The Problem
Preliminary data suggested that emergency department (ED) length of stay (LOS) is longest for youth waiting for transfer to inpatient psychiatric units. Therefore, we developed a predictive model to identify individuals at risk for admission to an ED or inpatient psychiatric unit using Medicaid physical and behavioral healthcare claims from Alliance Behavioral Healthcare (ABH).
Our Solution
We developed a predictive model using Medicaid health insurance claims to identify individuals at risk for admission to EDs or inpatient psychiatric units. For youth identified by the model, we intervened with care management to improve engagement in community-based behavioral health services, thus potentially reducing use ofED and inpatient psychiatric services.
Impact
In this project, 65% of cases that were reviewed by the DUHS-ABH multidisciplinary care management team were successfully engaged with community-based behavioral health services by the end of the project period. Specifically, over 3 months, the multidisciplinary team met for one hour weekly and reviewed clinical and social data for 31 at-risk youth identified by the predictive model. Following these multidisciplinary case reviews and outreach to families, approximately 65% of these at-risk youth were engaged in community-based behavioral health care (defined as having attended at least two visits with a single treatment provider). In some cases, there is continued outreach to families to support engagement in community-based services, thus potentially improving treatment engagement for this cohort.
At the multidisciplinary team meetings, care management objectives were identified for each at-risk youth in review –objectives related to treatment referral, addressing SDOH needs, or other issues. In approximately 75% of cases, the primary care management objective was completed. While there is still room for improvement, this 75% goal completion rate provides evidence that services are available –and interventions capable of being deployed–on behalf of youth identified by predictive modeling to be at risk for ED or inpatient psychiatric admission.
