Patient predictors and utilization of health services within a medical home for homeless persons

Background: The Veterans Health Administration (VHA) established a patient-centered medical home model of care for Veterans experiencing homelessness called Homeless Patient Aligned Care Teams (HPACTs) to improve engagement with primary care and reduce utilization of hospital-based services. To evaluate the impact of HPACT, we compare the number and type of health care visits in the twelve months before and after enrollment in one HPACT, and explore patient characteristics associated with increases and decreases in visits.

Methods: We conducted a chart review of VHA medical records for all patients enrolled in an HPACT in Pittsburgh, PA between May 2012 and December 2013 (n = 179). Multivariable mixed effect logistic regressions estimated differences in having any visit in the 0-6 months and 7-12 months before and after HPACT enrollment, and multinomial logistic regressions predicted increases or decreases versus no change in number of visits over 12 months.

Results: Compared to 0-6 months prior to HPACT, patients were more likely to visit primary care in the 0-6 months (OR = 4.91, CI = 2.94-8.20) and 7-12 months (OR = 2.30, CI = 1.42-3.72) following HPACT. Patients were less likely to visit the emergency department or to be hospitalized in the 0-6 months (OR = 0.57, CI = 0.34-0.94; and OR = 0.55, CI = 0.25-0.76) and 7-12 months following HPACT (OR = 0.43, CI = 0.33-0.91; and OR = 0.45, CI = 0.26-0.80). Patients were less likely to visit mental health (OR = 0.35, CI = 0.20-0.60) and addiction specialists (OR = 0.39, CI = 0.18-0.84) in the 7-12 months following HPACT. Overall, 59% of patients had increases in primary care visits following HPACT. Female patients and those with self-housing were less likely to have increases versus no change in primary care visits (RRR = 0.15, CI = 0.03-0.74; and RRR = 0.35, CI = 0.14-0.90).

Conclusions: An integrated HPACT model was successful in engaging homeless Veterans in primary care for one year, potentially contributing to reductions in ED use. More tailored approaches may be needed for vulnerable populations experiencing homelessness, including homeless women.