higher after laparoscopic anterior resection. 3 years: OS, DFS and local recurrence differences not significant; higher CRM positivity did not translate into local recurrence. 5 and 10 years: no significant differences. Conversion 29% and associated with worse outcomes (not randomized). Practice impact: supported oncologic safety of laparoscopy for colon cancer; for rectum, routine use was deemed premature in 2005, with supportive long-term reports — rectal evidence is assessed separately with COLOR II/ALaCaRT/Z6051.

Long-term update

Jayne JCO 2007 (3 years): OS, DFS and local recurrence differences not significant. Jayne BJS 2010 (5 years): no difference; port-site recurrence 2.4%; conversion associated with worse OS. Green BJS 2013 (median follow-up 62.9 months): median OS 78.3 vs 82.7 months (P=.780); median DFS 89.5 vs 77.0 months (P=.589). The authors state results support oncologic safety in colon and rectal cancer — not presented as proof of equivalence.

Sample size

794

Registry

ISRCTN74883561

PMID

15894098

DOI

10.1016/S0140-6736(05)66545-2

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the Lancet 2005 primary paper and the 3-, 5- and 10-year follow-up papers. Information that could not be verified is stated explicitly under “Limitations”.

Funding

The trial carries the Medical Research Council (MRC) name; funding details could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Colon and rectal patients in one trial; subgroup split and rectal CRM numbers: could not be verified from the primary source. Formal hypothesis/margin not verified. 2:1 randomization. Early-era laparoscopy (29% conversion). Direction of 3-year differences not determinable from abstract. Conversion analyses are non-randomized. CLASICC ≠ COST ≠ COLOR; rectal component not pooled with COLOR II/ALaCaRT/Z6051.
Physician-approved (conditional) · 14 Nov 2026Upper gastrointestinal

Dutch D1D2

Does wider lymph node removal in stomach cancer surgery extend life?

Why conditional approval?

This is an older trial; removal of the spleen and pancreas in the extended (D2) surgery group increased complications.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Historical trial; splenectomy/pancreatectomy in the D2 arm increased morbidity; there is an inconsistency in the length-of-stay data.

Key finding

The wider operation caused more complications and deaths in the short term; the overall survival difference was not significant, but deaths from stomach cancer were fewer at 15 years.

Why was this study done?

The study compared, for survival, removing only the lymph nodes right around the stomach (D1 lymph node dissection) with also removing lymph nodes from a wider area (D2 lymph node dissection) during stomach cancer surgery.

What did it find?

With the wider operation, complications occurred in about 43 vs 25 in 100 patients and operation-related death in about 10 vs 4 in 100; part of this was linked to also removing the spleen and tail of the pancreas. At five years, about 47 vs 45 in 100 patients were alive; there was no significant difference in overall survival. At fifteen years, the wider operation had less cancer returning in the same area and fewer deaths from stomach cancer.

What did longer follow-up show? At fifteen years, patients who had the extended (D2) operation had less cancer coming back in the same area and fewer deaths from gastric cancer; the overall survival difference was not statistically significant.

What does it mean for me?

A wider lymph node dissection may reduce death from stomach cancer in the long term, but carries higher surgical risk. In 2010 the study authors recommended the wider operation with the spleen preserved. The extent of surgery is decided with the surgical team based on tumour location and stage.

Which patients does it apply to?

Patients under 85 at 80 Dutch hospitals with stomach cancer that had not spread to distant organs. The trial was conducted in 1989–1993.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-DUTCH-D1D2

Full scientific title

Dutch D1D2 — D1 versus D2 lymphadenectomy for gastric cancer

Clinical question

In curative gastric cancer surgery, does extended (D2) lymph-node dissection improve survival over limited (D1) dissection, and at what cost in morbidity and mortality?

Themes

Surgical technique · Evidence that changed practice

Study design

Multicentre randomized controlled trial

Primary endpoint

Survival (D1 vs D2); short-term morbidity and mortality reported separately

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • Short-term morbidity and mortality (Lancet 1995)

    Result: Morbidity D2 43% vs D1 25% (p<.001); postoperative mortality D2 10% (32) vs D1 4% (15) (p=.004); reoperation 18% vs 8%. Median stay 25 vs 18 days in the 1995 paper, 16 vs 14 days in the 1999 paper — discrepancy: could not be verified from the primary source.

    Statistical interpretation: Randomized D1–D2 comparison; how much of the excess D2 morbidity/mortality is due to pancreatic/splenic resection cannot be separated from these figures.

    Endpoint met?: Not applicable

  • Overall survival — 5, 11 and 15 years

    Result: 5 years: D1 45% vs D2 47% (95% CI for difference −9.6 to +5.6). 11 years: 30% vs 35% (p=.53). 15 years: 21% vs 29% (p=.34).

    Statistical interpretation: No overall survival benefit shown; the 1999 and 2004 authors note high postoperative mortality offset any long-term benefit.

    Endpoint met?: Not met

  • Recurrence and gastric-cancer-related death — 15 years

    Result: Gastric-cancer-related death D1 48% vs D2 37% (significant). Local recurrence 22% vs 12%; regional recurrence 19% vs 13%. 5-year cumulative relapse after R0 resection 43% vs 37% (95% CI −2.4 to +14.4).

    Statistical interpretation: 15-year paper shows locoregional control and cancer-specific death favouring D2; overall survival difference not significant.

    Endpoint met?: Not applicable

  • Effect of splenectomy and pancreatectomy (not randomized)

    Result: In the 2004 final analysis, splenectomy, pancreatectomy, D2 and age >70 were significant risk factors for morbidity and mortality (risk ratio >1); no numbers in abstract. Which cases had pancreatic/splenic resection (protocol): could not be verified from the primary source.

    Statistical interpretation: Organ resection was not a randomized variable; the association is observational and cannot be causally separated from the effect of D2 dissection itself.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

80 Dutch hospitals; August 1989–July 1993. Histologically proven gastric adenocarcinoma without distant metastasis, age <85. 1078 randomized; 996 eligible; 711 curative (380 D1 / 331 D2); 285 palliative. Eleven supervising surgeons decided on curability and assisted with operations. NCI register DUT-KWF-CKVO-8905 (EU-90003).

Intervention

D2 lymphadenectomy (extended)

Comparator

D1 lymphadenectomy (perigastric)

Results

D2 had higher morbidity (43% vs 25%) and mortality (10% vs 4%). No significant overall survival difference (5-year 47% vs 45%; 15-year 29% vs 21%). At 15 years, D2 had lower local (12% vs 22%) and regional recurrence and gastric-cancer-related death (37% vs 48%). Part of the morbidity was associated with pancreatic/splenic resection (observational). Practice impact: routine D2 was not supported in 1999; in 2010 the authors recommended spleen-preserving D2.

Long-term update

Hartgrink JCO 2004 (11 years): OS 30% vs 35% (p=.53); only N2 patients may benefit from D2 (subgroup; N2 hard to predict). Songun Lancet Oncol 2010 (15 years): OS 21% vs 29% (p=.34); gastric-cancer-related death 48% vs 37%; local recurrence 22% vs 12%; regional recurrence 19% vs 13%. Authors recommend spleen-preserving D2 for curable gastric cancer.

Sample size

1078

Registry

DUT-KWF-CKVO-8905

PMID

10089184

DOI

10.1056/NEJM199903253401202

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the NEJM 1999 primary paper and the Lancet 1995, JCO 2004 and Lancet Oncol 2010 papers. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Dutch Health Insurance Funds Council and The Netherlands Cancer Foundation (2010 paper).

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Pancreatic/splenic resection not randomized; cannot be causally separated from the D2 effect. D2 performed with instruction/supervision across 80 hospitals (centre experience: could not be verified from the primary source). Overall survival not significant; 15-year cancer-specific difference is a secondary analysis. Hospital stay inconsistent between papers. N2 subgroup finding is hypothesis-generating. COI: could not be verified from the primary source.
Physician-approved (conditional) · 14 Nov 2026Upper gastrointestinal

TIME

Does keyhole oesophageal cancer surgery reduce lung infection?

Why conditional approval?

This is a small trial (115 patients); 30- and 90-day death rates and technical details were not verified.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Small sample (115); 30/90-day mortality and technical details were not verified.

Patients with a lung infection in the first two weeks after surgery (per 100 patients, approximately)
Time point: First 2 weeksOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With keyhole surgery, about 20 fewer lung infections per 100 patients.

The difference was statistically significant. There was no significant difference in three-year survival.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For oesophageal (food pipe) cancer surgery, the study asked whether surgery done through small cuts in the chest and abdomen with a camera (minimally invasive) reduces lung infection (pneumonia) after surgery compared with standard open surgery.

What did it find?

In the first two weeks, lung infection occurred in about 29 in 100 patients after open surgery and about 9 in 100 after minimally invasive surgery; minimally invasive surgery was shown to be better. There was no significant difference in three-year survival or disease-free survival.

What did longer follow-up show? At three years, no significant difference in survival or disease-free survival was found between keyhole and open surgery; the trial was not large enough to answer this question definitively.

What does it mean for me?

At experienced centres, minimally invasive oesophageal surgery can reduce lung problems after surgery. The trial was small and did not definitively show that cancer outcomes are the same.

Which patients does it apply to?

Patients aged 18–75 at five centres in three countries with removable cancer of the oesophagus within the chest or where the oesophagus joins the stomach.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-TIME

Full scientific title

TIME — minimally invasive versus open oesophagectomy for oesophageal cancer

Clinical question

In resectable oesophageal or gastro-oesophageal junction cancer, does minimally invasive transthoracic oesophagectomy reduce postoperative pulmonary infection compared with open transthoracic oesophagectomy?

Themes

Minimally invasive · Technology comparison · Perioperative safety

Study design

Multicentre, open-label randomized controlled trial

Primary endpoint

Pulmonary infection within the first 2 weeks and during the hospital stay

Primary endpoint met?

Met

Study outcomes (by outcome)

  • PrimaryPulmonary infection — primary (Lancet 2012)

    Result: First 2 weeks: open 16/56 (29%) vs MIE 5/59 (9%); RR 0.30 (95% CI 0.12–0.76); P=.005. During hospital stay: 34% vs 12%; RR 0.35 (0.16–0.78); P=.005.

    Statistical interpretation: Primary short-term endpoint met; small sample (115) and open-label design.

    Endpoint met?: Met

  • In-hospital mortality and other complications

    Result: In-hospital deaths: open 1 (anastomotic leak) vs MIE 2 (aspiration; mediastinitis after anastomotic leak). 30/90-day mortality, major complications, stay, R0 and node count not in abstracts: could not be verified from the primary source.

    Statistical interpretation: Not powered for mortality; very few events.

    Endpoint met?: Not applicable

  • 3-year OS and DFS (Ann Surg 2017)

    Result: OS open 40.4% vs MIE 50.5%; P=.207; HR 0.883 (0.540–1.441). DFS 35.9% vs 40.2%; HR 0.691 (0.389–1.239).

    Statistical interpretation: No significant difference ≠ oncologic equivalence; the trial was not powered for survival.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Five centres in three countries; 1 June 2009–31 March 2011; open-label. Age 18–75, resectable intrathoracic oesophageal or GOJ cancer (including Siewert I). 115 randomized (56 open / 59 MIE); ITT. Histology and neoadjuvant therapy distribution: could not be verified from the primary source.

Intervention

Minimally invasive transthoracic oesophagectomy

Comparator

Open transthoracic oesophagectomy

Results

Primary: 2-week pulmonary infection 29% vs 9% (RR 0.30; P=.005) → favouring MIE. In-hospital deaths 1 vs 2. 3-year OS 40.4% vs 50.5% and DFS 35.9% vs 40.2% — no significant difference. Practice impact: randomized evidence of a short-term pulmonary benefit of MIE; no claim of oncologic equivalence.

