Relevance of exclusion criteria in antidepressant clinical trials: a replication study

Abstract

Introduction: Patients in clinical randomized controlled trials (RCTs) of antidepressants are different from those typically treated in clinical practice, which may affect the generalizability of data from RCTs. We attempted to replicate the work of Zimmerman and colleagues (Zimmerman M, Mattia JI, Posternak MA. Are subjects in phamacological treatment trials of depression representative of patients in routine clinical practice? Am J Psychiatry. 2002;159:469-473), demonstrating that most patients seeking clinical treatment of depression would not qualify for an RCT based on common exclusion criteria.

Methods: Eight hundred seventeen patients presenting to an outpatient private practice were evaluated by retrospective chart review. The 11 exclusion criteria outlined in the previous study were applied to a sample of 348 depressed adults to determine the percentage that would have qualified for an RCT.

Results: We have closely replicated the study of Zimmerman et al, finding that 91% of our sample would not qualify for an RCT based on presence of any of the 11 exclusion criteria. Prevalence of 7 criteria applied were found significantly different in our population when compared with the study of Zimmerman et al, yet exclusion rates came within 0.2%. When only the 5 most common criteria were applied, exclusion rates remained high.

Conclusions: Some exclusion criteria are essential for ethical or diagnostic purposes or to reduce heterogeneity; others are somewhat arbitrary, widening the gap between research and clinical practice. Ninety-one percent of patients presenting for treatment of depression would not qualify for RCTs if the 11 exclusion criteria identified were applied; if a standard severity cutoff score and 4 criteria considered relevant to safety or diagnostic validity were used, nearly 75% would not qualify. Use of antidepressants in a typical clinical population is an extrapolation from research data.