Indomethacin with or without prophylactic pancreatic stent placement to prevent pancreatitis after ERCP: a randomised non-inferiority trial.
Elmunzer BJ, Foster LD, Serrano J, Coté GA, Edmundowicz SA, Wani S, Shah R, Bang JY, Varadarajulu S, Singh VK, Khashab M, Kwon RS, Scheiman JM, Willingham FF, Keilin SA, Papachristou GI, Chak A, Slivka A, Mullady D, Kushnir V, Buxbaum J, Keswani R, Gardner TB, Forbes N, Rastogi A, Ross A, Law J, Yachimski P, Chen YI, Barkun A, Smith ZL, Petersen B, Wang AY, Saltzman JR, Spitzer RL, Ordiah C, Spino C, Durkalski-Mauldin V, SVI Study Group
- DOI
- 10.1016/S0140-6736(23)02356-5
- Record issued
- 2026-08-16
- Engine
- 7.39.0
- Exported
- 2026-09-20
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/c1286736-9676-4e81-88bb-c7c3096860bd is authoritative.
How this rating was calculated
- StatisticsStatistic did not reproduce ×3−1.5★
- IntegrityIntegrity concern−0.5★
- The numeric-impossibility checks (GRIM/GRIMMER/DEBIT/SPRITE) did not run: 12 reported means were read, and their group size is not stated where the values are printed. These checks need the count the mean was averaged over, so none was performed.
- 01Printed percentage does not match its own count
41.8% does not match the reported count 366/975
“366 (41.8%)”
Table 1 - 02Printed percentage does not match its own count
41.8% does not match the reported count 363/975
“363 (41.8%)”
Table 1 - 03Significance claim does not survive recomputation
Non-inferiority p-value for primary outcome (risk difference 3.6%, 95% CI 0.6% to 6.6%)
“risk difference, 3.6%; 95% confidence interval, 0.6%−6.6%; p=0.18 for non-inferiority”
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-conducted multicenter randomized non-inferiority trial with strong design, clear reporting, and a concrete data sharing plan. The main weaknesses are minor reporting gaps (randomization method, informed consent, statistical software, CONSORT mention) and a potentially serious inconsistency in the primary non-inferiority p-value that warrants verification.
Both reviewers agreed on study type (interventional) and on most dimensions. The only substantive divergence was in statistical analysis, where the deterministic verification component found an inconsistency that overrides the reviewers' pass. The statistics verification covered only a subset of tests (those with test statistics/CI); other reported statistics were not independently verified.
Numerical inconsistencies
2 findings · worst highValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Summary statistic impossible for the stated N (GRIM/GRIMMER)Recomputed
- Internal contradictions in the reported numbersAssessed
Recomputed 3 tests: 2 consistent, 1 inconsistent (1 change significance at p<.05); 3 via agent-written checks. 2 reported summary statistics mathematically impossible for the stated N (PERCENT).
- PERCENT41.8% does not match the reported count 366/975
“366 (41.8%)”
Table 1 - PERCENT41.8% does not match the reported count 363/975
“363 (41.8%)”
Table 1
- CONSISTENTreported p = .010 · recomputed p = .019Reviewer 1Post hoc superiority p-value for primary outcome (risk difference 3.6%, 95% CI 0.6% to 6.6%)
“a post hoc one-sided test of the comparison of the two proportions indicated that indomethacin alone was inferior to indomethacin plus a prophylactic stent (p=0.01)”
Taken as given: The risk difference is 0.036; The 95% CI is two-sided; The p-value is for a one-sided test of superiority (risk difference > 0)Method: Computed one-sided p-value from the estimate and CI using pCI, which gives the two-sided p; the one-sided p is half of that.How we recomputed it: pCI(0.036, 0.006, 0.066, 0) - CONSISTENTreported p = .010 · recomputed p = .020Reviewer 2Post hoc test for inferiority (one-sided)
“a post hoc one-sided test of the comparison of the two proportions indicated that indomethacin alone was inferior to indomethacin plus a prophylactic stent (p=0.01)”
Taken as given: The test is a one-sided z-test for difference in proportions; The z-statistic is approximately 2.33Method: Used pZ with z=2.33 to approximate one-sided p-value.How we recomputed it: pZ(2.33)
- lowinternal contradictionThe abstract states 'p=0.18 for non-inferiority' while the results section states 'p=0.18 for non-inferiority' - consistent. However, the post hoc p=0.01 is described as 'one-sided' in results but the abstract says 'p=0.01' without specifying one-sided.
p=0.18 for non-inferiority ... p=0.01
Abstractreviewer’s wording
Overstated conclusions
1 finding · worst highConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Significance claim flips when recomputedRecomputed
- MORE SIGNIFICANT ON RECHECKINCONSISTENTreported p = .180 · recomputed p = .019Reviewers 1, 2Non-inferiority p-value for primary outcome (risk difference 3.6%, 95% CI 0.6% to 6.6%)Recomputing from the paper’s own numbers lands below p = 0.05 — more significant than the printed value. Usually benign (the reported figure is conservative), but the two don’t match.
