We evaluated the results of twelve hematological and plasma protein determinations in 450 to 500-milliliter volumes of shed blood that had been collected with or without acid-citrate-dextrose anticoagulant (National Institutes of Health Formula A) from knees and hips during the first twelve hours after arthroplasty. We also evaluated the effects on the recipients when the blood was used for reinfusion. The findings in the units that had been obtained in less than four hours, in between four and six hours, and in more than six hours after the arthroplasty were similar whether or not the acid-citrate-dextrose anticoagulant had been used. The mean values for the collected units were: in the blood, a concentration of hemoglobin of 115 grams per liter, a hematocrit of 0.34, a white blood-cell count of 4.8 x 10(9) per liter, and a red blood-cell count of 3.7 x 10(12) per liter, and, in the plasma, a level of hemoglobin of 160 grams per liter, a level of fibrinogen of less than 0.2 gram per liter, a level of factor-V clotting protein of less than 10 per cent of normal, a level of factor-VIII clotting protein that was 45 per cent of normal, a level of antithrombin III that was 45 per cent of normal, a level of plasminogen that was 55 per cent of normal, a level of protein C that was 100 per cent of normal, and a level of fibrin-degradation products of 1000 micrograms per milliliter of plasma. The clinical response of the patient was assessed after the reinfusion of a total of 205 units of unwashed shed blood into 153 patients. In addition, in 126 of the 153 patients, hematological and plasma-protein measurements were analyzed before the autotransfusion and one and twenty-four hours afterward. Each of these patients had received one to four units of shed blood that had been filtered but not washed. Only two (2 per cent) of the ninety-nine patients who received shed blood that had been collected six hours or less after the operation had a febrile reaction, whereas twelve (22 per cent) of the fifty-four patients who received blood that had been collected six to twelve hours after the operation had such a reaction. Factor-VIII levels were significantly lower in the first and second units of shed blood that had been collected after the procedures on the knee compared with the levels in the first and second units of blood that had been collected after the procedures on the hip. We have no explanation for this finding. In the shed blood that had been collected from the hips and knees with the acid-citrate-dextrose anticoagulant, the levels of factor VIII and protein C were significantly lower than those in the shed blood without the anticoagulant. There were no other significant differences in the measurements that had been made on the shed blood. Shed blood that had been reinfused through a forty-micrometer screen filter produced no clinical evidence of bleeding or abnormalities in clotting proteins or platelet counts, whether or not the anticoagulant had been used. No coagulopathy developed after reinfusion of an average of 1.3 units of autologous non-anticoagulated unwashed filtered shed blood that had been collected after knee or hip arthroplasties. The only adverse reaction to infusion of autologous red blood cells from this source was a febrile episode, which resolved rapidly.
