Veteran characteristics associated with risk of 30-day inpatient mental health (IMH) readmissions

Abstract: Repeated admissions to psychiatric inpatient services affect patients’ quality of life and may be indicative of the quality of care received during the previous inpatient admissions and/or follow-up outpatient care. Therefore, identifying patient characteristics associated with higher risk of readmission is critical to facilitate appropriate inpatient and follow-up outpatient care to reduce readmissions. In this study, we investigated veteran socio-economic and clinical characteristics associated with the odds of 30-day readmission to Veterans Health Administration Inpatient Mental Health services. Logistic regressions were performed on the total study cohort (N = 88,954), and four sub-groups defined by aggregated principal diagnoses. Our findings suggest that the risk of a 30-day readmission increased for veterans who had a principal diagnosis of schizophrenia or other psychotic disorder, had co-occurring physical and mental health comorbidities, and were at risk of experiencing homelessness. We also found consistently lower readmission rates among non-Hispanic black veterans. Our data is limited to associative relationships, but future work should investigate potential mechanisms including quality of inpatient services and follow-up outpatient care and identity-specific barriers to receiving care.Repeated admissions to psychiatric inpatient services affect patients’ quality of life and may be indicative of the quality of care received during the previous inpatient admissions and/or follow-up outpatient care. Therefore, identifying patient characteristics associated with higher risk of readmission is critical to facilitate appropriate inpatient and follow-up outpatient care to reduce readmissions. In this study, we investigated veteran socio-economic and clinical characteristics associated with the odds of 30-day readmission to Veterans Health Administration Inpatient Mental Health services. Logistic regressions were performed on the total study cohort (N = 88,954), and four sub-groups defined by aggregated principal diagnoses. Our findings suggest that the risk of a 30-day readmission increased for veterans who had a principal diagnosis of schizophrenia or other psychotic disorder, had co-occurring physical and mental health comorbidities, and were at risk of experiencing homelessness. We also found consistently lower readmission rates among non-Hispanic black veterans. Our data is limited to associative relationships, but future work should investigate potential mechanisms including quality of inpatient services and follow-up outpatient care and identity-specific barriers to receiving care.