Rehospitalization risk factors for psychiatric treatment among elderly Medicaid beneficiaries following hospitalization for a physical health condition

Abstract

Objectives: Hospitalizations for physical health problems can cause great mental health challenges. We examined risk factors and expenditures for early readmission for inpatient psychiatric treatment following a hospitalization for a non-psychiatric condition.

Method: We used eight years of Florida Medicaid enrollment and claims data for (FY 2003–2011) as well as data obtained from the Florida Center for Health Information and Policy Analysis, Inpatient Hospital Database, to identify episodes of hospitalization for all Medicaid-enrolled older adults. There were a total of 781,660 index hospitalizations for non-behavioral health-related hospitalizations, 2690 (0.3%) of which resulted in a behavioral health-related rehospitalization within 30 days after discharge. We estimated the relative risk of early rehospitalization using Cox proportional hazards regression.

Results: Participants (n = 276,254) were 76.8 ± 8.3 years old; 68.4% female and 43.5% White. As expected, we found that having a preexisting behavioral health disorder (including severe mental illness, substance use disorder, or dementia) greatly increased the risk of readmission. We also found that male gender increased the risk of early readmission. Older age, being Black and having one or more comorbid medical conditions as identified by the Charlson Index were associated with substantially reduced risk.

Conclusion: Identifying hospitalized patients with psychiatric conditions, as well as those who are relatively young within this age group, male, White rather than Black or of other race, alcohol abusers, and those without comorbid conditions, and providing them with care to address their behavioral needs may help in efforts to reduce early rehospitalizations for psychiatric conditions.

Keywords:

• Medicaid
• early readmission
• elderly
• psychiatric treatment

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Florida Agency for Health Care Administration, contract MED134: Medicaid Behavioral Health Studies.