Advertisement for orthosearch.org.uk
Orthopaedic Proceedings Logo

Receive monthly Table of Contents alerts from Orthopaedic Proceedings

Comprehensive article alerts can be set up and managed through your account settings

View my account settings

Visit Orthopaedic Proceedings at:

Loading...

Loading...

Full Access

TOO GOOD TO BE TRUE? AN EVALUATION OF STATISTICALLY SIGNIFICANT FINDINGS IN ORTHOPAEDIC TRAUMA RANDOMISED TRIALS



Abstract

We aimed to quantify the sample sizes and magnitude of treatment effects in a review of orthopaedic randomised trials with statistically significant findings.

We conducted a comprehensive search (PubMed, Cochrane) for all randomised controlled trials between 1/1/95 to 12/31/04. For continuous outcome measures (ie functional scores), we calculated effect sizes (mean difference/standard deviation). Dichotomous variables (ie infection, nonunion) were summarised as absolute risk differences and relative risk reductions (RRR). Effect sizes > 0.80 and RRRs> 50% were defined as large effects.

Our search yielded 433 RCTs, of which 76 RCTs with statistically significant findings on 184 outcomes (122 continuous/62 dichotomous outcomes) met study eligibility criteria. The mean effect size across studies with continuous outcome variables was 1.7 (95% confidence interval: 1.43–1.97). Almost one in three results, despite being reported as statistically significant did meet the definition of a large effect size (ES< 0.80). For dichotomous outcomes, the mean risk difference was 30% (95%confidence interval:24%–36%) and the mean relative risk reduction was 61%.

Correspondence should be addressed to: Cynthia Vezina, Communications Manager, COA, 4150-360 Ste. Catherine St. West, Westmount, QC H3Z 2Y5, Canada