05657107

January 07, 2026, 08:06

Reader’s expertise on the topic of the manuscript

Does the reader have a conflict of interest?

The overall quality of the manuscript, based on the above-listed criteria, should be evaluated and classified according to the following five categories

Language quality (style, grammar, and spelling) should be evaluated and classified according to the following five categories.

This meta-analysis systematically retrieved and synthesized evidence from 30 randomized controlled trials (RCTs) involving nearly 17,000 patients, providing the most comprehensive assessment to date on the efficacy of indomethacin for preventing post-ERCP pancreatitis (PEP). It offers valuable, up-to-date evidence-based references for clinical practice, and the authors' efforts are highly commendable. Nevertheless, while acknowledging its contributions, two critical methodological limitations must be highlighted, which may compromise the interpretation and generalizability of its findings. The present commentary aims to identify two key methodological flaws in this meta-analysis that seriously undermine the statistical validity and clinical interpretability of its results. First, the authors inappropriately disaggregated seven multi-arm randomized controlled trials into multiple independent pairwise comparisons for inclusion in the analysis. This practice directly violates the core assumption of data independence in meta-analyses: different comparison groups derived from the same trial are correlated due to the shared control arm. Treating these as independent samples artificially inflates the total sample size, misestimates the weight of each study, and leads to an inappropriate narrowing of confidence intervals, thereby increasing the risk of Type I or Type II errors. Second, the definition of the "control group" in the study encompasses interventions with extremely high clinical heterogeneity, including placebo, normal saline, other active medications (e.g., diclofenac, somatostatin), and invasive procedures (e.g., pancreatic duct stenting). Pooling these controls with vastly different mechanisms of action and therapeutic efficacies renders the reported pooled relative risk (RR = 0.85) clinically meaningless. Furthermore, the high heterogeneity observed (I² = 79%) is most likely attributable to this flawed methodological design. In summary, the aforementioned issues cast doubt on the statistical credibility of the primary conclusion—that "indomethacin does not significantly reduce the incidence of PEP"—and also make it difficult to provide a reasonable clinical interpretation for practice. Given that this review incorporates multiple interrelated interventions for comparison, network meta-analysis would represent a more appropriate methodological framework. It can rigorously integrate data from multi-arm trials and simultaneously evaluate the relative efficacy of all relevant preventive strategies.

Replied on January 19, 2026, 05:57

Thank you for this thoughtful and detailed commentary on our meta-analysis. We sincerely appreciate you engaging so deeply with our work and raising these important methodological points, which certainly warrant careful consideration. You are correct to highlight the challenges inherent in analyzing multi-arm trials and in defining the control group. Regarding the first point, our decision to treat comparisons from multi-arm trials separately was made to avoid losing valuable data on alternative interventions, though we acknowledge this introduces complexity. A sensitivity analysis excluding these trials, however, yielded a consistent direction and magnitude of effect. Concerning the control group heterogeneity, we agree it is significant. Our intent was to address the fundamental clinical question faced at the procedural bedside: "Compared to all other common preventive strategies (be it placebo, other drugs, or stents), what is the overall efficacy of indomethacin?" While this approach creates statistical heterogeneity, we believe the consistent, non-significant effect direction across studies remains a clinically relevant finding, suggesting indomethacin may not be the singularly superior strategy. We fully agree that a network meta-analysis would be an excellent next step to disentangle the comparative efficacy of all these interventions, and we hope our synthesized data can contribute to such a future effort. Your insights are invaluable for interpreting our conclusions and guiding subsequent research.