Estimates of Workload Associated With Suicide Risk Alerts After Implementation of Risk-Prediction Model

Andrea H Kline-Simon,Monique Does,Stacy Sterling,Vincent Liu,Gregory Simon,Yun Lu,Kelly Young-Wolff

doi:10.1001/jamanetworkopen.2020.21189

Abstract

This cohort study attempts to validate the Mental Health Research Network suicide risk–prediction model and estimate associated workloads.

Highlights

From 1999 to 2017, the US suicide rate increased by 35% across age, sex, and geographic groups.1 Suicide risk–prediction models using data from electronic health records provide a promising approach for identifying and assisting individuals at risk.2 The Mental Health Research Network (MHRN) developed highly discriminative suicide risk–prediction models using data from 20 million mental health care visits across 7 health systems.3,4 the clinical and operational requirements for implementing the models in practice are unknown
Over 1 year, we identified 1 408 683 mental health encounters (254 779 unique patients with mean [SD] age 40.7 [18.7] years, including 89 857 men [35.3%], 63 110 individuals who were Hispanic or Black [24.8%], and 35 267 individuals who had Medicare coverage in the previous year [13.8%])
Model discrimination was comparable to that found by Simon et al3 using the original MHRN sample

Summary

Introduction

From 1999 to 2017, the US suicide rate increased by 35% across age, sex, and geographic groups. Suicide risk–prediction models using data from electronic health records provide a promising approach for identifying and assisting individuals at risk. The Mental Health Research Network (MHRN) developed highly discriminative suicide risk–prediction models using data from 20 million mental health care visits across 7 health systems. the clinical and operational requirements for implementing the models in practice are unknown. From 1999 to 2017, the US suicide rate increased by 35% across age, sex, and geographic groups.. Suicide risk–prediction models using data from electronic health records provide a promising approach for identifying and assisting individuals at risk.. The Mental Health Research Network (MHRN) developed highly discriminative suicide risk–prediction models using data from 20 million mental health care visits across 7 health systems.. The clinical and operational requirements for implementing the models in practice are unknown. We sought to externally validate the MHRN risk model and provide clinical workload estimates for implementation. Author affiliations and article information are listed at the end of this article

Methods

Results

Conclusion