The primary aim of this study is to reduce the incidence of nonelective readmissions to the medical service at Wishard Hospital by increasing nursing and social support for needs identified on the patients' first admission during the study period. Nonelective readmissions are those for urgent or emergent conditions as opposed to those elective readmissions which are scheduled for specific procedures. Secondary aims include: (1) improving the efficiency of utilization of medical services by increasing return visits to the clinic and reducing emergency visits; (2) improving the functional status and risk factors for readmission (defined from previous studies at Wishard); (3) demonstrating the cost-effectiveness of the intervention; and (4) refining our predictive model for readmission by considering the functional status of patients at discharge as well as their medical characteristics. All medicine inpatients fulfilling criteria will be interviewed using a previously validated instrument (Long Term Care Information System or LTCIS) to identify social service, nursing and medical problems patients have at discharge. Patients will be randomly assigned to two groups. Those in the control group will receive standard care. For those in the invervention group, the LTCIS will be available to physicians as part of their outpatient charts in the general medicine clinic. In addition, the medicine clinic social worker and/or nurse practioners will aggressively follow these patients by telephone and subsequent clinic visits. Also, the social worker and nurse practioners will care for nursing or social service problems identified. We wish to show that a system of formally assessing patients' status at discharge, making this information available to their outpatient providers, providing a mechanism by which inpatient and outpatient services are more continuous, and having outpatient providers document the status of problems and care will reduce medical readmission, improve patients' functional status, and be cost effective.
