Making the Most of it: Application of Planned Missingness Design to Increase the Efficiency of Diagnostic Assessment

Multimethod assessment is recommended as “best practice” in clinical assessment and is often implemented through the combined use of symptom rating scales and structured interviews. While this approach increases confidence in the validity of assessment, it also increases burden, expense, and leads to the accumulation of redundant information. To address this problem, we evaluate the use of a planned missingness design within the framework of adult Attention Deficit/Hyperactivity Disorder (ADHD) assessment. In a sample of 169 young adults, we fit a two-method measurement (TMM) model using ADHD symptoms obtained from rating scales and a structured diagnostic interview. Based on an estimated 8:1 differential between the cost of conducting an in-person diagnostic interview vs. completing questionnaires online, we conducted a series of Monte Carlo simulations to determine the utility of combining TMM with a planned missingness design. We find that even when costs are kept constant, statistical power of the TMM/planned missingness design was equal to the power that would have been obtained had nearly twice the number of participants with complete data been recruited. Conversely, costs could be decreased by 20–25%, while maintaining statistical power equivalent to a design with complete data. Our results suggest the TMM design is a promising technique for reducing the cost and burden of diagnostic assessment within research settings.