“risk difference, 3.6%; 95% confidence interval, 0.6%−6.6%; p=0.18 for non-inferiority”
Taken as given: The risk difference is 0.036 (3.6%); The 95% CI is two-sided; The p-value is for a one-sided non-inferiority test with margin 5%Method: Computed p-value from the estimate and CI using the pCI function, which assumes a normal approximation for the risk difference.How we recomputed it: pCI(0.036, 0.006, 0.066, 0)
6 major claims checked against the paper's own evidence: all adequately supported.
- supportedReviewer 1Indomethacin alone is not as effective as indomethacin plus prophylactic stent for preventing post-ERCP pancreatitis.The primary outcome shows a higher rate of pancreatitis in the indomethacin alone group, and non-inferiority was not demonstrated.Evidence: Primary outcome: 145/975 (14.9%) vs 110/975 (11.3%), risk difference 3.6% (95% CI 0.6-6.6%), p=0.18 for non-inferiority.
“Post-ERCP pancreatitis occurred in 145 of 975 patients (14.9%) in the indomethacin alone group and in 110 of 975 (11.3%) in the indomethacin plus stent group (risk difference 3.6%; 95% confidence interval, 0.6%−6.6%; p=0.18 for non-inferiority).”
AbstractFind in source - supportedReviewer 1The relative benefit of stent placement is more prominent among patients at highest risk for pancreatitis.Subgroup analyses show increasing absolute risk reduction with higher risk scores, though the interaction was not statistically significant.Evidence: NNT decreased from 44 (risk score 1-1.5) to 7 (risk score 3+).
“The absolute risk reduction associated with indomethacin plus stent ranged from a number needed to treat (NNT) to prevent one case of ERCP-related pancreatitis of 44 when the risk score was 1–1.5, to 38 when the risk score was 2–2.5, to 7 (95% CI: 3.9–56) when the risk score was 3+.”
ResultsFind in source - supportedReviewer 1The findings support the use of prophylactic pancreatic stent placement in addition to rectal indomethacin.The trial demonstrates that indomethacin alone is inferior, supporting the combination strategy.Evidence: Primary outcome results and post hoc superiority test (p=0.01).
“In summary, a strategy of indomethacin alone was not as effective as a strategy of indomethacin plus prophylactic pancreatic stent placement.”
DiscussionFind in source - supportedReviewer 2Indomethacin alone was not as effective as indomethacin plus prophylactic pancreatic stent placement for preventing post-ERCP pancreatitis.The primary outcome shows a higher rate of pancreatitis in the indomethacin alone group, and non-inferiority was not demonstrated.Evidence: Primary outcome: 145/975 (14.9%) vs 110/975 (11.3%), risk difference 3.6% (95% CI 0.6-6.6%), p=0.18 for non-inferiority.
“a strategy of indomethacin alone was not as effective as a strategy of indomethacin plus prophylactic pancreatic stent placement.”
AbstractFind in source - supportedReviewer 2The relative benefit of stent placement was more prominent among patients at highest risk for pancreatitis.Subgroup analyses show increasing benefit with higher risk scores, though interaction tests were not all significant.Evidence: NNT decreased from 44 to 7 across risk score categories; subgroup analyses show consistent benefit.
“The relative benefit of stent placement was generally consistent across study subgroups but appeared more prominent among patients at highest risk for pancreatitis.”
AbstractFind in source - supportedReviewer 2The widespread abandonment of prophylactic stent placement is not justified.The trial demonstrates benefit of stent placement, supporting the claim.Evidence: Primary outcome results and discussion.
“The findings of this trial indicate that this widespread abandonment of prophylactic stent placement is not justified.”
Discussion ¶4Find in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary endpoint is post-ERCP pancreatitis, a clinical outcome defined by a consensus definition (new or increased abdominal pain, elevation of pancreatic enzymes ≥3 times upper limit of normal, and hospitalization for ≥2 nights). This is a hard clinical outcome, not a surrogate biomarker. The trial also reports secondary outcomes including moderate-severe pancreatitis, severe pancreatitis, and pancreatitis-related death, all clinical outcomes.
“The primary outcome measure was post-ERCP pancreatitis, based on a widely-validated consensus definition that was applied as a diagnostic framework.”
