A prediction rule for the use of postdischarge medical services

To develop and validate a prediction rule screening instrument, easily incorporated into the routine hospital admission assessment, that could facilitate discharge planning by identifying patients at the time of admission who are most likely to need postdischarge medical services.

Prospective cohort study with separate phases for prediction rule development and validation.

Urban teaching hospital.

General medical service patients, 381 in the derivation phase and 323 in the validation phase, who provided self-reported medical history, health status, and demographic data as a part of their admission nursing assessment, and were subsequently discharged alive.

Use of postdischarge medical services such as visiting nurse or physical therapy, medical equipment, or placement in a rehabilitation or long-term care facility was determined. A prediction rule based on a patient's age and Medical Outcomes Study 36-Item Short Form Health Survey (SF-36) physical function and social function scores stratified patients with regard to their risk of using postdischarge medical services. In the validation set, the rate of actual postdischarge medical service use was 15% (15 of 97), 36% (39 of 107), and 58% (57 of 98) among patients characterized by the prediction rule as being at "low", "intermediate," and "high" risk of using postdischarge medical services, respectively.

This prediction rule stratified general medical patients with regard to their likelihood of needing discharge planning to arrange for postdischarge medical services. Further research is necessary to determine whether prospective identification of patients likely to need discharge planning will make the hospital discharge planning process more efficient.