Closed wound drainage in total hip or total knee replacement. A prospective, randomized study.

Abstract
We prospectively randomized 415 total joint replacements for either a closed wound-drainage system or no postoperative drainage. Drainage was not used in 200 total joint replacements, of which 138 were total knee replacements and sixty-two, total hip replacements. Drainage was used in 215 total joint replacements, of which 137 were total knee replacements and seventy-eight, total hip replacements. All patients were evaluated for the presence of excessive postoperative drainage that necessitated cessation of the range-of-motion exercises, the amount of transfused blood (homologous and autologous), and the preoperative and postoperative hemoglobin levels. The range of motion was assessed daily in the patients who had a total knee replacement. No statistical difference was found in the number of patients who had excessive postoperative drainage from a drained or non-drained wound. There was also no statistical difference with respect to the amount of transfused blood and the preoperative and postoperative hemoglobin levels. Furthermore, in the patients who had a total knee replacement, there were no statistical differences between drained and non-drained wounds with respect to the daily range of motion during the first seven days postoperatively. The mean amount of blood transfused was 157 milliliters in the total knee replacements with drains, 160 milliliters in the total knee replacements without drains, 188 milliliters in the total hip replacements with drains, and ninety-three milliliters in the total hip replacements without drains. Consequently, we believe that postoperative drainage systems offer little advantage in the outcome of primary total knee or total hip replacements.A savings of $21,500 (215 drainage units at $100 per unit) would have resulted if drains had not been used at all in this series. Our series was large; however, the findings should be considered preliminary.

This publication has 7 references indexed in Scilit: