No differential attrition was found in randomized controlled trials published in general medical journals: a meta-analysis

Abstract

Objective: Differential attrition is regarded as a major threat to the internal validity of a randomized controlled trial (RCT). This study identifies the degree of differential attrition in RCTs covering a broad spectrum of clinical areas and factors that are related to this.

Study design and setting: A PubMed search was conducted to obtain a random sample of 100 RCTs published between 2008 and 2010 in journals from the ISI Web of Knowledge(SM) category of medicine, general and internal. Eligibility criteria for selecting studies were primary publications of two-arm parallel randomized clinical trials, containing human participants and one or multiple follow-up measurements whose availability depended on the patients' willingness to participate.

Results: A significant amount of differential attrition was observed in 8% of the trials. The average differential attrition rate was 0.99 (95% confidence interval: 0.97-1.01), indicating no general difference in attrition rates between intervention and control groups. Moreover, no indication of heterogeneity was found, suggesting that the occurrence of differential attrition in the published literature is mostly a chance finding, unrelated to any particular design factors.

Conclusion: Differential attrition did not generally occur in RCTs covering a broad spectrum of clinical areas within general and internal medicine.

Keywords: Bias; Differential attrition; Internal validity; Loss to follow-up; Meta-analysis; RCT.