Risk of Bias and Its Impact on Intervention Effect Estimates of Randomized Controlled Trials in Endodontics

Introduction: The aims of this study were to evaluate the methodological quality of randomized controlled trials (RCTs) recently published in endodontics and to investigate the influences of methodological characteristics on the magnitude of treatment effects. Methods: PubMed was searched for RCTs published from October 2013 to October 2018 in 3 leading endodontic journals. The methodological quality of the included studies was determined by using the Cochrane Collaboration risk of bias (RoB) tool. The estimates of intervention effects were expressed or calculated as odds ratios and the standardized mean difference for binary and continuous outcomes, respectively. Meta-regression analyses and Monte Carlo permutation tests were performed to identify the association between RoB and intervention effect estimates. Results: A total of 121 RCTs were identified as eligible for the current study. For both the studies with binary and continuous outcome measures, the domain of blinding of participants and personnel had the highest percentage of high RoB. For binary outcomes, methodological deficiencies in allocation concealment tended to produce exaggerated treatment effects. For continuous outcomes, risk regarding blinding of participants and personnel and incomplete outcome data were more likely to provide overestimated trial results. Conclusions: The methodological quality of RCTs within endodontics is suboptimal, and these methodological deficiencies could exaggerate intervention effect estimates in endodontic RCTs. Better trial methodology and more explicit reporting are needed to improve the reliability of evidence in endodontic RCTs.