51. DEMOGRAPHIC AND CLINICAL PREDICTORS OF PARTICIPANT ENGAGEMENT IN SUSTAIN

Abstract

Introduction

This study utilized data from the well-established Supporting Seniors Receiving Treatment And Intervention (SUSTAIN) program. SUSTAIN is sponsored by the Commonwealth of Pennsylvania Department of Aging, and is delivered by the University of Pennsylvania as a free clinical service. Through this partnership, SUSTAIN provides geriatric mental health care management and other behavioral health services by telephone to enrollees in the Pennsylvania Department of Aging’s Pharmaceutical Assistance Contract for the Elderly (PACE) and PACE Needs Enhancement Tier (PACENET). PACE and PACENET, respectively, provide prescription drug coverage for low- and moderately low-income elderly (age 65+) residents of Pennsylvania. Weekly random samples of PACE and PACENET cardholders that have been newly prescribed an antidepressant, anxiolytic, or antipsychotic medication are offered participation in SUSTAIN. SUSTAIN has demonstrated benefits for patients in reducing psychiatric symptom burden and improving quality of life when they engage in services. To support broader implementation of this effective intervention it is important to explore patient engagement in services. This analysis aimed to examine demographic and clinical predictors of patient engagement.

Current literature suggests that geriatric patients who are younger, white, female, and live in an urban area are more likely to engage in mental health services. We examined participant data to determine if the participants who engage in SUSTAIN follow these demographic trends. Clinically, previous studies have demonstrated that geriatric patients with higher symptom burden or previous positive experience with mental health treatment engage more frequently with mental health services. However, these predictors may be less applicable to the older adult population as psychiatric symptoms may be attributed to other medical illnesses and engagement may be influenced by stigma against participating in behavioral health services. We sought to explore if the clinical factors of prior health service utilization and patient identified resource needs were predictors of engagement in treatment.

Methods

This study analyzed a sample of 1071 SUSTAIN participants who were enrolled in the program as indicated by completing a baseline measure between the years of 2018 and 2024. Their level of engagement was denoted by the number of telephone contacts with a behavioral health clinician (nurses and/or licensed clinical social workers) for a care management session. To evaluate demographic predictors of program engagement, participants’ age, gender, race, ethnicity, and geographic area were correlated with the number of care management sessions attended.

To evaluate clinical predictors of program engagement, two sections of the comprehensive baseline assessment were reviewed: service utilization and a needs assessment. To evaluate medical service utilization, participants were asked if within the past three months they received care in any of the following settings: (1) ER or urgent care visit, (2) hospital admission for medical or surgical care, (3) hospital admission for psychiatric care, (4) outpatient visit to a primary, medical, or surgical clinic, (5) outpatient visit to a mental health therapist, or (6) outpatient visit with a psychiatrist. Overall service utilization was defined as the total number of settings in which a participant endorsed receiving care. This was then split up into hospital services (items 1-3) and outpatient services (items 4-6) which were each correlated with the number of care management sessions attended. To evaluate patient needs, participants were provided with a list of 6 resource needs and could endorse a need in any of those categories. The categories were as follows: (1) home safety repairs and modifications, (2) mood or behavioral health services, (3) financial support or legal assistance, (4) community-based resources, (5) general health resources, and (6) other. The number of needs endorsed across all categories was correlated with the number of sessions attended.

Results

The average age of participants at the baseline assessment was 75.88 years old (SD = 6.56 years). Eighty-one percent of the participants were female and 19% were male. The majority of participants self-identified their race as White (92%) and ethnicity as non-Hispanic (97.7%). 35% of participants lived in a rural-designated county while 66% lived in an urban-designated county. There were no demographic characteristics that were significantly associated with number of sessions engaged.

The average number of service settings utilized was 1.60 (SD = 0.87). Participants used an average of 0.46 hospital service settings (SD = 0.72) and 1.14 outpatient service settings (SD = 0.51). Overall medical service utilization was not significantly associated with engagement in sessions, r(1013)= -.01, p = .825. Further analysis revealed that there was a small association between hospital service utilization and engagement in SUSTAIN services, where people who endorsed fewer hospital service setting use engaged in more SUSTAIN sessions of care, r(1013)= -.06, p = .041. There was also a small association between outpatient service utilization and engagement in SUSTAIN services, where people who endorsed using a greater number of outpatient service settings engaged in more SUSTAIN sessions of care, r(1013)= .08, p = .012.

The average number of resource needs endorsed was 1.37 (SD = 1.12). 24.6% of SUSTAIN care management participants endorsed no needs across all six categories, 36.0% endorsed a need in one category, 22.4% endorsed a need in two categories, 12.6% endorsed a need in three categories, 4.1% endorsed a need in four categories, 0.3% endorsed a need in five categories. The number of resource needs endorsed was not significantly associated with engagement in sessions, r(315)= -.02, p = .757.

Conclusions

Clinically, the associations between inpatient and outpatient service utilization to the number of SUSTAIN sessions engaged were as expected. As SUSTAIN is an outpatient service itself, it was predicted that higher utilizers of outpatient care would be more engaged. Yet, the hospital service utilization predictor may be an opportunity to focus outreach attempts on participants who are more clinically appropriate for the service. Though there was no association between patient identified needs and number of sessions attended, there may be an opportunity to explore correlations between engagement and type of need identified in the future.

In this study, we identified clinical predictors of engagement that may be explored further in future projects. Though a number of the trends observed in this data analysis were not statistically significant, they contribute important observations to the under-studied area of behavioral health treatment in older adults.