Long-term update

Straatman, Ann Surg 2017;266(2):232-236 (PMID 28187044): 3-year OS open 40.4% vs MIE 50.5% (P=.207; HR 0.883); DFS 35.9% vs 40.2% (HR 0.691). Authors: no difference in DFS or OS. Small trial not powered for survival.

Sample size

115

Registry

NTR TC 2452

PMID

22552194

DOI

10.1016/S0140-6736(12)60516-9

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the Lancet 2012 primary paper and the Ann Surg 2017 3-year paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Digestive Surgery Foundation, Unit of Digestive Surgery, VU University Medical Centre (Lancet 2012).

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Small sample (115); open-label; five centres. Technical details, histology and neoadjuvant therapy: could not be verified from the primary source. Not powered for long-term survival; no significant difference ≠ equivalence. TIME (total MIE) ≠ MIRO (hybrid); results are not pooled.
Physician-approved (conditional) · 14 Nov 2026Upper gastrointestinal

MIRO

Does doing the abdominal part of oesophageal cancer surgery by keyhole reduce serious complications?

Why conditional approval?

The hybrid technique was performed in experienced centres; death and quality-of-life results were not verified.

Verified source: Verified at the level of the primary publication abstract and the follow-up publication in PMC.

Conditions kept on this card: The hybrid technique was performed in experienced centres; mortality and quality-of-life results were not verified.

Patients with a serious complication in the 30 days after surgery (per 100 patients, approximately)
Time point: 30 daysOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: With hybrid surgery, about 28 fewer serious complications per 100 patients.

The difference was statistically significant. The difference in five-year survival was not significant.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For oesophageal (food pipe) cancer surgery, the study asked whether a ‘hybrid’ operation, with the abdominal part done by keyhole (laparoscopic) surgery and the chest part done open, reduces serious complications in the 30 days after surgery compared with fully open surgery.

What did it find?

Serious complications occurred in about 36 in 100 patients after hybrid surgery and about 64 in 100 after open surgery; hybrid surgery was shown to be better. Serious lung complications were also fewer (about 18 vs 30 in 100). At five years, about 59 in 100 were alive in the hybrid group and about 47 in 100 in the open group, but this difference was not statistically significant.

What did longer follow-up show? At five years, survival was 59% after hybrid surgery and 47% after open surgery; this difference was not statistically significant, and the trial was not designed for this question.

What does it mean for me?

At experienced centres, hybrid surgery can reduce serious problems after surgery. A survival benefit was not shown. ‘Totally minimally invasive’ surgery, where the chest part is also done by keyhole, is a different method and these results do not directly apply to it.

Which patients does it apply to?

Patients aged 18–75 at 13 French centres with removable cancer in the middle or lower third of the oesophagus.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-MIRO

Full scientific title

MIRO — hybrid minimally invasive versus open oesophagectomy for oesophageal cancer

Clinical question

In middle- and lower-third oesophageal cancer, does hybrid minimally invasive oesophagectomy (laparoscopic abdominal phase + open right thoracotomy) reduce 30-day major complications compared with open oesophagectomy?

Themes

Minimally invasive · Technology comparison · Perioperative safety

Study design

Phase III randomized controlled trial

Primary endpoint

Intraoperative or postoperative Clavien-Dindo grade ≥II major complication within 30 days

Primary endpoint met?

Met

Study outcomes (by outcome)

  • PrimaryMajor complication — primary (NEJM 2019)

    Result: Hybrid 37 (36%) vs open 67 (64%); OR 0.31 (95% CI 0.18–0.55); P<.001. Major pulmonary complications 18/102 (18%) vs 31/103 (30%).

    Statistical interpretation: Primary short-term endpoint met (favouring hybrid).

    Endpoint met?: Met

  • Mortality

    Result: 30- and 90-day mortality not in the PubMed abstracts or the 5-year full text: could not be verified from the primary source.

    Statistical interpretation: Not verified.

    Endpoint met?: Unknown / unverified

  • 3- and 5-year OS / DFS (post hoc follow-up)

    Result: 3 years (NEJM 2019): OS 67% (57–75) vs 55% (45–64); DFS 57% (47–66) vs 48% (38–57). 5 years (JAMA Surg 2021): OS 59% (48–68) vs 47% (37–57), HR 0.71 (0.48–1.06); DFS 52% (42–61) vs 44% (34–53), HR 0.81 (0.55–1.17).

    Statistical interpretation: 5-year analysis is post hoc; protocol not designed or powered for long-term outcomes. No significant difference ≠ equivalence; the trend favouring hybrid is not evidence of superiority.

    Endpoint met?: Not applicable

  • Quality of life

    Result: Separate publication (Mariette Ann Surg 2020, PMID 31404005); results not verified in this review.

    Statistical interpretation: Not verified.

    Endpoint met?: Unknown / unverified

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

13 centres in France; October 2009–April 2012; open-label. Age 18–75, resectable middle- or lower-third oesophageal cancer suitable for Ivor Lewis. 207 randomized (103 hybrid / 104 open); ITT. Adenocarcinoma 59%, squamous 41%. Neoadjuvant chemotherapy 42%, chemoradiotherapy 32%, none 27%. 85% men, median age 61.

Intervention

Hybrid minimally invasive oesophagectomy: laparoscopic gastric mobilization + open right thoracotomy (Ivor Lewis)

Comparator

Open transthoracic oesophagectomy (Ivor Lewis)

Results

Primary: 30-day major complications 36% vs 64%; OR 0.31; P<.001 → favouring hybrid. Major pulmonary complications 18% vs 30%. 5-year OS 59% vs 47% (HR 0.71; 0.48–1.06) and DFS 52% vs 44% — post hoc, no significant difference. Practice impact: randomized evidence that hybrid oesophagectomy reduces perioperative morbidity versus open surgery; no claim of oncologic superiority or equivalence.

Long-term update

Nuytens, JAMA Surg 2021;156(4):323-332 (PMID 33595631): 5-year OS 59% vs 47% (HR 0.71; 95% CI 0.48–1.06); 5-year DFS 52% vs 44% (HR 0.81; 0.55–1.17). In multivariable analysis, major complications (OS HR 2.21) and major pulmonary complications (OS HR 1.94) were associated with worse survival. Authors: post hoc analysis, original protocol not designed for long-term outcomes; lack of significance largely due to insufficient power.

Sample size

207

Registry

NCT00937456

PMID

30625052

DOI

10.1056/NEJMoa1805101

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the NEJM 2019 primary paper; PubMed abstract and PMC full text of the JAMA Surg 2021 5-year paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Original trial: PHRC / French National Cancer Institute. The 5-year follow-up received no additional funding; according to the paper, the funder had no role (publication statement).

Conflicts of interest

JAMA Surg 2021 full text: some authors reported national research grants and personal fees or in-kind support from drug/device companies. The statement of the NEJM 2019 paper could not be verified from the primary source.

Limits

Open-label; 207 patients. Long-term outcomes post hoc and unpowered. Mortality and quality-of-life results: could not be verified from the primary source. Only middle/lower-third tumours suitable for Ivor Lewis. MIRO (hybrid) ≠ TIME (total MIE); results are not pooled.
✓ Physician-approved · 9 Nov 2026Upper gastrointestinal

KLASS-01

For early-stage stomach cancer, is keyhole surgery an acceptable alternative to open surgery for five-year survival?

Patients alive at five years (per 100 patients, estimated)
Time point: 5 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Keyhole surgery: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For early-stage stomach cancer, the study asked whether removing the lower part of the stomach by keyhole (laparoscopic) surgery is unacceptably worse than open surgery for five-year survival.

What did it find?

At five years, about 94 in 100 patients after keyhole surgery and about 93 in 100 after open surgery were alive. Keyhole surgery stayed within the prespecified acceptance margin, meaning it was not unacceptably worse than open surgery for five-year survival. This does not mean the two operations are equal or that keyhole surgery is better. In a separate publication, overall complications in the first 30 days after surgery were fewer after keyhole surgery (about 13 vs 20 in 100). These rates come from the analysis of patients who had the operation they were assigned to (per-protocol analysis); similar results were reported in an analysis covering a wider group of patients.

What does it mean for me?

For early-stage stomach cancer, keyhole surgery by experienced teams can be considered an acceptable alternative to open surgery. The results do not directly apply to more advanced tumours or operations removing the whole stomach.

Which patients does it apply to?

Patients aged 20–80 at 13 South Korean institutions with stage I stomach cancer suitable for removal of the lower part of the stomach.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-KLASS-01

Full scientific title

KLASS-01 — laparoscopic versus open distal gastrectomy for clinical stage I gastric cancer

Clinical question

In clinical stage I gastric cancer, is laparoscopic distal gastrectomy non-inferior to open distal gastrectomy for 5-year overall survival within a −5% margin?

Themes

Minimally invasive · Technology comparison · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

5-year overall survival (non-inferiority; margin −5%)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Primary5-year OS — primary NI (JAMA Oncol 2019)

    Result: Laparoscopic 94.2% vs open 93.3%; log-rank P=.64; difference 0.9 percentage points (one-sided 97.5% CI lower bound −1.6). Lower bound −1.6 > −5 margin → NI met. Per-protocol consistent with ITT.

    Statistical interpretation: Non-inferiority only; not superiority or equivalence. The authors acknowledge the −5% margin was large relative to observed 5-year mortality.

    Endpoint met?: Met

  • Secondary5-year cancer-specific survival (secondary)

    Result: 97.1% vs 97.2%; log-rank P=.91; difference −0.03 points (one-sided 97.5% CI lower bound −1.8).

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • 30-day complications and mortality (Ann Surg 2016)

    Result: Per-protocol population (n=1256; laparoscopic 644, open 612): overall complications 13.0% vs 19.9% (P=.001); wound complications 3.1% vs 7.7% (P<.001); major intra-abdominal complications 7.6% vs 10.3% (P=.095); mortality 0.6% vs 0.3% (P=.687); conversion 0.9%. The Ann Surg 2016 abstract reports these rates for the per-protocol analysis and states that modified ITT (n=1384) results were similar; mITT rates are not in the abstract and were not verified against the Ann Surg full text.

    Statistical interpretation: Overall and wound complications favoured laparoscopy; no significant difference in major complications or mortality.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

South Korea; 13 institutions, 15 surgeons; January 2006–August 2010; phase 3 open-label NI RCT. Age 20–80; clinical stage I gastric adenocarcinoma (cT1N0M0, cT1N1M0 or cT2aN0M0); suitable for distal gastrectomy; ASA ≤3. 1416 randomized (705 lap / 711 open); mITT 1384; per-protocol 1256 (644/612). 85 crossovers (63 randomized to open → lap; 22 randomized to lap → open). Median follow-up 99.8 months.

Intervention

Laparoscopic distal gastrectomy (D1+β or D2)

Comparator

Open distal gastrectomy (D1+β or D2)

Results

Primary: 5-year OS 94.2% vs 93.3%; lower bound −1.6 > −5 → NI met. Cancer-specific survival 97.1% vs 97.2%. Overall complications 13.0% vs 19.9% (favouring lap) and 30-day mortality 0.6% vs 0.3% — Ann Surg 2016, per-protocol population (644/612). Practice impact: supported laparoscopic distal gastrectomy as an oncologically safe alternative for clinical stage I distal gastric cancer in experienced hands.

Sample size

1416

Registry

NCT00452751

PMID

30730546

DOI

10.1001/jamaoncol.2018.6727

Physician approval

Approved (9 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

PubMed abstracts PMID 26352529 + 30730546 and PMC6459124 full text (inclusion, lymphadenectomy, accreditation, funding, COI, limitations). Ann Surg 2016 COI → not verified from the primary source.

