Predicting elderly cardiac patients at risk for readmission.

Elder patients with cardiac disease are at high risk for physical deterioration during post hospital recovery and suffer frequent early readmission. It is important to identify such patients who frequently need help with discharge planning from social workers during their first admission. This study utilized computerized data on 628 patients, 238 of whom were readmissions. Question was raised as to what factors (functional, psychological, social and environmental), differentiated patients who were readmitted from those who were not. Using logistic regression, three variables: marital status, presence of coping difficulty and age of patients were identified as predictors of readmission within three months. Those who were married were less likely to be readmitted. Those with coping difficulties and older individuals were more likely to be readmitted. The accuracy of prediction, using these three factors, was 61 percent. Of those patients predicted as not being readmitted, sixty-nine percent were correctly predicted, while 39 percent were readmitted. Of patients predicted as readmissions, 49 percent were correctly predicted, while 51 percent were not. The major limitation of this study was that key physiological determinants of readmission were not collected. It is imperative that a valid screening device for predicting who is at risk for readmission should include physiological preconditions as well as functional and psychosocial data.