Chih, M.Y., Patton, T., McTavish, F.M., Isham, A.J., Judkins-Fisher, C.L., Atwood, A.K., & Gustafson, D.H. (2014). Predictive modeling of addiction lapses in a mobile health application. Journal of Substance Abuse Treatment, 46(1): 29-35. PMCID: 3963148.
The Addiction- Comprehensive Health Enhancement Support System (A-CHESS) smartphone application is designed to provide continuing care for patients with alcohol use disorders. To provide immediate and targeted help to patients in risk of relapsing, data from a recent study of A-CHESS was used to create a predictive model of relapse. Study participants enrolled in A-CHESS completed the Weekly Check-In, a survey used to monitor patient’s recovery and substance use. From this data, researchers created a statistical model that accurately predicted risk for relapsing in the upcoming week. This model was programmed into the A-CHESS intervention and used to notify patients and clinicians when the risk of relapse was high.