Funding

National Cancer Control Programs, Ministry of Health & Welfare, Republic of Korea (grant 0520310). The funder had no role (publication statement).

Conflicts of interest

JAMA Oncol 2019 statement: no conflicts of interest reported.

Limits

Clinical stage I and distal gastrectomy only; not transferable to advanced disease or total gastrectomy. 85 crossovers. −5% NI margin large relative to observed mortality. Highly experienced Korean centres; lower BMI than Western patients. Extracorporeal reconstruction. KLASS-01 (stage I) ≠ CLASS-01 (locally advanced); results are not pooled.
✓ Physician-approved · 14 Nov 2026Upper gastrointestinal

CLASS-01

Is keyhole surgery as safe as open surgery for locally advanced stomach cancer?

Patients alive without disease at three years (per 100 patients, estimated)
Time point: 3 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Keyhole surgery: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior. Because the margin was wide, the result should be interpreted cautiously.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For locally advanced stomach cancer that has grown into the deeper layers of the stomach wall, the study asked whether removing the lower part of the stomach with wide lymph node clearance (D2 lymph node dissection) by keyhole (laparoscopic) surgery is unacceptably worse than open surgery for three-year disease-free survival.

What did it find?

At three years, about 77 in 100 patients after keyhole surgery and about 78 in 100 after open surgery were alive without disease. Keyhole surgery stayed within the prespecified acceptance margin. Complications and operation-related death were similar. At five years, about 73 vs 76 in 100 were alive; the difference was not significant.

What did longer follow-up show? Five-year survival was 73% after keyhole and 76% after open surgery; the difference was not statistically significant, but a small difference has not been ruled out.

What does it mean for me?

At experienced centres, keyhole surgery may be an option for locally advanced stomach cancer. Because the acceptance margin was wide, results should be interpreted cautiously; a small difference was not ruled out.

Which patients does it apply to?

Patients aged 18–75 at 14 Chinese centres in good general condition, with cancer in the lower or middle stomach that had grown into the muscle layer or deeper but not into nearby organs or to distant organs, and who had not received drug treatment before surgery.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-CLASS-01

Full scientific title

CLASS-01 — laparoscopic versus open D2 distal gastrectomy for locally advanced gastric cancer

Clinical question

In locally advanced (cT2–4a N0–3 M0) gastric cancer, is laparoscopic distal gastrectomy with D2 lymphadenectomy non-inferior to open surgery for 3-year disease-free survival within a −10% margin?

Themes

Minimally invasive · Technology comparison

Study design

Non-inferiority randomized controlled trial

Primary endpoint

3-year disease-free survival (non-inferiority; margin −10%)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Primary3-year DFS — primary NI (JAMA 2019)

    Result: Laparoscopic 76.5% vs open 77.8%; difference −1.3% (one-sided 97.5% CI −6.5% to ∞). Lower bound −6.5 > −10 margin → NI met.

    Statistical interpretation: Non-inferiority only; not superiority or equivalence. The −10% margin is wide; a difference of up to −6.5% favouring open is not excluded.

    Endpoint met?: Met

  • Secondary3-year OS and recurrence (secondary)

    Result: OS 83.1% vs 85.2%; adjusted HR 1.19 (95% CI 0.87–1.64; P=.28). Recurrence 18.8% vs 16.5%; subHR 1.15 (0.86–1.54; P=.35).

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

  • 5-year OS (JAMA Surg 2022)

    Result: 72.6% vs 76.3%; log-rank P=.19; HR 1.17 (95% CI 0.93–1.48). Gastric-cancer death HR 1.14 (0.87–1.49); other causes HR 1.23 (0.74–2.05).

    Statistical interpretation: No significant difference ≠ equivalence; point estimate favours open and the CI does not exclude a clinically relevant difference favouring open.

    Endpoint met?: Not applicable

  • 30-day morbidity and mortality (JCO 2016)

    Result: Morbidity 15.2% vs 12.9%; difference 2.3% (95% CI −1.9 to 6.6; P=.285). Mortality 0.4% vs 0%; P=.249. Conversion to open surgery 6.4% (protocol definition: minilaparotomy >10 cm; source JAMA 2019, PMC6547120).

    Statistical interpretation: No significant difference ≠ equivalence.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

14 centres, 15 experienced surgeons in China; September 2012–December 2014; open-label NI RCT. Age 18–75, ECOG 0–1; lower- or middle-third gastric adenocarcinoma; cT2–4a N0–3 M0 (AJCC Ib–IIIc; T1 and T4b excluded); bulky nodes >3 cm excluded; no neoadjuvant therapy. 1056 randomized (528/528); 1039 operated (519/520). Pathologic stage III ≈42%, stage I ≈29% in both arms.

Intervention

Laparoscopic distal gastrectomy + D2 lymphadenectomy

Comparator

Open distal gastrectomy + D2 lymphadenectomy

Results

Primary: 3-year DFS 76.5% vs 77.8%; difference −1.3% (lower bound −6.5 > −10) → NI met. 3-year OS 83.1%/85.2%; 5-year OS 72.6%/76.3% (HR 1.17; not significant). Morbidity 15.2%/12.9%; mortality 0.4%/0%; conversion 6.4%. Practice impact: randomized evidence supporting laparoscopic D2 distal gastrectomy for locally advanced distal gastric cancer in experienced centres; interpreted cautiously given the wide NI margin and the 5-year point estimate.

Long-term update

Huang, JAMA Surg 2022;157(1):9-17 (PMID 34668963): 1039 patients; 5-year OS 72.6% vs 76.3% (log-rank P=.19; HR 1.17, 95% CI 0.93–1.48). Gastric-cancer death HR 1.14; other causes HR 1.23. Authors: similar 5-year OS. No equivalence claim.

Sample size

1056

Registry

NCT01609309

PMID

31135850

DOI

10.1001/jama.2019.5359

Physician approval

Approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract and PMC full text of the JAMA 2019 primary paper; PubMed records of the JCO 2016 and JAMA Surg 2022 papers. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Chinese government science programmes (e.g. Guangdong Provincial Science and Technology Key Project, National Key Technology R&D Program) and a Johnson & Johnson Medical Ltd programme. According to the paper, the sponsors had no role (publication statement).

Conflicts of interest

JAMA 2019 and JAMA Surg 2022: “None reported.”

Limits

Wide NI margin (−10%). China only, high-volume centres and experienced surgeons; limited transfer to Western and proximal-tumour practice. No neoadjuvant therapy (differs from current perioperative chemotherapy practice). No patient-reported outcomes. Minilaparotomy ≤10 cm. CLASS-01 (locally advanced) ≠ KLASS-01 (stage I); CLASS-01 ≠ CLASSIC (adjuvant chemotherapy).
Physician-approved (conditional) · 14 Nov 2026Upper gastrointestinal

JCOG9501

Does also removing lymph nodes around the aorta in stomach cancer extend life?

Why conditional approval?

The trial was done in high-volume centres in Japan; subgroup results and conflict-of-interest statements were not verified.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: High-volume centres in Japan; subgroup results and conflict-of-interest statements were not verified.

Patients alive at five years (per 100 patients, estimated)
Time point: 5 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

No survival benefit of the extra removal was shown; the operation took longer and caused more blood loss.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For stomach cancer surgery, the study asked whether also removing lymph nodes around the main artery (aorta), in addition to standard wide lymph node clearance (D2 lymph node dissection), lengthens survival.

What did it find?

At five years, about 69 in 100 patients were alive after standard surgery and about 70 in 100 after the extra clearance; no survival benefit of the extra clearance was shown. The wider operation took longer and caused more blood loss; complications were numerically higher (about 28 vs 21 in 100).

What does it mean for me?

For curative stomach cancer surgery, routinely removing lymph nodes around the aorta did not provide benefit. The study was done at very experienced Japanese centres.

Which patients does it apply to?

Patients at 24 Japanese hospitals suitable for curative surgery, with stomach cancer that had grown into the outer layers of the stomach wall or beyond (stage T2b–T4).

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-JCOG9501

Full scientific title

JCOG9501 — D2 versus D2 plus para-aortic nodal dissection for gastric cancer

Clinical question

In curable T2b–T4 gastric cancer, does adding para-aortic nodal dissection (PAND) to D2 lymphadenectomy improve overall survival?

Themes

Surgical technique · De-escalation

Study design

Multicentre randomized controlled trial

Primary endpoint

Overall survival

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • Primary5-year OS — primary (NEJM 2008)

    Result: D2 69.2% vs D2 + PAND 70.3%; HR 1.03 (95% CI 0.77–1.37); P=.85.

    Statistical interpretation: PAND did not improve survival. This is not proof of equivalence, but no benefit of the additional dissection was shown.

    Endpoint met?: Not met

  • SecondaryRelapse-free survival (secondary)

    Result: HR 1.08 (95% CI 0.83–1.42); P=.56.

    Statistical interpretation: No significant difference.

    Endpoint met?: Not applicable

  • Complications and mortality (JCO 2004; NEJM 2008)

    Result: Surgery-related complications D2 20.9% vs PAND 28.1% (P=.07). No significant difference in anastomotic leak, pancreatic fistula, abscess or pneumonia. Hospital/30-day mortality 0.8% in each group. PAND added a median 63 min of operating time and 230 mL of blood loss.

    Statistical interpretation: Very low mortality reflects experienced Japanese surgeons; not transferable to Western series.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

24 hospitals in Japan, experienced surgeons only; July 1995–April 2001. Curable T2b (subserosal), T3 or T4 gastric adenocarcinoma; surgically fit patients. 523 randomized (263 D2 / 260 D2 + PAND).

Intervention

D2 lymphadenectomy + para-aortic nodal dissection

Comparator

D2 lymphadenectomy alone

Results

Primary: 5-year OS 69.2% vs 70.3%; HR 1.03 (0.77–1.37) → no survival benefit from PAND. RFS HR 1.08. Complications 20.9% vs 28.1% (P=.07); mortality 0.8% in each group; PAND was longer with more blood loss. Practice impact: did not support routine prophylactic para-aortic dissection for curable gastric cancer; D2 remained standard.

Sample size

523

Registry

NCT00149279

PMID

18669424

DOI

10.1056/NEJMoa0707035

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the NEJM 2008 primary paper and the JCO 2004 paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Japanese high-volume centres and experienced surgeons only; Western patient profile (BMI, tumour location, stage distribution) differs → limited generalizability. Pre-perioperative-chemotherapy era. Curable T2b–T4 only; the question differs for patients with gross para-aortic involvement. The result is ‘no benefit’, not proof of equivalence. Funding/COI: could not be verified from the primary source.
Physician-approved (conditional) · 14 Nov 2026Upper gastrointestinal

JCOG0110

Is it safe to preserve the spleen in upper stomach cancer surgery?

Why conditional approval?

The result applies only to cancers of the upper stomach that have not spread to the greater curvature; complication figures were not verified.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Proximal gastric cancer without greater-curvature invasion; morbidity figures were not verified.

Patients alive at five years (per 100 patients, estimated)
Time point: 5 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Spleen preservation: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior. Removing the spleen led to more complications.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In patients with cancer in the upper stomach who need removal of the whole stomach, the study asked whether the spleen must also be removed to reach the lymph nodes around it. The question was whether preserving the spleen is unacceptably worse for survival.

What did it find?

At five years, about 75 in 100 patients with the spleen removed and about 76 in 100 with the spleen preserved were alive. Preserving the spleen stayed within the prespecified acceptance margin, meaning it was not worse. Removing the spleen led to more complications and blood loss.

What does it mean for me?

If the tumour does not extend to the stomach’s greater curvature, preserving the spleen is a reasonable option. This does not mean spleen preservation is superior. The decision is made with the surgical team based on tumour location.

Which patients does it apply to?

