Computerized Adaptive Suicidal Risk Stratification and Prediction

Project Summary The software to be created with this SBIR, the Computerized Adaptive Test for Suicide Scale – Expanded (CAT-SSE), will help healthcare systems to identify and monitor suicide risk in a more robust and valid way than is currently done. In the final stage of Phase 2, we proposed to integrate the validated CAT-SSE with Epic, the EHR used at the University of Massachusetts Memorial Healthcare (UMass) system. Notably, this two-way, deep integration will allow us to import the MHRN indicators and use them in the CAT-SSE risk stratification equations and deliver the results and clinical decision support in the EHR for clinicians to access and act on at the point of care. Through the first administrative supplement (R44MH118780-01A1), our aims with the Epic integration were to (1) build the de-identification model (see Figure), and (2) build the tools to facilitate translation of this approach at other sites. We have since deployed a multi-tier architecture for building the CAT-SSE into Epic such that patient identifiers are not shared with ATT. We were able to successfully transfer data between ATT and the UMass Epic servers within an Epic test environment. In this Phase 2 administrative supplement, we are requesting funds for UMass IT to complete the refinement of the existing integration of the CAT-SSE into their Epic EHR instance. The main outcomes of the integration refinement will be: (1) Development of elements that ensure clinical usability and interpretability are maximized. These elements include but are not limited to data visualization, guidance language, and Epic alerts. (2) Development of remote and in-person CAT-SSE delivery options for CAT-SSE participants. (3) Successful real-time import of CAT-SSE scores from ATT that leverage MHRN data to calculate a suicide risk prediction. It is estimated that a total of 370 hours of IT work are required to complete the above outcomes. Unfortunately, Due to COVID-19, the project encountered major enrollment delays, which forced the study team to revise the initial enrollment targets and divert all available funds to maximize subject enrollments. In addition, UMass IT department changed its priorities to advance COVID-related patient care projects (e.g. telehealth) over research projects. Because of the increased IT labor post-pandemic, and utilizing all available resources to recover the enrollment loss during COVID peaks, we will need additional funds to complete this important part of the project. Details on how these funds will be used are summarized in the budget justification.