Blood loss in total knee arthroplasty.

Over a two-year period, 112 consecutive primary total knee arthroplasties (TKA) were performed according to a standard protocol and were evaluated to determine the effects of intraoperative tourniquet deflation, suction drainage, and immediate continuous passive motion (CPM) on blood loss and the need for blood transfusions. It has been reported that these measures lead to increased morbidity following TKA. The tourniquet was deflated before closure, hemostasis was obtained, and two suction drains were placed for 24 to 48 hours. CPM was started immediately, and continued for a mean of seven days. Serial hematocrits (Hct) were obtained. Mean age of the 64 males and 48 females was 65 years. The diagnosis was osteoarthritis in 86% and rheumatoid arthritis in 14%. Seventy-three percent of the prostheses were cemented and 27% were uncemented. The mean tourniquet time was 104 minutes and the mean length of surgery was 140 minutes. Blood loss measured at surgery averaged 222 ml. Total calculated blood loss averaged 794 ml, but was higher for patients with a cemented versus uncemented prosthesis (p less than .05), and males versus females (p less than .005). It was not related to diagnosis, tourniquet time, or length of surgery. Of the 38% of patients requiring a transfusion (mean, 2.2 units), there was not a significantly greater drop in Hct (from 36.6 to 26.5; change, 10.1) when compared to patients not receiving a transfusion (from 43.1 to 32.5; change, 10.6). Therefore, the need for transfusion was related to preoperative Hct and not the intraoperative or postoperative blood loss.(ABSTRACT TRUNCATED AT 250 WORDS)