Patients at 36 Japanese institutions with cancer of the upper stomach not involving the greater curvature, grown into the muscle layer or deeper (stage T2–T4) and not spread to distant organs, in whom complete removal was confirmed possible during surgery and abdominal washings showed no cancer cells.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-JCOG0110

Full scientific title

JCOG0110 — splenectomy versus spleen preservation in total gastrectomy for proximal gastric cancer

Clinical question

In proximal gastric cancer not invading the greater curvature, is spleen preservation during total gastrectomy non-inferior to splenectomy for overall survival?

Themes

Surgical technique · De-escalation · Organ preservation

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Overall survival (non-inferiority of spleen preservation; HR margin 1.21; one-sided alpha 5%)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Primary5-year OS — primary NI (Ann Surg 2017)

    Result: Splenectomy 75.1% vs spleen preservation 76.4%; HR 0.88 (CI 0.67–1.16); P=.025 for NI. The abstract prints the CI level as ‘90.7%’ → possible typesetting error, to be checked against full text (could not be verified from the primary source).

    Statistical interpretation: Non-inferiority of spleen preservation confirmed; this does not mean spleen preservation is superior.

    Endpoint met?: Met

  • SecondaryMorbidity and blood loss (secondary)

    Result: Splenectomy associated with higher morbidity and greater blood loss; operating time similar. No numeric values, pancreatic fistula rate or RFS in abstract: could not be verified from the primary source.

    Statistical interpretation: Magnitude of the morbidity difference not verified.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

36 institutions in Japan; June 2002–March 2009. Proximal gastric adenocarcinoma not invading the greater curvature, T2–4 N0–2 M0; R0 resection feasibility and negative lavage cytology confirmed intraoperatively. 505 randomized (254 splenectomy / 251 spleen preservation).

Intervention

Spleen-preserving total gastrectomy

Comparator

Total gastrectomy with splenectomy (for splenic hilar node dissection)

Results

Primary: 5-year OS 75.1% (splenectomy) vs 76.4% (spleen preservation); HR 0.88; NI P=.025 → spleen preservation non-inferior. Splenectomy had more morbidity and blood loss (numbers not verified). Practice impact: the authors recommend avoiding splenectomy in proximal gastric cancer not invading the greater curvature.

Sample size

505

PMID

27280511

DOI

10.1097/SLA.0000000000001814

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the Ann Surg 2017 primary paper and the JCOG0110 supplementary analysis. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Tumours invading the greater curvature excluded. Japanese high-volume centres; limited transfer to Western patient profiles. CI level in abstract (‘90.7%’) unclear; morbidity numbers, registration, funding/COI: could not be verified from the primary source. The JCOG0110 name is not in the primary record (matched via supplementary analysis). Not pooled with the observational pancreas/spleen findings of Dutch D1D2.
Physician-approved (conditional) · 14 Nov 2026Breast cancer

NSABP B-06

Is keeping the breast as safe as removing the whole breast in breast cancer?

Why conditional approval?

This is an older trial carried out in 1976–1984; radiotherapy dose and drug treatments differed from today’s.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Historical trial (1976–1984); radiotherapy dose and systemic treatment differ from current practice.

Patients whose cancer came back in the same breast within twenty years — among those who had only the tumour removed (per 100 patients, approximately)
Time point: 20 yearsOrange dot: a patient who had this outcomeGrey dot: a patient who did not

Absolute difference: Adding radiotherapy meant cancer came back in the same breast in about 25 fewer people per 100.

There was no significant difference in survival between the three groups, including whole-breast removal.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

For breast cancer up to 4 cm, the study compared long-term results of removing the whole breast with removing only the tumour with a clear margin (lumpectomy), with or without radiation therapy (radiotherapy).

What did it find?

At twenty years, there was no significant difference between the three groups in disease-free or overall survival. After lumpectomy, cancer returned in the same breast in about 39 in 100 patients without radiation and about 14 in 100 with radiation.

What did longer follow-up show? At twenty years, there was no significant survival difference between breast-conserving surgery and removing the whole breast; radiotherapy clearly reduced cancer returning in the same breast.

What does it mean for me?

For suitable patients, breast-conserving surgery plus radiation is a reasonable alternative to removing the whole breast. Lumpectomy without radiation clearly increases the risk of cancer returning in the same breast. The right operation is decided with the surgical team based on tumour size, location and personal preferences.

Which patients does it apply to?

Patients with stage I–II breast cancer with tumours up to 4 cm; all had their armpit lymph nodes removed. The trial began in 1976; today’s treatments differ.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-NSABP-B06

Full scientific title

NSABP B-06 — total mastectomy versus lumpectomy with or without breast irradiation

Clinical question

In stage I–II (≤4 cm) breast cancer, does lumpectomy (with or without radiotherapy) compromise disease-free or overall survival compared with total mastectomy, and does breast irradiation reduce ipsilateral breast recurrence?

Themes

Organ preservation · Surgical technique · Evidence that changed practice

Study design

Three-arm randomized controlled trial

Primary endpoint

Disease-free, distant-disease-free and overall survival (3 arms); ipsilateral breast tumour recurrence (lumpectomy arms)

Primary endpoint met?

Not applicable

Study outcomes (by outcome)

  • Survival — 20 years (3 arms)

    Result: No significant differences in disease-free, distant-disease-free or overall survival among arms. Death HR: lumpectomy alone vs mastectomy 1.05 (95% CI 0.90–1.23; p=.51); lumpectomy+RT vs mastectomy 0.97 (0.83–1.14; p=.74).

    Statistical interpretation: Absence of a significant difference is not a formal equivalence test; the confidence intervals do not exclude small differences.

    Endpoint met?: Not applicable

  • Ipsilateral breast tumour recurrence — 20 years

    Result: Lumpectomy+RT 14.3% vs lumpectomy alone 39.2% (p<.001). In the 1985 analysis, breast-tumour-free rate 92.3% vs 72.1%.

    Statistical interpretation: RT was a randomized variable; the local-control benefit of RT after breast-conserving surgery is direct randomized evidence. Effect of RT on death in margin-negative patients HR 0.91 (0.77–1.06; p=.23) — not significant.

    Endpoint met?: Met

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Started 1976; stage I–II, tumours ≤4 cm. 1851 patients in the 2002 report (1843 in the 1985 analysis). All arms had axillary dissection; tumour-free margins were required for lumpectomy. Node-positive patients received chemotherapy (agents: could not be verified from the primary source). No trial registration number found.

Intervention

Lumpectomy (segmental mastectomy) + axillary dissection, with or without breast RT (RT randomized)

Comparator

Total mastectomy + axillary dissection

Results

At 20 years, no significant differences among the three arms in disease-free, distant-disease-free or overall survival. RT reduced ipsilateral breast recurrence from 39.2% to 14.3%. Practice impact: in suitable patients, lumpectomy + RT was adopted as a reasonable alternative to mastectomy; lumpectomy without RT carries high local recurrence.

Long-term update

Fisher NEJM 2002 (20 years): ipsilateral breast recurrence 14.3% vs 39.2% (p<.001); death HRs vs mastectomy 1.05 (lumpectomy alone) and 0.97 (lumpectomy+RT), not significant.

Sample size

1851

PMID

12393820

DOI

10.1056/NEJMoa022152

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the NEJM 1985, 1989 and 2002 papers. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

No significant survival difference is not formal equivalence. 1976 era: imaging, pathology, systemic therapy and RT technique differ from today. All patients had axillary dissection. RT dose, complications, funding/COI: could not be verified from the primary source.
Physician-approved (conditional) · 14 Nov 2026Breast cancer

Milan (Veronesi)

For small breast cancers, is removing only part of the breast enough?

Why conditional approval?

This is an older trial covering only tumours 2 cm or smaller; radiotherapy and drug treatments differed from today’s.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Historical trial; tumours ≤2 cm only; differs from current radiotherapy and systemic treatment.

Key finding

At twenty years, survival was similar in both groups; cancer returning in the same breast was somewhat more common after breast conservation.

Why was this study done?

For small breast cancers, the study compared long-term results of a radical operation removing the breast together with the chest muscle versus removing only the quarter of the breast containing the tumour followed by breast radiation therapy (radiotherapy).

What did it find?

At twenty years, death from any cause occurred in about 42 in 100 patients in both groups, and death from breast cancer in about 26 vs 24 in 100; these were similar. Cancer returning in the same breast was more common after breast conservation (about 9 vs 2 in 100).

What did longer follow-up show? At twenty years, there was no survival difference between breast-conserving surgery + radiotherapy and radical mastectomy; cancer returning in the same breast was more common in the conserving group.

What does it mean for me?

For small tumours, breast-conserving surgery plus radiation did not shorten survival compared with radical surgery, but the chance of cancer returning in the same breast was somewhat higher. The decision is made with the surgical team based on tumour features and preferences.

Which patients does it apply to?

Women in Milan in 1973–1980 with tumours up to 2 cm and no palpable armpit lymph nodes. Today’s treatments differ.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-MILANO-VERONESI

Full scientific title

Milan (Veronesi) — Halsted radical mastectomy versus quadrantectomy + axillary dissection + RT

Clinical question

In small (≤2 cm) breast cancer without palpable axillary nodes, does quadrantectomy + axillary dissection + breast radiotherapy compromise survival compared with Halsted radical mastectomy?

Themes

Organ preservation · Surgical technique · Evidence that changed practice

Study design

Randomized controlled trial

Primary endpoint

Overall survival and breast-cancer death; local recurrence reported separately

Primary endpoint met?

Not applicable

Study outcomes (by outcome)

  • Death — 20 years

    Result: All-cause death quadrantectomy arm 41.7% vs radical mastectomy 41.2% (p=1.0); breast-cancer death 26.1% vs 24.3% (p=0.8). No differences in contralateral cancer, distant metastases or second primary tumours.

    Statistical interpretation: No significant difference; this should not be presented as a non-inferiority or equivalence test.

    Endpoint met?: Not applicable

  • Local recurrence — 20 years

    Result: Same-breast recurrence 8.8% (30 women) in the quadrantectomy arm vs chest-wall recurrence 2.3% (8 women) after radical mastectomy; p<.001.

    Statistical interpretation: Local recurrence significantly higher in the breast-conserving arm; this did not translate into a survival difference.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Milan, 1973–1980; tumours <2 cm (≤2 cm in the 2002 paper), no palpable axillary nodes. 701 women: 349 Halsted radical mastectomy / 352 quadrantectomy. From 1976, node-positive patients received adjuvant chemotherapy (regimen: could not be verified from the primary source). No trial registration number.

Intervention

Quadrantectomy + axillary dissection + RT to the ipsilateral breast (all patients)

Comparator

Halsted radical mastectomy

Results

At 20 years, all-cause and breast-cancer death were similar (41.7% vs 41.2%; 26.1% vs 24.3%); local recurrence was higher in the quadrantectomy arm (8.8% vs 2.3%). Practice impact: supported replacing Halsted radical mastectomy with breast-conserving surgery + RT for small tumours.

Long-term update

Veronesi NEJM 2002 (20 years): death 41.7% vs 41.2% (p=1.0); breast-cancer death 26.1% vs 24.3% (p=0.8); local recurrence 8.8% vs 2.3% (p<.001).

Sample size

701

PMID

12393819

DOI

10.1056/NEJMoa020989

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the NEJM 1981 and 2002 papers. Information that could not be verified is stated explicitly under “Limitations”.

Funding

No funding information in the PubMed record; the full text could not be read, so the funding source could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Single centre; RT not randomized (surgery and RT effects cannot be separated). No difference is not formal equivalence. Comparator is Halsted mastectomy, no longer used. Only small (≤2 cm), clinically node-negative tumours. RT dose, complications, funding/COI: could not be verified from the primary source.
✓ Physician-approved · 14 Nov 2026Breast cancer

ACOSOG Z0011

If cancer is found in 1–2 sentinel nodes, do the other armpit nodes need to be removed?

