Latent class analysis of emergency department patients engaged in telehealth peer recovery support services and associations of identified classes with post-discharge outcomes

Abstract

Background: People with substance use disorders (SUDs) frequently use emergency department (ED) services. Despite evidence demonstrating that post-discharge SUD treatment linkage effectively reduces the number of ED re-presentations, relatively few hospitals have implemented interventions to identify and connect patients with SUDs to appropriate care. ED-based peer recovery support specialist (PRSS) interventions have emerged as a promising approach for hospitals, but more research is needed to understand the extent to which these interventions meet the needs of patients who present to the ED for different reasons and with various underlying concerns.

Method: A retrospective cohort analysis used data from a telehealth PRSS program in 15 EDs within one Indiana hospital system. The study included 2950 ED patients who engaged with telehealth PRSS services between September 2018 and September 2021. Latent class analysis identified patterns of patient characteristics associated with post-discharge PRSS engagement and ED re-presentations. Covariate predictors and distal outcomes were assessed to examine the associations between class membership, demographic factors, and patient outcomes.

Results: The study team selected a six-class model as the best fit for the data. Class 1, representing patients with opioid use disorder and mental health diagnoses who presented to the ED for an opioid overdose, was used as the reference class for all other statistical tests. Multinomial logistic regression analysis demonstrated significant associations between covariate predictors, outcomes, and class membership. Regression results also demonstrate PRSSs had greater success contacting patients with prior year ED use and patients with a successful post-discharge PRSS contact were less likely to re-present to the ED.

Conclusion: Results highlight the heterogeneity of patients with SUDs and emphasize the need for tailored interventions to address patient-specific needs more effectively. They also provide support for the perceived utility of PRSS engagement for ED patients.

Keywords: Emergency department; Latent class analysis; Opioid use disorder; Peer support services; Substance use disorder; Telehealth.