I enjoyed reading Trager and colleagues' 1 recently published analysis of spinal manipulative therapy (SMT) outcomes when compared with ibuprofen use for lower back pain (LBP) patients. I commend the authors for their methodological transparency through the provision of Supporting Information File detailing exactly how they designed patient cohorts in TriNetX. This level of transparency is not always apparent in TriNetX-based studies 2 and allows readers to more intimately understand how the cohorts analyzed were constructed. When examining these Supporting Information, I noticed some potential oversights in the authors' query design, which I outline below. The authors' Supporting Information File 1 provides a graphical representation of their study design, which closely matches the permitted querying logic of the TriNetX platform. Notably, an inclusion criterion for LBP was missing from this study design, despite this condition being the main focus of the paper. On October 20, 2025, I replicated their study using their exact logic (Table S1). Since the authors conducted their study, the TriNetX Research Network has grown in size by 8.5%, now containing data on 153 million patients. This, in combination with the increased time for patients to fulfill the follow-up visit inclusion criterion, resulted in a slightly larger cohort of SMT patients (n = 29,454) compared with the original study (n = 25,274). However, when a single record of any LBP code before or on the same day as the patients' index SMT event was required for patient inclusion (Table S2), only 13,773 patients remained. This indicates that potentially less than half of the included patients received SMT for an LBP indication. This limitation is not acknowledged in the article. When I replicated the ibuprofen cohort, including patients with an LBP diagnosis at any time (Table S3), I produced a cohort of 723,444 patients, which is indicative of replicating the original authors' cohort of 597,406 patients when accounting for growth in the TriNetX network. However, when I restricted the query to only include patients with an LBP diagnosis prior to or on the index ibuprofen prescription (Table S4), the cohort size reduced to 494,730 patients. This suggests that approximately 17% of the control cohort were not diagnosed with LBP at study entry. This limitation is not acknowledged in the article. It is therefore important for readers to appreciate that the findings of Trager and colleagues are potentially not specific to LBP patients. The analysis of opioid use disorder outcomes could be repeated using these amended cohorts to determine if this lack of LBP indication influences the conclusions of the original study. Regardless, I once again thank the authors for their design transparency and hope that this postpublication dialogue can assist other health science researchers in designing and reviewing TriNetX-based studies. Joshua Wang: conceptualization, investigation, writing – original draft, writing – review and editing, methodology. The author has nothing to report. The author has nothing to report. The author declares no conflicts of interest. The lead/corresponding author (Joshua Wang) affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained. The full cohort definitions used to construct the cohorts discussed are provided in the Supporting Information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Joshua Wang (Wed,) studied this question.