Patients alive at ten years (per 100 patients, estimated)
Time point: 10 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Not removing the other nodes: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In women with cancer in only 1–2 of the first armpit lymph nodes that cancer reaches (sentinel nodes), the study asked whether the remaining armpit nodes also need to be removed (axillary dissection). The question was whether not removing them is unacceptably worse for survival.

What did it find?

At ten years, about 86 in 100 patients without further node removal and about 84 in 100 with node removal were alive. Not removing the other nodes stayed within the prespecified acceptance margin, meaning it was not worse. Cancer returning in the armpit was rare in both groups.

What did longer follow-up show? At ten years, survival in the group without full underarm node removal was not unacceptably worse than with full removal; cancer returning in the underarm was rare in both groups.

What does it mean for me?

For patients meeting these conditions, full clearance of the armpit is often not needed. This does not mean skipping it is superior. The decision is made with the surgical team.

Which patients does it apply to?

Women with tumours up to 5 cm (stage T1–T2), no palpable armpit nodes, cancer in only 1–2 sentinel nodes, and planned breast-conserving surgery with whole-breast radiation. It does not directly apply to mastectomy, 3 or more positive nodes, or drug treatment before surgery.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-Z0011

Full scientific title

ACOSOG Z0011 — ALND versus SLNB alone for 1–2 positive sentinel nodes

Clinical question

In women with cT1–T2 tumours, no palpable adenopathy, 1–2 positive sentinel nodes and planned lumpectomy + tangential whole-breast RT, does omitting axillary dissection (ALND) cause an unacceptable loss of overall survival?

Themes

De-escalation · Surgical technique · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Overall survival (non-inferiority; one-sided HR margin <1.3)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Overall survival — 10 years

    Result: SLNB alone 86.3% vs ALND 83.6%; HR 0.85 (one-sided 95% CI 0–1.16); non-inferiority p=.02. 5 years: 92.5% vs 91.8%.

    Statistical interpretation: Non-inferiority shown; this does not mean equivalence or superiority. The trial closed early with 891 randomized (target 1900; fewer deaths than expected); the 10-year OS analysis is extended follow-up not prespecified.

    Endpoint met?: Met

  • Disease-free survival and regional recurrence

    Result: 10-year DFS 80.2% vs 78.2%; HR 0.85 (95% CI 0.62–1.17; p=.32). Locoregional recurrence-free survival 83.0% vs 81.2%. One regional recurrence in the SLNB-alone arm during years 5–10. At 6.3 years (Ann Surg 2010) local recurrence p=.11, regional recurrence p=.45.

    Statistical interpretation: No significant differences in secondary endpoints; regional recurrence stayed low despite additional nodal macrometastases in 27.3% of the ALND arm.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

115 sites, May 1999–December 2004. cT1–T2, no palpable adenopathy, 1–2 positive sentinel nodes; lumpectomy + tangential whole-breast RT planned. 891 randomized (446 SLNB / 445 ALND); 35 withdrew consent → 856; ITT 436/420. Median age 55; ~84% of known cases ER+; adjuvant systemic therapy 97%/96%. Micrometastases 44.8% in SLNB arm vs 37.5% in ALND arm. Mastectomy, neoadjuvant therapy, ≥3 positive SLNs and palpable nodes were excluded.

Intervention

SLNB alone (no ALND)

Comparator

Completion ALND

Results

At 10 years, OS 86.3% vs 83.6% (HR 0.85; non-inferiority p=.02); DFS 80.2% vs 78.2% (p=.32); regional recurrence low in both arms. Practice impact: supported omitting routine ALND in patients meeting Z0011 criteria (cT1–T2, cN0, 1–2 positive SLNs, breast-conserving surgery + whole-breast RT, systemic therapy). Not generalizable to all SLN-positive patients.

Long-term update

Giuliano JAMA 2017: 10-year OS 86.3% vs 83.6%; DFS 80.2% vs 78.2%; locoregional recurrence-free survival 83.0% vs 81.2%; one regional recurrence in the SLNB-alone arm in years 5–10.

Sample size

891

Registry

NCT00003855

PMID

28898379

DOI

10.1001/jama.2017.11470

Physician approval

Approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts and PMC full texts of the JAMA 2011 and JAMA 2017 papers. Information that could not be verified is stated explicitly under “Limitations”.

Funding

National Cancer Institute U10 grants. According to the JAMA 2017 statement, the NCI had a role in design, conduct, analysis and writing; neither the Alliance nor the NCI had the right to veto submission of the manuscript.

Conflicts of interest

JAMA 2017 full text: the authors completed the ICMJE form; no other disclosures were reported. JAMA 2011 PubMed record: the authors indicated no potential conflicts of interest.

Limits

Closed early with 891 of a planned 1900; few events. The 10-year OS analysis was not prespecified. Off-protocol nodal RT (18.9%) and no RT (11%); high-tangent coverage uncertain. Only breast-conserving surgery + whole-breast RT; mastectomy, neoadjuvant therapy, palpable nodes and ≥3 positive SLNs excluded. Low-risk, mostly ER+ population with a high proportion of micrometastases. Lymphoedema rates: could not be verified from the primary source.
✓ Physician-approved · 14 Nov 2026Breast cancer

AMAROS (EORTC 10981-22023)

If the sentinel node has cancer, can armpit radiation replace armpit surgery?

Key finding

Cancer returning in the armpit was very rare with both methods; arm swelling was about half as common with radiation.

Why was this study done?

In patients with cancer in the first armpit lymph node that cancer reaches (sentinel node), the study compared surgical removal of the remaining armpit nodes (axillary dissection) with radiation therapy (radiotherapy) to the armpit.

What did it find?

At five years, cancer returned in the armpit in about 0.4 in 100 patients after surgery and about 1.2 in 100 after radiation; both were very rare. Because there were so few events, the statistical comparison of the two methods was underpowered. Arm swelling (lymphoedema) occurred in about 23 in 100 after surgery and about 11 in 100 after radiation. There was no significant difference in survival.

What did longer follow-up show? At ten years, cancer returning in the underarm was about 0.9% after surgery and about 1.8% after radiotherapy; arm swelling (lymphoedema) was less in those who had radiotherapy.

What does it mean for me?

Radiation to the armpit can be an option with less arm swelling than surgical node removal. This does not prove the two methods are equal. The right approach is decided with the surgeon and the radiation specialist (radiation oncologist).

Which patients does it apply to?

Patients at 34 centres in nine countries with tumours up to 5 cm (stage T1–T2), no palpable armpit nodes and cancer in the sentinel node.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-AMAROS

Full scientific title

AMAROS (EORTC 10981-22023) — axillary RT versus ALND for sentinel-node-positive patients

Clinical question

In T1–T2 breast cancer without palpable nodes and a positive sentinel node, does axillary radiotherapy (ART) provide axillary control comparable to axillary dissection (ALND) with less morbidity?

Themes

De-escalation · Surgical technique · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

5-year axillary recurrence (non-inferiority; HR margin 2)

Primary endpoint met?

Unknown / unverified

Study outcomes (by outcome)

  • Axillary recurrence — 5 years (SN-positive ITT)

    Result: ALND 0.43% (4/744) vs ART 1.19% (7/681). The planned non-inferiority test required 52 events; only 11 occurred. One-sided 95% CI for HR 0.00–5.27.

    Statistical interpretation: Because of very few events the non-inferiority test was underpowered; the upper CI exceeds the margin of 2. Formal non-inferiority therefore cannot be claimed; axillary recurrence was very low in absolute terms in both arms.

    Endpoint met?: Unknown / unverified

  • SecondaryLymphoedema (secondary)

    Result: Clinical signs of lymphoedema at 5 years ALND 23% vs axillary RT 11% (p<.0001; Lancet Oncol 2014). 1 year ALND 28% vs ART 15%; 3 years ALND 23% vs ART 14%. Arm circumference increase ≥10% at 5 years ALND 13% vs ART 5%.

    Statistical interpretation: Significantly less lymphoedema with ART; no difference in quality of life through 5 years.

    Endpoint met?: Not applicable

  • Survival — 5 and 10 years

    Result: 5-year OS ALND 93.3% vs ART 92.5%, HR 1.17 (0.85–1.62; p=.34); DFS ALND 86.9% vs ART 82.7%, HR 1.18 (0.93–1.51; p=.18). 10-year OS HR 1.17 (0.89–1.52); DFS HR 1.19 (0.97–1.46).

    Statistical interpretation: No significant differences; point estimates lean against ART but CIs are wide. No equivalence claim can be made.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

34 centres in 9 countries, February 2001–April 2010. T1–T2, no palpable nodes; randomized before sentinel-node biopsy. 4823 enrolled / 4806 eligible (2402 ALND / 2404 ART); SN-positive ITT 744 ALND / 681 ART. Breast-conserving surgery 82% in both arms, mastectomy 17%/18%. SN burden (ALND/ART): macrometastases 59%/62%, micrometastases 29%/29%, isolated tumour cells 12%/10%.

Intervention

Axillary RT: axillary levels I–III + medial supraclavicular region, 25 × 2 Gy

Comparator

Completion ALND

Results

5-year axillary recurrence ALND 0.43% vs ART 1.19% — both very low; the non-inferiority test was underpowered because of too few events. Lymphoedema at 5 years ALND 23% vs ART 11%. No significant differences in OS or DFS. At 10 years axillary recurrence ALND 0.93% vs ART 1.82%. In exploratory analysis, second primary cancers 12.1% ART vs 8.3% ALND. Practice impact: axillary RT was adopted as an alternative to ALND in SN-positive cT1–T2 patients, with less arm morbidity.

Long-term update

Bartels JCO 2023 (10 years): axillary recurrence ALND 0.93% vs ART 1.82%, HR 1.71 (95% CI 0.67–4.39); OS HR 1.17 (0.89–1.52); DFS HR 1.19 (0.97–1.46); 5-year lymphoedema ALND 24.5% vs ART 11.9% (updated analysis; primary-publication value ALND 23% vs ART 11%). Authors recommend ART as preferred for SN-positive cT1–T2 patients.

Sample size

4823

Registry

NCT00014612

PMID

25439688

DOI

10.1016/S1470-2045(14)70460-7

Physician approval

Approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract and PMC full text of the Lancet Oncol 2014 primary paper; PubMed abstract of the JCO 2023 10-year paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

EORTC Charitable Trust (Lancet Oncol 2014). According to the paper, the funder had no role in the study (publication statement).

Conflicts of interest

Lancet Oncol 2014 full text: “We declare no competing interests.”

Limits

Far fewer axillary recurrences than expected (11/52); non-inferiority could not be statistically proven. OS/DFS point estimates lean against ART with wide CIs. The second-primary difference is exploratory. Nearly 40% of SN burden was micrometastasis/ITC; transfer to patients with heavy macrometastatic burden is limited. Population without neoadjuvant therapy.
✓ Physician-approved · 14 Nov 2026Breast cancer

SOUND

For small breast cancers with a normal armpit ultrasound, can sentinel node biopsy be skipped?

Patients alive without spread to distant organs at five years (per 100 patients, estimated)
Time point: 5 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Skipping the biopsy: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In patients with small breast cancers and a normal pre-operative armpit ultrasound, the study asked whether it is safe to skip removing and examining the first armpit lymph node that cancer reaches (sentinel node biopsy). The question was whether skipping the biopsy is unacceptably worse.

What did it find?

At five years, about 98 in 100 patients were alive without spread to distant organs, both with and without biopsy. Skipping the biopsy stayed within the prespecified acceptance margin, meaning it was not worse. Cancer returning in the armpit occurred in fewer than 1 in 100 patients in both groups.

What does it mean for me?

