Relationship of Mental Illness and Ambulatory Care Sensitive Condition (ACSC)-Related Hospitalizations among Veterans seen in VHA Primary Care

Trivedi RB, Sylling PW, Wong ES, Liu C, Hebert PL, Kivlahan DR, Post EP, Pomerantz AS, Fihn SD, Nelson KM. Relationship of Mental Illness and Ambulatory Care Sensitive Condition (ACSC)-Related Hospitalizations among Veterans seen in VHA Primary Care. Poster session presented at: AcademyHealth Annual Research Meeting; 2014 Jun 9; San Diego, CA.

Related HSR&D Project(s)

• CDA 13-024 – Economic Impact of Dual Use and Patient Choice in Primary Care
• CDA 09-206 – Developing a Couples Coping Intervention for Veterans with Heart Failure

Veterans with a mental health condition have higher hospitalization rates for ambulatory care sensitive conditions (ACSCs) compared to those without a mental illness. The Veterans Health Administration (VHA) implemented its patient centered medical home, Patient Aligned Care Teams (PACT), in April 2010. Our objective was to assess the association of PACT on ACSC-related hospitalization rates among Veterans with depression and post-traumatic stress disorder (PTSD). Study Design: An interrupted time series model examined the association between PACT and ACSCs using VHA administrative data aggregated to the facility-mental health cohort-quarter level. Initial and medium-term changes in ACSCs were measured using PACT indicators corresponding to 4/1/2010 (2010-Quarter 2(Q2)) and 4/1/2011 (2011-Q2). Models were adjusted for seasonality, secular hospitalization trends, patient age, sex, health risk, facility size, and facility area economic climate. Random effects allowed intercept and time-trend parameters to differ by facility and mental health cohort. We defined ACSC rate as the total number of ACSC-related hospitalizations among Veterans with each diagnosis assigned to a clinic, divided by the number of clinic patients with that diagnosis. We compared ACSC rates between Veterans with and without mental illness across our sample period and also compared changes in ACSC hospitalizations after PACT, relative to existing pre-PACT trends, among Veterans with PTSD or depression to Veterans at each clinic without that diagnosis. Due to significant utilization of non-VHA care for Veterans 65 and older, we estimated separate models for Veterans under 65 years and those 65 years and older. Models for PTSD and depression cohorts were performed separately.