Does the frequency of follow-up assessments affect clinical trial outcome? A meta-analysis and meta-regression of placebo-controlled randomized trials

The number and temporal distribution of follow-up assessments during a clinical trial is a critical factor which may influence the outcome as well as the overall cost of a trial. Therefore, we aimed to examine whether the overall and differential frequency of study observations during the course of a clinical trial affects the risk ratio (RR) of responding to antidepressants vs. placebo, specifically in trials for major depressive disorder (MDD). Medline/Pubmed publication databases were searched for randomized, double-blind, placebo-controlled trials of antidepressants for adults with MDD (1 January 1980-11 May 2010). A total of 142 manuscripts involving 256 drug-placebo comparison were pooled (n=38 860). We found that higher overall frequency (OF, frequency of assessments during the entire trial) and higher late frequency (LF, frequency of assessments after the first 3 wk of the trial), but not higher early frequency (EF, frequency of assessments during the first 3 wk of the trial), of follow-up visits predicted a significantly greater RR of responding to antidepressant vs. placebo (coefficient=0.213, p=0.014; coefficient=0.238, p=0.003; and coefficient=0.021, p=0.755, respectively, for OF, LF and EF). None of the measures of frequency examined (OF, EF, LF) significantly predicted the RR of discontinuing antidepressant vs. placebo. These findings suggest that increasing the number of follow-up visits, specifically after the third week rather than within the first 3 wk of the trial, may be an effective approach to improve the likelihood of success in placebo-controlled clinical trials for MDD.