Prospective validation of a point score system for predicting blood transfusion following hip or knee replacement

BACKGROUND: The purpose of this study was to validate a previously published point score system for predicting the likelihood of a postoperative blood transfusion following hip or knee replacement. STUDY DESIGN AND METHODS: Data were collected prospectively on 460 sequential patients undergoing elective hip and knee replacement at two academic hospitals. Blood transfusion frequency was determined for patients in each of the four risk strata, as defined by the point score system. The accuracy of the system was validated by calculating the area under the receiver operating characteristic (ROC) curve for each site. Data were then combined and inappropriate blood transfusions were eliminated, by using the American College of Physicians guidelines. The frequencies of blood transfusion within each strata were recalculated along with an ROC curve. RESULTS: The point score system accurately predicted the likelihood of Mood transfusion at both hospitals, despite marked differences in overall transfusion frequencies. The calculated areas under the ROC curves were 0.78 and 0.79 for the two sites. The point score system also proved valid when only appropriate blood transfusions were considered, with a calculated area under the ROC curve of 0.74. CONCLUSION: The point score system can accurately predict the likelihood of postoperative blood transfusion following hip or knee replacement. Such a system can be used to target high-risk patients for preoperative autologous blood donation.