In selected low-risk patients, armpit surgery can be skipped. However, the biopsy gives information that can affect drug treatment decisions in some patients; about 14 in 100 biopsied patients had cancer in the node. The decision is made with the surgical team.

Which patients does it apply to?

Patients with tumours up to 2 cm, a normal armpit ultrasound and planned breast-conserving surgery plus radiation. Most tumours were hormone-sensitive. It does not directly apply to patients having a mastectomy.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-SOUND

Full scientific title

SOUND — SLNB versus no axillary surgery in small breast cancer with negative axillary ultrasound

Clinical question

In patients with breast cancer ≤2 cm, negative preoperative axillary ultrasound and planned breast-conserving surgery + RT, does omitting sentinel-node biopsy (SLNB) cause an unacceptable loss of distant disease-free survival?

Themes

De-escalation · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

5-year distant disease-free survival (non-inferiority; absolute margin 2.5%, HR <1.74)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Distant disease-free survival — 5 years

    Result: SLNB 97.7% vs no axillary surgery 98.0%; HR 0.84 (two-sided 90% CI 0.45–1.54); non-inferiority p=.02.

    Statistical interpretation: Non-inferiority shown; this does not mean equivalence or superiority. Few events; a 10-year analysis is planned.

    Endpoint met?: Met

  • SecondarySecondary endpoints — 5 years

    Result: DFS 94.7% vs 93.9% (p=.30); OS 98.2% vs 98.4% (p=.72). Axillary recurrence as first event 0.4% in both arms (3 vs 5); distant metastasis 1.8% vs 2.0%. Positive nodes in the SLNB arm 13.7%.

    Statistical interpretation: No significant differences in secondary endpoints.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Italy, Switzerland, Spain, Chile; February 2012–June 2017. No age limit (median 60). Tumour ≤2 cm; negative preoperative axillary ultrasound — a single doubtful node required negative fine-needle aspiration (57 patients). Breast-conserving surgery + RT planned. 1463 randomized; ITT 1405 (SLNB 708 / no axillary surgery 697). ER+/HER2− 87.8%; median tumour 1.1 cm; grade 3 ~18%. Chemotherapy 20.1% vs 17.5%; endocrine therapy 90.7% vs 93.0%.

Intervention

No axillary surgery

Comparator

Sentinel-node biopsy (SLNB)

Results

5-year distant disease-free survival 97.7% vs 98.0% (non-inferiority p=.02); axillary recurrence 0.4% in both arms. Practice impact: supports omitting SLNB in small, mostly ER+/HER2− tumours with negative axillary ultrasound treated with breast-conserving surgery + RT; however, nodal staging from SLNB (13.7% positive here) may influence adjuvant treatment decisions.

Sample size

1463

Registry

NCT02167490

PMID

37733364

DOI

10.1001/jamaoncol.2023.3759

Physician approval

Approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract and PMC full text of the JAMA Oncol 2023 primary paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

The JAMA Oncol 2023 full text has no Funding/Support section; the funding source could not be verified from the primary source.

Conflicts of interest

JAMA Oncol 2023 full text: three authors reported personal fees from drug and device companies outside the submitted work; no other disclosures were reported.

Limits

Low-risk, mostly ER+/HER2− population; few events. Heterogeneous RT (about 11% partial-breast only); nodal RT unknown. Not applicable to mastectomy. Design predates Z0011. Registered after 492 patients. 5-year results; 10-year analysis awaited.
Physician-approved (conditional) · 14 Nov 2026Breast cancer

INSEMA

When the armpit looks normal in breast cancer, can sentinel node biopsy be skipped?

Why conditional approval?

The main result is based only on patients who followed the protocol (per-protocol analysis); age, tumour-subtype distribution and arm-problem rates were not verified.

Verified source: Verified at the level of the primary publication abstract and the trial registry.

Conditions kept on this card: Per-protocol analysis; age, ER/HER2 distribution and arm-morbidity rates were not verified.

Patients alive without spread or new cancer at five years (per 100 patients, estimated)
Time point: 5 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Skipping the biopsy: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior. Arm swelling and restricted movement were less common without biopsy.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In breast cancer patients whose armpit looked normal on examination and ultrasound, the study asked whether it is safe to skip removing and examining the first armpit lymph node that cancer reaches (sentinel node biopsy). The question was whether skipping the biopsy is unacceptably worse.

What did it find?

At five years, about 92 in 100 patients in both groups were alive without spread or new cancer. Skipping the biopsy stayed within the prespecified acceptance margin. Cancer returning in the armpit was slightly more common without biopsy (about 1 vs 0.3 in 100) but rare; arm swelling and restricted movement were less common.

What does it mean for me?

In selected patients, armpit surgery can be skipped. The information a biopsy provides may affect drug treatment decisions in some patients. The decision is made with the surgical team.

Which patients does it apply to?

Women aged 18 or older in Germany and Austria with cancer smaller than 5 cm in one breast (stage T1–T2; most smaller than 2 cm), a normal armpit on examination and ultrasound, and planned breast-conserving surgery with whole-breast radiation. It does not directly apply to mastectomy or partial-breast radiation only.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-INSEMA

Full scientific title

INSEMA — omission of SLNB in clinically node-negative early breast cancer

Clinical question

In clinically and sonographically node-negative T1–T2 (<5 cm) breast cancer planned for breast-conserving surgery + whole-breast RT, does omitting SLNB cause an unacceptable loss of invasive disease-free survival?

Themes

De-escalation · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

5-year invasive disease-free survival (non-inferiority; iDFS ≥85% and HR upper CI <1.271)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • Invasive disease-free survival — 5 years (per-protocol)

    Result: SLNB omitted 91.9% vs SLNB 91.7%; HR 0.91 (95% CI 0.73–1.14). Median follow-up 73.6 months.

    Statistical interpretation: Prespecified non-inferiority criteria were met; this does not mean equivalence or superiority.

    Endpoint met?: Met

  • Axillary recurrence, death and arm morbidity

    Result: Axillary recurrence 1.0% (no SLNB) vs 0.3% (SLNB); death 1.4% vs 2.4%. Less lymphoedema, better arm mobility and less pain with SLNB omission (rates: could not be verified from the primary source).

    Statistical interpretation: Axillary recurrence numerically higher with omission but low in absolute terms; arm morbidity lower with omission.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Germany and Austria; registry start September 2015. Women ≥18; unilateral invasive breast cancer, estimated tumour <5 cm (cT1/T2; cT1 90%, pT1 79%), irrespective of hormone-receptor and HER2 status. cN0/iN0: clinically and sonographically negative axilla; a suspicious node on ultrasound required negative core biopsy or FNA before randomization. Planned breast-conserving surgery (R0). 1:4 randomization (no SLNB : SLNB); 5502 patients; per-protocol 4858 (962 / 3896).

Intervention

No axillary surgery (SLNB omitted)

Comparator

Sentinel-node biopsy (SLNB)

Results

5-year iDFS 91.9% vs 91.7% (HR 0.91; non-inferiority met). Axillary recurrence 1.0% vs 0.3%; less arm morbidity with SLNB omission. Practice impact: supports omitting SLNB in a broader cN0 population than SOUND (T1–T2, any receptor status, mandatory whole-breast RT); however, most tumours were T1 and loss of nodal staging information may affect adjuvant treatment decisions.

Sample size

5502

Registry

NCT02466737

PMID

39665649

DOI

10.1056/NEJMoa2412063

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the NEJM 2024 primary paper; ClinicalTrials.gov NCT02466737. Information that could not be verified is stated explicitly under “Limitations”.

Funding

German Cancer Aid (Deutsche Krebshilfe) (primary paper abstract).

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Primary analysis per-protocol; small omission arm (962) due to 1:4 randomization. Mostly T1 (cT1 90%); limited evidence for T2. Mandatory whole-breast RT and breast-conserving surgery; not applicable to mastectomy or partial-breast RT alone. Age, ER/HER2 distribution, arm-morbidity rates and COI: could not be verified from the primary source (no PMC full text). Second-randomization results not yet available.
Physician-approved (conditional) · 14 Nov 2026Breast cancer

SENOMAC

If a sentinel node has a clear cancer deposit, do the other armpit nodes need to be removed?

Why conditional approval?

The main result, overall survival, has so far been reported only in a congress abstract (ASCO 2026); there is no peer-reviewed full paper yet.

Verified source: Verified at the level of the primary publication abstract and a congress abstract.

Conditions kept on this card: The primary overall-survival result is available only as an ASCO 2026 congress abstract, with no peer-reviewed full publication; recurrence-free survival (NEJM 2024) is a secondary outcome.

Patients alive at five years (per 100 patients, estimated; 2026 congress abstract — not a peer-reviewed paper)
Time point: 5 yearsTeal dot: a patient who reached this outcomeGrey dot: a patient who did not

Not removing the other nodes: It stayed within the prespecified acceptance margin (not unacceptably worse); this does not mean it is superior. This is a 2026 congress result; a peer-reviewed paper has not yet been published.

Dots and numbers are rounded, approximate values, not exact patient counts.

Why was this study done?

In patients with a cancer deposit larger than 2 mm in 1–2 of the first armpit lymph nodes that cancer reaches (sentinel nodes), the study asked whether the remaining armpit nodes also need to be removed (axillary dissection). The question was whether not removing them is unacceptably worse for survival.

What did it find?

At five years, about 93 in 100 patients with node removal and about 94 in 100 without were alive. Not removing the other nodes stayed within the prespecified acceptance margin; this does not mean “better”. In questionnaires completed by patients, arm movement and arm complaints (such as pain, swelling or weakness) were clearly better at 3 and 5 years in those whose other nodes were not removed; overall quality-of-life scores were similar in both groups. In both groups, about 88–90 in 100 patients received radiation to the lymph node areas. This result was presented at a congress in 2026; a peer-reviewed paper has not yet been published.

What does it mean for me?

For patients with these features who will receive radiation to the lymph node areas, extensive armpit surgery is often not needed. The decision is made with the surgeon and the radiation specialist (radiation oncologist).

Which patients does it apply to?

Adults with tumours not grown into surrounding tissues (stage T1–T3), a normal armpit on examination, and a deposit larger than 2 mm in only 1–2 sentinel nodes; both breast-conserving surgery and mastectomy patients were included.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-SENOMAC

Full scientific title

SENOMAC — omission of ALND for 1–2 sentinel-node macrometastases

Clinical question

In clinically node-negative T1–T3 breast cancer with 1–2 sentinel-node macrometastases (>2 mm), does omitting completion axillary dissection (ALND) compromise survival?

Themes

De-escalation · Surgical technique · Evidence that changed practice

Study design

Non-inferiority randomized controlled trial

Primary endpoint

Overall survival (primary; non-inferiority: upper bound of one-sided 90% CI for HR <1.44). Result reported only as an ASCO 2026 congress abstract; no peer-reviewed full paper

Primary endpoint met?

Met

Study outcomes (by outcome)

  • PrimaryOverall survival — 5 years (PRIMARY endpoint; CONGRESS ABSTRACT — ASCO 2026 LBA503, not a peer-reviewed full paper)

    Result: Original ASCO 2026 congress abstract (J Clin Oncol 2026;44(17_suppl):LBA503; not a peer-reviewed full paper): median follow-up 60.1 months; 196 deaths (75 from breast cancer). 5-year OS ALND 93.4% vs no ALND 94.4%; HR 0.84 (95% CI 0.64–1.12). 5-year breast-cancer-specific survival 97.3% vs 97.8%; HR 0.86 (95% CI 0.55–1.34). Prespecified non-inferiority margin met.

