Behavioral Health Risk Assessment and Estimation: Validating an Integrated, Multi-Risk Factor Approach aided by Technology

High rates of behavioral health problems in the U.S. require integrated, multi-dimensional approaches. The study of behavioral health risk assessment and estimation aided by technology has the potential to inform assessment and management of behavioral health problems toward the goal of reducing adverse outcomes. The objective of this study is to inform evidence-based behavioral health risk assessment and estimation.
This research examines the U.S. Army Medical Command data within the Behavioral Health Risk Management module (BHRM) to explore behavioral health risk assessment and estimation aided by technology. Analyses are conducted on BHRM data from the records of 30,263 U.S. Army active duty, Guard and Reserve service members assigned to military medical units (U.S. Army Warrior Transition Units) between September 1, 2009 and November 12, 2013. To test risk assessment, responses on the BHRM intake tool (Behavioral Health Risk Assessment-Questionnaire / BHRA-Q) are used to test prevalence, associations, internal reliability and questionnaire’s factor group structure. To examine risk estimation, statistical tests are completed on the prevalence and correlations of risk estimates by the BHRM and clinical providers as well as the predictive properties of demographic variables toward risk estimation. Hypotheses are supported for significant relationships among behavioral health risk variables (r =  .40); good fit of the data to the eight-factor group structure of the BHRA-Q (Comparative Fit Index = 0.969; Tucker-Lewis Fit Index = 0.967; Root Mean Square Error of Approximation = .029 [90% Confidence Interval 0.029 - 0.030]); significant correlations among BHRM and provider risk estimates (large or medium effect size of BHRM on provider estimates); and three significant demographic predictors of risk estimation (race, religion and military service component). Internal reliability of BHRA-Q is supported (Cronbach’s α = .897). This study tests data related to an integrated, multi-risk factor behavioral health risk assessment questionnaire (BHRA-Q) and risk estimation aided by technology (BHRM). Findings support behavioral health risk assessment and estimation using evidence-based / informed multi-risk factor assessment, aided by technology, to inform clinical decision making. Although demographic variables are not strong predictors of risk estimation, as grouped and tested, further study is recommended.