Do academic oral and maxillofacial surgeons comply with best practices for reporting the results of randomized clinical trials

Abstract Purpose To address shortcoming of reporting p-values, several leading medical journals have revised their authors’ guidelines in reporting results, specifically mandating the use of confidence intervals. The purpose of this study was to evaluate the frequency of reporting confidence intervals in randomized clinical trials (RCTs) by academic oral and maxillofacial surgeons. Methods The authors implemented a retrospective cohort study and enrolled a sample composed of RCTs published in three oral and maxillofacial surgery (OMS) journals in 2009, 2012, 2015, and 2018. The authors identified RCTs using medical subject headings (MeSH) search in MedLine®. Studies were included in the sample if they were randomized and reported a measure of association. Predictor variables included the year published, journal, OMS focus area (e.g. dentoalveolar surgery, anesthesia/facial pain, or craniomaxillofacial trauma), reported confounding factors, funding sources, conflict of interest, study region, number of institutions involved, number of authors, and academic rank of authors. The primary outcome of interest was reporting of the confidence interval. Descriptive and bivariate statistics were computed. A p-value 0.05 was considered significant. Results The sample was composed of 102 publications, the p-value was uniformly reported (102, 100%) in all publications. Confidence intervals were reported in 28.4% (n = 29, 95% CI: 19.9-38.2) of the papers. OMS focus area was associated with reporting confidence intervals (p = 0.008). Anesthesia/facial pain studies were significantly less likely to report the confidence interval (12.0%) compared to studies on craniomaxillofacial deformities (100%, Bonferroni p = 0.02). There were no other significant associations between the predictors and confidence interval reporting (p > 0.10). Conclusions Confidence intervals are reported in a small proportion of RCTs within OMS. Given the shortcomings of reporting only the p-values, there is a significant need for improving statistical reporting standards among OMS journals.