- ADEQUATEEffect sizeThe primary effect is a risk difference of 3.6% (95% CI 0.6%–6.6%) in post-ERCP pancreatitis between indomethacin alone and indomethacin plus stent, with a relative risk of 1.32 (95% CI 1.05–1.66). The effect is statistically significant (p=0.01 for inferiority) and clinically meaningful as it represents a 32% relative increase in the primary outcome, with a number needed to treat of 7 in the highest risk subgroup. The trial was designed with a non-inferiority margin of 5%, and the upper bound of the CI exceeded this, indicating a clinically important difference.
“The indomethacin alone group had a 32% higher risk of developing post-ERCP pancreatitis relative to the combination (RR 1.32; 95% CI 1.05–1.66).”
Data authenticity concerns
None foundAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
Checked — nothing surfaced.
Reporting gaps
None foundRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
Checked — nothing surfaced.
The introduction cites prior trials on NSAIDs and stents, acknowledges limitations of prior studies (open-label, expert centers), and provides a rationale for testing indomethacin alone. The hypothesis follows logically from the cited evidence.
“the perceived benefit of prophylactic stent placement may be exaggerated as prior trials were un-blinded in design and conducted by a small number of expert centers, limiting generalizability to broader practice settings”
“To reappraise the need for prophylactic stent placement among patients receiving rectal NSAIDS, we performed a comparative effectiveness trial of rectal indomethacin alone vs. the combination of indomethacin plus a prophylactic pancreatic stent”
“the perceived benefit of prophylactic stent placement may be exaggerated as prior trials were un-blinded in design and conducted by a small number of expert centers, limiting generalizability to broader practice settings”
“To reappraise the need for prophylactic stent placement among patients receiving rectal NSAIDS, we performed a comparative effectiveness trial of rectal indomethacin alone vs. the combination of indomethacin plus a prophylactic pancreatic stent”
Randomization method is not explicitly described but is implied by 'randomized in a 1:1 ratio'. Blinding is well described. Power analysis is detailed. Inclusion/exclusion criteria are pre-specified. Outlier handling is addressed via per-protocol analysis. Controls are inherent in the comparator arm. Independent replication is not applicable for a single pivotal trial.
“Patients, treating clinicians, and outcomes assessors were blinded to study group assignment.”
“We thus estimated that a sample of 1430 patients would be required to give the trial at least 85% power to rule out a 5% higher rate of post-ERCP pancreatitis in the indomethacin alone arm, at a one-sided error rate of 0.025, and assuming 5% treatment crossover.”
“patients were randomized in a 1:1 ratio to receive a prophylactic stent or not”
Sex, age, race, and comorbidities are reported in Table 1. Since both sexes are enrolled, sex_justified is not applicable. Species/strain and housing are not applicable for a human trial.
“Age, Mean (SD) | 55.8 (16.3) | 55.6 (16.4)”
“Age, Mean (SD) | 55.8 (16.3) | 55.6 (16.4) | | Male Sex, n (%) | 379 (38.9%) | 376 (38.6%)”
The paper states approval from human studies review committee at each institution. Informed consent is implied by the trial's conduct, though not explicitly described. Regulatory compliance is implied by the trial's conduct.
“we enrolled eligible patients at 20 referral centers in the United States and Canada after approval from the human studies review committee at each institution.”
“after approval from the human studies review committee at each institution”
Indomethacin is named with dose and route. Stent types are described in Table 1. No other biological/chemical resources are used. Software tools are not mentioned, but no bespoke software is used.
“All patients received two 50 mg indomethacin suppositories.”
“All patients received two 50 mg indomethacin suppositories.”
“Prophylactic pancreatic stent caliber in those who received a stent, n/N (%)”
Tests are named (risk difference, non-inferiority test). Assumptions are handled by design. Exact p-values are reported. Effect sizes with CIs are provided. Software is not identified, but this is a minor omission. Data presentation is adequate for a clinical trial. Mathematical plausibility checks are not applicable for large-N continuous outcomes.
“p=0.18 for non-inferiority”
“risk difference, 3.6%; 95% confidence interval, 0.6%−6.6%”
“risk difference, 3.6%; 95% confidence interval, 0.6%−6.6%; p=0.18 for non-inferiority”
“a post hoc one-sided test of the comparison of the two proportions indicated that indomethacin alone was inferior to indomethacin plus a prophylactic stent (p=0.01)”
The data sharing statement specifies that de-identified participant data will be shared as a public use dataset via NIH, with a timeframe. This is a concrete access route. Repository deposit and accession numbers are not applicable for patient-level data. Code sharing is not applicable.
“All individual de-identified participant data collected during the conduct of this trial and a corresponding data dictionary will be shared as part of a public use dataset in accordance with US National Institutes of Health policies.”
“All individual de-identified participant data collected during the conduct of this trial and a corresponding data dictionary will be shared as part of a public use dataset in accordance with US National Institutes of Health policies.”
Trial registration number is provided. Methods are detailed. Limitations are explicitly discussed. Conclusions are proportional to the evidence. Funding and COI are disclosed.