    Statistical interpretation: NI criterion (original congress abstract): upper bound of the one-sided 90% CI for the HR <1.44, with ALND as reference. The reported 95% CI upper bound of 1.12 is below this margin. This is a non-inferiority result, not equivalence or superiority (CI includes 1). Evidence level: congress abstract. No peer-reviewed full paper of the primary endpoint yet; this result is not at the same evidence level as the peer-reviewed NEJM 2024 RFS data.

    Endpoint met?: Met

  • SecondaryRecurrence-free survival — 5 years (secondary endpoint, per-protocol; PEER-REVIEWED — NEJM 2024)

    Result: ALND omitted 89.7% vs ALND 88.7%; country-adjusted HR 0.89 (95% CI 0.66–1.19); non-inferiority p<.001 (margin 1.44). Median follow-up 46.8 months; 191 events.

    Statistical interpretation: Non-inferiority shown for a secondary endpoint; this does not mean equivalence or superiority and does not replace the primary OS result.

    Endpoint met?: Met

  • SecondaryPatient-reported arm morbidity — 3 and 5 years (secondary; CONGRESS ABSTRACT — ASCO 2026 LBA503)

    Result: Original ASCO 2026 abstract: Lymph-ICF arm physical-function difference favouring no ALND 6.14 (5.00–7.43) points at 3 years and 5.71 (4.46–7.25) at 5 years (p<.001). EORTC QLQ-BR23 arm-symptom score higher with ALND: 20.80 vs 10.83 at 3 years, 19.40 vs 9.31 at 5 years (p<.001). No difference in global quality of life. Questionnaire response 81.3% (3 y) / 74.6% (5 y).

    Statistical interpretation: Patient-reported secondary endpoint; taken from the original congress abstract, not yet verified in a peer-reviewed full paper. Interpret cautiously because response rates fell over time.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

Sweden, Denmark, Germany, Greece, Italy; January 2015–December 2021. Age ≥18; clinically node-negative (palpation; preoperative ultrasound performed), T1–T3 invasive breast cancer; macrometastases (>2 mm) in 1–2 sentinel nodes. 2766 enrolled; per-protocol 2540 (1335 SLNB only / 1205 ALND). Per congress report, mastectomy 36% (919/2540) and 87% ER+/HER2− (ASCO Post; not peer-review verified). Neoadjuvant therapy status: could not be verified from the primary source.

Intervention

SLNB alone (no ALND)

Comparator

Completion ALND

Results

Primary endpoint — CONGRESS ABSTRACT (ASCO 2026 LBA503; no peer-reviewed full paper): 5-year overall survival ALND 93.4% vs no ALND 94.4%; HR 0.84 (95% CI 0.64–1.12); prespecified NI margin met. Patient-reported arm function was better and arm symptoms fewer without ALND; no difference in global quality of life. Secondary endpoint — PEER-REVIEWED (NEJM 2024): 5-year recurrence-free survival no ALND 89.7% vs ALND 88.7%; country-adjusted HR 0.89 (95% CI 0.66–1.19); NI margin (1.44) met; median follow-up 46.8 months. Practice impact: peer-reviewed RFS data, with a consistent congress OS result, support ALND omission in some patients outside Z0011 criteria (T3 and mastectomy included; 1–2 macrometastases only). Because most patients received nodal RT, the result should not be transferred to patients not receiving nodal RT. The peer-reviewed publication of the primary OS result is awaited.

Sample size

2766

Registry

NCT02240472

PMID

38598571

DOI

10.1056/NEJMoa2313487

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the NEJM 2024 paper; for the primary overall-survival result, the text of ASCO 2026 congress abstract LBA503; ClinicalTrials.gov NCT02240472. The overall-survival result is at congress-abstract level only; there is no peer-reviewed full paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Swedish Research Council and others (NEJM 2024 abstract); the full list could not be verified from the primary source.

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

The primary OS result is reported only as an ASCO 2026 congress abstract; no peer-reviewed full paper. Reporting note: some secondary congress news items gave a different OS HR (0.89; 0.67–1.17); the card uses 0.84 from the original abstract. The NEJM 2024 RFS result is a separate secondary endpoint (median follow-up 46.8 months). Per-protocol analysis. High nodal RT rate (~89%) — not transferable to patients without nodal RT. Only 1–2 macrometastases; neoadjuvant status and COI: could not be verified from the primary source. The “severe arm problems 12.6% vs 3.6%” reported in some congress news is not in the original abstract and is not used on the card; mastectomy and ER/HER2 rates in the population field rest on congress news only. The “primary endpoint met” status reflects only the primary OS non-inferiority result from the congress abstract; peer-reviewed publication is awaited.
Physician-approved (conditional) · 14 Nov 2026Peritoneum and abdominal cavity · Colon and rectal cancer (incl. its spread)

PRODIGE 7

Does adding heated oxaliplatin chemotherapy (HIPEC) to surgery for bowel cancer spread to the abdominal lining extend life?

Why conditional approval?

The result applies only to the 30-minute oxaliplatin HIPEC regimen used in this trial and cannot be transferred to other HIPEC regimens.

Verified source: Verified at the level of the primary publication abstract and the ClinicalTrials.gov registry protocol.

Conditions kept on this card: The result applies only to the oxaliplatin-based HIPEC regimen studied (30 minutes; registry protocol); PCI subgroups and the completeness-of-cytoreduction (CC) distribution could not be verified.

Key finding

The oxaliplatin HIPEC used in this study did not extend survival when added to surgery and increased serious complications. Both groups had surgical removal of the tumours; the value of this operation itself was not tested in this study.

Why was this study done?

When colon or rectal cancer spreads to the lining of the abdomen (peritoneum), all visible tumour can be removed surgically in suitable patients; this is called cytoreductive surgery. Some centres give heated chemotherapy into the abdomen at the end of the operation (hyperthermic intraperitoneal chemotherapy, HIPEC). The study asked whether adding HIPEC with a drug called oxaliplatin to this operation lengthens survival.

What did it find?

Survival was similar in both groups; adding the oxaliplatin HIPEC used in this study showed no additional benefit. The time by which half of the patients in a group were still alive (median survival) was about 41 months in both groups. This does not mean every patient will live this long. Serious problems within 2 months after surgery were more common with HIPEC: about 26 vs 15 in 100 patients.

What does it mean for me?

The oxaliplatin HIPEC used in this study added no benefit to surgery and increased serious complications. The study did not show that complete surgical removal of tumours is worthless; both groups had this operation.

Which patients does it apply to?

Patients aged 18–70 with colorectal cancer, in good general condition, with a Peritoneal Cancer Index (PCI, a score of how widespread the disease is on the abdominal lining) of 25 or less, whose tumours were completely or almost completely removed at surgery. This result applies only to HIPEC with this drug and this method; it does not directly apply to HIPEC with other drugs or different durations, or to other cancer types. The right treatment is decided with the surgical and oncology team according to the patient’s situation.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-PRODIGE-7

Full scientific title

PRODIGE 7 — cytoreductive surgery with or without oxaliplatin HIPEC for colorectal peritoneal metastases

Clinical question

In patients with colorectal peritoneal metastases in whom complete cytoreduction is achieved, does adding oxaliplatin-based HIPEC to cytoreductive surgery prolong overall survival?

Themes

Surgical technique · Perioperative safety · Evidence that changed practice

Study design

Phase III randomized controlled trial

Primary endpoint

Overall survival (ITT)

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • PrimaryOverall survival (primary)

    Result: Median OS: cytoreductive surgery + HIPEC 41.7 months (95% CI 36.2–53.8) vs cytoreductive surgery alone 41.2 months (35.1–49.7); HR 1.00 (95.37% CI 0.63–1.58); stratified log-rank p=.99. Median follow-up 63.8 months.

    Statistical interpretation: No survival benefit from adding oxaliplatin HIPEC was shown. Because both arms had cytoreductive surgery, the trial does not test the effect of cytoreductive surgery itself. The CI is wide; this is not proof of equivalence either.

    Endpoint met?: Not met

  • Morbidity and mortality

    Result: Treatment-related deaths at 30 days: 2 (2%) in each arm. Grade ≥3 adverse events at 30 days 42% (56/133) vs 32% (42/132) (p=.083); at 60 days 26% (34/131) vs 15% (20/130) (p=.035).

    Statistical interpretation: Adding HIPEC significantly increased serious late complications at 60 days; the 30-day difference was not significant.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

17 cancer centres in France; February 2008–January 2014. Age 18–70, histologically proven colorectal cancer with peritoneal metastases, WHO performance status 0–1, Peritoneal Cancer Index (PCI) ≤25, eligible for 6 months of systemic chemotherapy. Randomization was intraoperative, after complete macroscopic resection or <1 mm residual disease. 265 patients (133 HIPEC / 132 surgery alone). PCI distribution and subgroup results could not be verified from the primary source. Secondary sources report that a post-hoc analysis found a difference favouring HIPEC in the PCI 11–15 subgroup; this is hypothesis-generating, no numbers are given on the card, and it should not be used for treatment decisions.

Intervention

Cytoreductive surgery + oxaliplatin HIPEC (closed technique 360 mg/m², open technique 460 mg/m²; IV fluorouracil 400 mg/m² + folinic acid 20 mg/m² 20 min before HIPEC)

Comparator

Cytoreductive surgery alone

Results

Median OS 41.7 vs 41.2 months (HR 1.00; p=.99): no survival benefit from adding oxaliplatin HIPEC was shown. Grade ≥3 adverse events at 60 days were more frequent with HIPEC (26% vs 15%; p=.035). Practice impact: routine addition of this oxaliplatin-HIPEC protocol for colorectal peritoneal metastases was not supported; the authors conclude that cytoreductive surgery alone should be the cornerstone of curative-intent strategy. The trial does not show that cytoreductive surgery is useless in suitably selected patients. Regimen: oxaliplatin 360 mg/m² (closed) / 460 mg/m² (open technique); IV 5-FU 400 mg/m² + folinic acid 20 mg/m² 20 minutes before HIPEC. HIPEC duration: 30 minutes of hyperthermia according to the ClinicalTrials.gov registry protocol; not verified from the paper’s methods section.

Sample size

265

Registry

NCT00769405

PMID

33476595

DOI

10.1016/S1470-2045(20)30599-4

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the Lancet Oncol 2021 primary paper; ClinicalTrials.gov NCT00769405 registry protocol. The Lancet Oncol full text could not be read in this review. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Institut National du Cancer, Programme Hospitalier de Recherche Clinique du Cancer, Ligue Contre le Cancer (primary paper abstract).

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

The result applies only to this oxaliplatin-HIPEC protocol; it cannot be generalized to other HIPEC drugs (e.g. mitomycin C), protocols or cancer types. Cytoreductive surgery was done in both arms, so its effect was not tested. All patients received systemic chemotherapy; the added benefit of HIPEC was assessed in that context. Age 18–70, PCI ≤25, complete/near-complete cytoreduction — a selected population. Wide CI; a small effect cannot be excluded. PCI subgroups, completeness-of-cytoreduction (CC) distribution and author conflict-of-interest disclosures could not be verified from the primary source (full text not read). The HIPEC duration is taken only from the registry protocol (30 minutes). No separate long-term publication found.
Physician-approved (conditional) · 14 Nov 2026Peritoneum and abdominal cavity · Ovarian cancer

OVHIPEC-1

Does adding heated chemotherapy (HIPEC) to surgery for advanced ovarian cancer extend life?

Why conditional approval?

The result applies only to stage III ovarian cancer and surgery performed between chemotherapy cycles (interval surgery); details of surgical complications were not verified.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Stage III ovarian cancer and interval surgery; 2018 conflict-of-interest statements and details of surgical complications were not verified.

Key finding

In selected patients who had chemotherapy before surgery, HIPEC delayed the return of disease and extended survival; serious side effects were similar.

Why was this study done?

