Characteristics of repeat users of an inpatient psychiatry service at a large military tertiary care hospital.

Objective: The purpose of this study was to identify clinical and demographic variables that correlated with readmission to a large tertiary care military psychiatric inpatient service located in the Washington, DC area.

Methods: Data from 983 consecutive inpatient admissions (comprised of 814 individual patients) over a 13-month period (July 1999-July 2000) were abstracted from medical records for retrospective analysis. Repeat users were defined as those individuals receiving two or more inpatient admissions to Walter Reed Army Medical Center during the study period. Statistical comparisons were made between repeat and single admission groups to identify variables predictive of rehospitalization.

Results: Of 814 individual patients under study, 117 (14%) were identified as repeat users. A history of childhood psychiatric problems, previous psychiatric hospitalization, current or past substance abuse, legal problems, and presence of a psychotic or nonbipolar mood disorder were associated with readmission after controlling for active duty status. Current comorbid substance use or personality disorder diagnoses were not predictive. Having at least one child was protective against readmission. The 117 repeat users accounted for 3,838 (37%) of the total 10,393 hospital bed-days during the study period. Logistic regression analysis of all variables correlated with readmission demonstrated statistical significance for past psychiatric hospitalization and age of onset of psychiatric problems before age 18 when active duty status, age, and gender variables were included in the equation.

Conclusions: This study is one of the largest to investigate predictors of rehospitalization in a population that enjoys universal "free" access to comprehensive mental health care services and other benefits of employment. This retrospective analysis documented three important facts: That there seem to be only minor differences between military and civilians with respect to variables associated with psychiatric rehospitalization, that a relatively small proportion of patients accounted for a disproportionately large number of inpatient bed-days, and that an unexpectedly high proportion of active duty patients admitted to this study site reported childhood psychiatric problems and psychiatric hospitalizations before their first hospitalization at Walter Reed Army Medical Center.