Adherence to guidance on registration of randomised controlled trials published in Anaesthesia

The International Committee of Medical Journal Editors recommends the prospective registration of interventional clinical trials. We aimed to assess the compliance with these guidelines for manuscripts submitted to and published by a single anaesthetic journal. We examined the rates of prospective trial registration, the incidence of discrepancies in primary outcome measure(s) and sample sizes, and the citation metrics of all randomised controlled trials published in Anaesthesia over a 3-year period (2014-2016). Of the 422 randomised controlled trials submitted during the study period, 115 (27.3%) were accepted for publication, of which 90 (78.3%) were patient studies, with the remaining 25 comprising manikin, simulation, volunteer, bench, cadaver and other non-patient intervention studies. Of the accepted patient studies, 64 (71.1%) were prospectively registered with a clinical trials registry, 20 (22.2%) were not registered and 6 (6.7%) were retrospectively registered after manuscript submission. There was no difference in the frequency of registration between accepted and rejected manuscripts (77.8% vs. 84.5%, respectively, p=0.143). The median (IQR [range]) time from registration of accepted manuscripts to journal submission was 701 (331-1341 [99-2436])days. There was no correlation between number of patients recruited to a study and time to submission. Fifty-two (81.3%) of the prospectively registered studies reported the same primary outcomes in both registration and submission, and 34 (53.1%) studies were published with the same powered sample size as that described in the registry. Eleven (12.2%) studies recruited more patients and 19 (21.1%) recruited fewer patients than described in the registration protocol. There was no difference in the median (IQR [range]) number of citations per month since publication between prospectively (0.27 (0.15-0.46 [0.00-1.59]), and retrospectively (0.39 (0.15-0.62 [0.10-0.67]); p=0.502) or unregistered (0.33 (0.10-0.52 [0.00-0.67]); p=0.867) studies. Our results suggest that prospective clinical trial registration has no influence on acceptance for publication by Anaesthesia or subsequent citation metrics. The international recommendation for prospective trial registration appears to have not been universally incorporated into anaesthetic-related research practice.