In advanced ovarian cancer, some patients first receive chemotherapy and then have surgery to remove the tumours; this is called interval cytoreductive surgery. The study asked whether giving heated chemotherapy into the abdomen at the end of this operation (hyperthermic intraperitoneal chemotherapy, HIPEC; here with a drug called cisplatin) delays the return of disease and lengthens survival.

What did it find?

In the HIPEC group, disease returned later and survival was longer; HIPEC was shown to be better. The time by which half of the patients in a group were still alive (median survival) was about 34 months with surgery alone and about 46 months with HIPEC; this does not mean every patient will live this long. Serious side effects occurred at similar rates (about 25 and 27 in 100 patients).

What did longer follow-up show? At about 10 years of follow-up, the survival advantage in the HIPEC group persisted. Over this period, about 88 in 100 patients in the surgery-only group and about 82 in 100 in the HIPEC group had died.

What does it mean for me?

For suitable patients with stage III ovarian cancer who receive chemotherapy first and then surgery, HIPEC is an option that can be discussed at an experienced centre.

Which patients does it apply to?

Patients aged 18–76 with stage III epithelial ovarian cancer whose disease did not progress after 3 cycles of chemotherapy and in whom complete or near-complete tumour removal was expected at surgery. It does not directly apply to patients whose first treatment is surgery, to other stages or to bowel cancer.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-OVHIPEC-1

Full scientific title

OVHIPEC-1 — interval cytoreduction with or without cisplatin HIPEC for stage III ovarian cancer

Clinical question

In stage III epithelial ovarian cancer treated with neoadjuvant chemotherapy, does adding cisplatin HIPEC to interval cytoreductive surgery prolong recurrence-free and overall survival?

Themes

Surgical technique · Timing and strategy · Evidence that changed practice

Study design

Phase III randomized controlled trial

Primary endpoint

Recurrence-free survival (NEJM 2018; termed progression-free survival in the 2023 paper)

Primary endpoint met?

Met

Study outcomes (by outcome)

  • PrimaryRecurrence-free survival — 2018 primary analysis

    Result: Recurrence or death: surgery alone 110/123 (89%) vs surgery + HIPEC 99/122 (81%); HR 0.66 (95% CI 0.50–0.87; p=.003). Median RFS 10.7 vs 14.2 months.

    Statistical interpretation: Significant difference in the primary endpoint favouring HIPEC (superiority design).

    Endpoint met?: Met

  • SecondaryOverall survival — 2018 (key secondary)

    Result: Deaths at median 4.7 years: surgery alone 76 (62%) vs surgery + HIPEC 61 (50%); HR 0.67 (95% CI 0.48–0.94; p=.02). Median OS 33.9 vs 45.7 months.

    Statistical interpretation: Significant secondary-endpoint difference favouring HIPEC.

    Endpoint met?: Not applicable

  • Adverse events — 2018

    Result: Grade 3–4 adverse events 25% (surgery alone) vs 27% (surgery + HIPEC); p=.76.

    Statistical interpretation: No significant difference; this is not proof of equivalent toxicity.

    Endpoint met?: Not applicable

  • Long-term survival — 2023 final analysis

    Result: Median follow-up ~10 years. PFS events 114 (93%) vs 109 (89%); median PFS 10.7 vs 14.3 months; HR 0.63 (95% CI 0.48–0.83; p=.0008). Deaths 108 (88%) vs 100 (82%); median OS 33.3 vs 44.9 months; HR 0.70 (95% CI 0.53–0.92; p=.011).

    Statistical interpretation: The long-term analysis supports the 2018 result; the difference favouring HIPEC persists.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

8 HIPEC centres in the Netherlands and Belgium; April 2007–April 2016. Age 18–76, primary stage III epithelial ovarian cancer not eligible for primary cytoreduction; at least stable disease after 3 cycles of neoadjuvant carboplatin + paclitaxel; WHO performance 0–2. Randomization was intraoperative, when complete (no visible disease) or optimal (≤10 mm residual) cytoreduction was anticipated. 245 patients (123 surgery alone / 122 surgery + HIPEC). EudraCT 2006-003466-34.

Intervention

Interval cytoreductive surgery + HIPEC (cisplatin 100 mg/m²); 3 postoperative cycles of carboplatin + paclitaxel

Comparator

Interval cytoreductive surgery alone; 3 postoperative cycles of carboplatin + paclitaxel

Results

2018: RFS HR 0.66 (median 10.7 → 14.2 months); OS HR 0.67 (median 33.9 → 45.7 months); grade 3–4 adverse events 25% vs 27%. 2023 (~10 years): PFS HR 0.63; OS HR 0.70 (0.53–0.92; median 33.3 → 44.9 months). Practice impact: provided randomized survival evidence for cisplatin HIPEC in selected stage III ovarian cancer patients undergoing interval cytoreduction after neoadjuvant chemotherapy (guideline status: could not be verified from the primary source). It cannot be generalized to all ovarian cancer patients or to colorectal peritoneal metastases.

Long-term update

Aronson Lancet Oncol 2023 (median follow-up 10.1 / 10.4 years): PFS HR 0.63 (0.48–0.83; p=.0008), median 10.7 vs 14.3 months; OS HR 0.70 (0.53–0.92; p=.011), median 33.3 vs 44.9 months. The authors state the long-term survival benefit is confirmed.

Sample size

245

Registry

NCT00426257

PMID

29342393

DOI

10.1056/NEJMoa1708618

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstracts of the NEJM 2018 primary paper and the Lancet Oncol 2023 long-term paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Dutch Cancer Society (KWF Kankerbestrijding).

Conflicts of interest

In the 2023 paper some authors reported institutional research support or consultancy fees from drug companies. The statement of the NEJM 2018 paper could not be verified from the primary source.

Limits

Only stage III epithelial ovarian cancer with interval cytoreduction after neoadjuvant chemotherapy; cannot be generalized to primary cytoreduction, stage IV or recurrent disease. The cisplatin-HIPEC result cannot be transferred to colorectal peritoneal metastases or other drugs. Open-label; 8 HIPEC centres (experienced-centre context). Relatively small sample (245). Enrolment 2007–2016; interaction with current systemic therapies (e.g. PARP inhibitors): could not be verified from the primary source. Surgical complication details, BRCA/HRD distribution: could not be verified from the primary source.
Physician-approved (conditional) · 14 Nov 2026Ovarian cancer

LION

In ovarian cancer surgery, do normal-looking lymph nodes need to be removed?

Why conditional approval?

The result applies only to lymph nodes that appeared clinically normal; removing them increased serious complications and early deaths.

Verified source: The primary publication was verified at abstract level; the full text was not reviewed.

Conditions kept on this card: Clinically normal lymph nodes only; lymphadenectomy increased serious complications and early deaths.

Key finding

Removing normal-looking lymph nodes did not extend survival; instead, repeat surgery and death after surgery were more common.

Why was this study done?

In surgery for advanced ovarian cancer, after all visible tumour was removed, normal-looking lymph nodes around the pelvis and main blood vessels were for many years also removed systematically (lymphadenectomy). The study asked whether this lengthens survival.

What did it find?

Removing the lymph nodes did not lengthen survival or delay the return of disease. The time by which half of the patients in a group were still alive (median survival) was about 69 months without node removal and about 66 months with it; this difference is not statistically significant and does not show how long any individual patient will live. With node removal, repeat abdominal surgery for complications was needed in about 12 vs 7 in 100 patients, and death within 2 months after surgery occurred in about 3 vs 1 in 100.

What does it mean for me?

When all tumour has been removed and the lymph nodes look normal, routinely removing them increases risk without benefit.

Which patients does it apply to?

Patients with newly diagnosed ovarian cancer of stage IIB–IV according to the International Federation of Gynecology and Obstetrics (FIGO) system, whose visible tumour was completely removed at surgery and whose lymph nodes looked normal before and during the operation. Removal of enlarged or suspicious lymph nodes was not the subject of this study.

Some information on this card has not yet been verified against a source. Details are in the scientific section.

Show scientific details

SURG-LION

Full scientific title

LION — systematic pelvic and para-aortic lymphadenectomy versus no lymphadenectomy in advanced ovarian cancer

Clinical question

In advanced ovarian cancer with macroscopically complete resection and lymph nodes clinically normal before and during surgery, does systematic pelvic and para-aortic lymphadenectomy prolong overall survival?

Themes

De-escalation · Surgical technique · Perioperative safety · Evidence that changed practice

Study design

Multicentre randomized controlled trial

Primary endpoint

Overall survival

Primary endpoint met?

Not met

Study outcomes (by outcome)

  • PrimaryOverall survival (primary)

    Result: Median OS: no lymphadenectomy 69.2 months vs lymphadenectomy 65.5 months; HR for death with lymphadenectomy 1.06 (95% CI 0.83–1.34; p=.65).

    Statistical interpretation: No survival benefit from systematic lymphadenectomy was shown. This is not proof that the approaches are equivalent; but together with the increased surgical risk, it does not support routine lymphadenectomy.

    Endpoint met?: Not met

  • Progression-free survival

    Result: Median PFS 25.5 months in both groups; HR with lymphadenectomy 1.11 (95% CI 0.92–1.34; p=.29).

    Statistical interpretation: No significant difference.

    Endpoint met?: Not applicable

  • Surgical complications and early mortality

    Result: Repeat laparotomy: lymphadenectomy 12.4% vs no lymphadenectomy 6.5% (p=.01). Death within 60 days after surgery: 3.1% vs 0.9% (p=.049). Other complication rates (e.g. lymphocyst, transfusion): could not be verified from the primary source.

    Statistical interpretation: Lymphadenectomy significantly increased serious postoperative complications and early mortality.

    Endpoint met?: Not applicable

Clinical interpretation

See the outcome-by-outcome results above (not a single mixed badge).

Population

International multicentre (number of centres and countries: could not be verified from the primary source); December 2008–January 2012. Newly diagnosed advanced ovarian cancer (FIGO IIB–IV), macroscopically complete resection, lymph nodes normal before and during surgery. Intraoperative randomization. 647 patients (323 lymphadenectomy / 324 no lymphadenectomy). Centres had to qualify for surgical skills before participation.

Intervention

Systematic pelvic and para-aortic lymphadenectomy (median 57 nodes removed: 35 pelvic, 22 para-aortic)

Comparator

No lymphadenectomy

Results

Median OS 69.2 (no lymphadenectomy) vs 65.5 months (lymphadenectomy); HR 1.06 (0.83–1.34). PFS 25.5 months in both groups. Lymphadenectomy increased repeat laparotomy (12.4% vs 6.5%) and 60-day mortality (3.1% vs 0.9%). Practice impact: supported abandoning routine systematic lymphadenectomy in advanced ovarian cancer with macroscopically complete resection and clinically normal nodes. Cannot be generalized to removal of suspicious/enlarged nodes.

Sample size

647

Registry

NCT00712218

PMID

30811909

DOI

10.1056/NEJMoa1808424

Physician approval

Conditionally approved (14 Nov 2026)

Scientific review

Physician review completed

Source verification

Sources listed; not all details have been verified against the full text

Verified against

Sources compared: PubMed abstract of the NEJM 2019 primary paper. Information that could not be verified is stated explicitly under “Limitations”.

Funding

Deutsche Forschungsgemeinschaft and Austrian Science Fund (primary paper abstract).

Conflicts of interest

No conflict-of-interest statement in the PubMed record; the full text could not be read, so author disclosures could not be verified from the primary source. This does not mean there were no conflicts.

Limits

Only patients with macroscopically complete resection and clinically normal nodes; cannot be generalized to removal of enlarged/suspicious nodes, early-stage disease or patients with residual tumour. No survival difference is not proof of equivalence. Centres were selected for surgical skill. Pathological node positivity in the lymphadenectomy arm, detailed complications and COI: could not be verified from the primary source (no full text). No separate long-term publication found. No landmark card.