Comment on: Effect of gender and age on bDMARD efficacy for axial spondyloarthritis patients: a meta-analysis of randomized controlled trials
Bibliographic record
Abstract
Dear Editor, We read with great interest the article by Xie et al. [1] entitled ‘Effect of gender and age on bDMARD efficacy for axial spondyloarthritis patients: a meta-analysis of randomized controlled trials’, which found a preferential response to bDMARDs in males and younger patients with axial spondyloarthritis (axSpA). We commend the authors for focusing on this important topic; however, we would like to highlight several issues related to the analysis, reporting and interpretation of results as presented in this article. First, in their meta-analysis, the authors pooled together trials that exhibited substantial variability in patient population (e.g. radiographic and non-radiographic axSpA) and study intervention (e.g. TNF and IL-17 inhibitors). Instead of using random effects models, which are typically used when variability in effect size is expected across trials, the authors opted for fixed effects models. This decision was seemingly justified by the authors due to the calculated low level of heterogeneity. However, we believe that the selection and justification for fixed effects models are methodologically flawed. The selection of random vs fixed effects method should be guided by the question of whether there is evidence that the true effect size varies across studies [2]. Fixed-effects models operate under the assumption of uniformity in interventions, treating differences in effect size across trials as mere chance variations. Conversely, random-effects models acknowledge that variability in effect size may stem from genuine differences in the interventions across trials, such as variations in the effects of study drugs. The Cochrane Handbook for systematic reviews explicitly cautions against using statistical tests for heterogeneity as the sole basis for choosing between fixed-effect and random-effect meta-analyses [3]. While both random and fixed effects methods may yield similar results in the absence of heterogeneity among studies, fixed effects models tend to underestimate the true confidence interval around the summary estimate when heterogeneity is present (I2 > 0). This discrepancy is evident in their meta-analysis of gender effects, as indicated by Figs 2 and 3, where heterogeneity across studies is apparent. Consequently, the true confidence intervals around the effect size are likely wider than those reported. Second, the authors stated that ASAS40 response served as the primary end point analysed in the meta-analysis. However, it becomes apparent that one of the studies included in the meta-analysis, GO-AHEAD (reference 16) [4], only reported ASAS20 response by gender. In this case, it seems that the authors pooled together the ASAS20 response data from GO-AHEAD with the results from the other studies that reported ASAS40 response, contradicting their stated methodology. As such, we believe that this reference should have been excluded from the meta-analysis, or at the very least, the meta-analysis should have explicitly acknowledged the amalgamation of different study endpoints to ensure clarity. Third, a notable finding in the article lies in the limited reporting of gender-disaggregated study results. Among the 129 RCTs identified in the systematic literature review, only 10 trials reported study endpoints by gender and were consequently included in the meta-analysis. This observation prompts concern regarding publication bias, wherein trials demonstrating significant gender differences in response are more inclined to be published and integrated into the meta-analysis. Despite this finding, the authors have made no attempt to evaluate the potential impact of publication bias on their findings or address it within the article. Overall, we believe that adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement (PRISMA) guidelines [5] will enhance transparency and elevate the quality of future meta-analyses, both in terms of their conduct and reporting. No new data were generated or analysed in support of this article. No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article. Disclosure statement: Lihi Eder received research and educational grants from Abbvie, UCB, Novartis, Pfizer, Janssen, Fresenius Kabi, Sandoz, Amgen and Eli Lilly. He has participated in advisory boards/consulted for Novartis, Janssen, Pfizer, Eli Lilly, UCB and BMS. Philip Mease has received research grants from Abbvie, Acelyrin, Amgen, Bristol Myers Squibb, Eli Lilly, Janssen, Novartis and UCB; consulting fees from Abbvie, Acelyrin, Amgen, Bristol Myers Squibb, Eli Lillly, Inmagene, Janssen, Moonlake Pharma, Novartis, Pfizer, UCB and Ventyx; and speaker fees from Abbvie, Eli Lilly, Janssen, Novartis, Pfizer and UCB. Lianne S. Gensler has received research grants from Novartis and UCB, and advisory/consulting honoraria for AbbVie, Acelyrin, Eli Lilly, Novartis, Pfizer and UCB. The remaining author has declared no conflicts of interest.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.026 | 0.159 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.004 | 0.008 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.003 | 0.004 |
| Open science | 0.006 | 0.002 |
| Research integrity | 0.024 | 0.019 |
| Insufficient payload (model declined to judge) | 0.015 | 0.007 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".