Bleeding complications with enoxaparin for deep venous thrombosis prophylaxis.

The safest and most efficacious method of deep venous thrombosis prophylaxis remains controversial. With the use of enoxaparin, a low-molecular-weight heparin, becoming ubiquitous in many institutions, we specifically examined bleeding complications related to its use. A case-control study was conducted on consecutive patients receiving enoxaparin prophylaxis after primary or revision total knee or total hip arthroplasty or hip hemiarthroplasty. Matched controls receiving no pharmacologic anticoagulation were identified. Patient and operative characteristics, hematologic values, and timing of enoxaparin dosing were analyzed as related to major and minor bleeding complications. A total of 152 procedures with enoxaparin and an equal number of control cases were included for a total of 304 patients. The enoxaparin group had a 23.7% total complication rate compared to 16.5% for the control group (P = .11). The power of the test was .35 and indicated that approximately 970 patients would need to be reviewed to have at least an 80% chance of finding a statistically significant difference. Major complications occurred in 5 patients (3.3%) in the enoxaparin group and 2 (1.3%) in the control group (P = .25, power = .21). Minor complications in the enoxaparin group were slightly higher but not significant at 20.4% versus 15.1% in the control group (P = .23). There were significantly fewer minor complications in the enoxaparin group after primary single-joint procedures (16.50%) than all other procedures (32.4%). Patients receiving the first dose of enoxaparin 10 hours or more postoperatively had significantly fewer complications (P = .05). The postoperative hematocrit drop was significantly greater for the enoxaparin group for all procedures (P = .003) as well as for primary single procedures (P = .0005). The postoperative transfusion requirement was significantly greater after primary single procedures (P = .02) in the enoxaparin group. One patient with an epidural catheter and receiving enoxaparin postoperatively developed an epidural hematoma. Although there were not significantly more complications with enoxaparin, there was evidence of significantly increased postoperative bleeding. The low-power analysis reveals that a large number of patients (970-1,700) are required to show a statistically significant difference in bleeding complications between the 2 groups. To minimize complications, a short period to allow initial hemiostasis postoperatively is recommended, as is preferential use of enoxaparin for primary single-joint replacements. Enoxaparin used in conjunction with an indwelling epidural catheter is not recommended.