The randomized controlled trial (RCT) may be the gold standard of clinical research, but not all RCTs are created equal. In the December 20, 2017 issue of The Journal, Kay et al. set out to determine the historical quality of more than 400 RCTs relating to anterior cruciate ligament (ACL) reconstruction published between 1985 and 2016.
The authors evaluated the studies using the Detsky quality index (on which a score of ≥75% is considered “high” methodological quality) and the CONSORT checklist. Fifty-two percent of the studies analyzed were surgical trials (graft choice, femoral fixation method, and single- vs double-bundle techniques), while 48% were nonsurgical trials (analgesic and rehabilitation regimens).
The mean Detsky score across all studies was 68.9% ± 13.2%. Forty-two percent of the studies were considered high-quality according to Detsky scores. There was no significant difference in mean Detsky scores between surgical and nonsurgical trials.
The authors noted a substantial increase in the number of ACL RCTs over time, although that number has not changed significantly since 2009, which is coincidentally the year that JBJS and many other journals required authors to use the CONSORT flow diagram. Notably, the reporting of methodologically sound randomization processes and the prospective calculation of sample sizes have also improved significantly in recent years.
Both the inclusion of a CONSORT flow diagram and the year published were significant predictors of the quality score. A mere 5% of trials published in 2008 or prior included the CONSORT flow diagram, while >40% of trials published since 2009 had a flow diagram.
While the authors are encouraged that “close to 50% of trials in this study reported blinding of outcome assessors,” they say areas for improvement include better “descriptions of allocation concealment and intent-to-treat analyses.”