“the results of this trial should be interpreted in the context of important limitations related to blinding and generalizability.”
“the results of this trial should be interpreted in the context of important limitations related to blinding and generalizability.”
Registered (1 ID: ClinicalTrials.gov). No reporting guideline cited.
Broken references and links
None foundReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Checked — nothing surfaced.
Checked 26 references by DOI: 25 verified — 1 no DOI (shown, not verified).
- NO DOIInflammatory mediators and cytokines—new aspects of pathophysiology and assessment of severity of acute pancreatitis?No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
1 data/code link checked; 1 live.
- datahttps://clinicaltrials.gov/ct2/show/NCT02476279LIVEHTTP 200Resolves, but the content could not be matched to the paper.
Copyediting
6 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 6 minor suggestions below.
6 copyedit issues flagged: mostly consistency, typo, clarity.
- MINORconsistencyAbstract, Findings“p=0.18 for non-inferiority”→ Consider reporting the p-value as p=0.18 for non-inferiority (one-sided) to clarify.The p-value is for a one-sided test, but the abstract does not specify.
- MINORtypoTable 1, footnote“millileters”→ Change to 'milliliters'.Spelling error.
- MINORclarityMethods, Statistical analysis“a post hoc one-sided test of the comparison of the two proportions indicated that indomethacin alone was inferior”→ Clarify that this is a post hoc test for superiority, not part of the primary non-inferiority analysis.The term 'post hoc' may be confusing; consider 'exploratory'.
- MINORtypoAbstract, Findings“p=0.18 for non-inferiority”→ Ensure p-value formatting is consistent (e.g., p = 0.18).Spacing around equals sign is inconsistent.
- MINORconsistencyResults, Study outcomes“p=0.01”→ Use consistent p-value formatting throughout.Same as above.
- MINORclarityMethods, Statistical analysis“a one-sided error rate of 0.025”→ Clarify that this is the alpha level for the non-inferiority test.Could be misinterpreted.
The published work is generally robust, but the inconsistency in the primary non-inferiority p-value (reported 0.18 vs recomputed ~0.019) is a validity concern that an informed reader should weigh heavily. A correction or independent re-analysis of the primary outcome is warranted. The reporting gaps (randomization method, consent, software) are minor and do not undermine the overall conclusions.
- 1.HIGHstatisticsRe-analyze the primary non-inferiority outcome and verify the reported p-value (0.18) against the recomputed value (~0.019); if the recomputed value is correct, issue a correction to the abstract and results.The verification component found a large discrepancy in the primary endpoint p-value, which could change the interpretation of the non-inferiority conclusion.
- 2.HIGHreportingAdd a statement in the Methods describing the randomization method (e.g., computer-generated sequence, block size, allocation concealment).The current description only says 'randomized in a 1:1 ratio', which is insufficient for reproducibility.
- 3.HIGHethicsAdd an explicit statement in the Methods that informed consent was obtained from all participants, including the type (written/oral) and any waiver details.Informed consent is a fundamental ethical requirement and its absence is a reporting gap that should be corrected.
- 4.HIGHreportingIdentify the statistical software used (e.g., SAS version, R version) in the Statistical analysis section.Software identification is part of standard reporting and aids reproducibility.
- 5.MEDIUMreportingMention adherence to CONSORT reporting guidelines and include a CONSORT flow diagram.Explicitly stating CONSORT adherence improves transparency and reader confidence.
- 6.MEDIUMcopyeditFix the typo 'millileters' to 'milliliters' in Table 1 footnote.Spelling errors detract from professionalism.
- 7.MEDIUMcopyeditClarify in the Abstract and Results that the non-inferiority p-value is one-sided, and use consistent p-value formatting (e.g., p = 0.18).The one-sided nature is important for interpretation and formatting consistency is expected.
- 8.MEDIUMcopyeditIn the Methods, clarify that the 'post hoc one-sided test' is an exploratory analysis, not part of the primary non-inferiority analysis.The term 'post hoc' may confuse readers about the analysis hierarchy.
- 9.MEDIUMreportingAdd a statement about the availability of the full study protocol and statistical analysis plan in the Data Sharing Statement.This enhances transparency and allows independent verification.
- 10.LOWreportingClarify the role of the funding agency in the design and conduct of the trial.This addresses potential bias concerns.
- 11.LOWreportingInclude a statement about the DSMB charter and its independence in the Methods.This is good practice for clinical trials.
- 12.LOWreportingSpecify the exact timing of indomethacin administration relative to ERCP in the protocol summary.This detail is clinically important and currently missing.
- 13.LOWstatisticsAdd a note about the handling of missing data for the primary outcome in the statistical analysis section.Missing data handling is a key methodological detail.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.