Risk of bias for randomized controlled trials in Journal of Clinical Monitoring and Computing.

Abstract

Well-designed randomized controlled trials (RCTs) are considered to represent a high level of evidence and influence medical decision-making in evidence-based medicine. When biases occur in study design, processing, and reporting of RCTs, however, it is difficult to interpret results and judge the impact of interventions. Accordingly, we evaluate the quality of RCT reporting published in the Journal of Clinical Monitoring and Computing (JCMC) using three assessment tools. Reporting quality of RCTs published in the JCMC was evaluated through December 31, 2020, using Jadad and van Tulder scales and the Cochrane Collaboration's risk of bias tool (CCRBT). Stepwise regression analysis was performed to identify factors associated with reporting quality. Database searches confirmed 132 RCTs in 1,507 original articles. The numbers of RCTs meeting criteria for high reporting quality were 97 (73.5%) using the Jadad scale, 99 (75.0%) using the van Tulder scale, and 19 (14.4%) with the CCRBT. Jadad scores [median score (interquartile range) = 3.0 (2.0-5.0), coefficients (95% CI) = 0.08 (0.04, 0.11), p < 0.001], van Tulder scores [median score (interquartile range) = 7.0 (5.0-8.75), coefficients (95% CI) = 0.15 (0.11, 0.20), p < 0.001], and CCRBT assessment [coefficients (95% CI) = 0.04 (0.02, 0.06), p < 0.001] increased significantly with publication year. The median score (interquartile range) of the last 5 years were 4.0 (3.0-5.0) in Jadad scores, and 8.0 (6.0-9.0) in van Tulder scores. Only 33.3% and 37.1% of articles described detailed blinding and allocation methods, respectively. Reporting quality increased over time, with consistently high reporting quality in recently published JCMC RCTs.