Contrasting a broad range of outcome and cost measures, the authors analyze the implications for sample sizes and study design using data from prior mental health and primary care studies that span a wide range of practice settings, patient populations and geographic areas.
