To The Editor:
We are writing in regard to the article "Thrombosis Prevention After Total Hip Arthroplasty: A Prospective, Randomized Trial Comparing a Mobile Compression Device with Low-Molecular-Weight Heparin" (2010;92:527-35), by Colwell et al. Major bleeding is the most important safety end point in studies on drugs for deep venous thrombosis prophylaxis, but an adequate definition of major bleeding in surgical studies is still an issue of debate1. Colwell et al. reported that, compared with low-molecular-weight heparin, the use of a mobile compression device for deep venous thrombosis prophylaxis following total hip arthroplasty resulted in a significant decrease in major bleeding events. However, this conclusion may be misleading. The definition of major bleeding in this study included bleeding that required rehospitalization or prolonged hospitalization, required any intervention such as surgery or hematoma aspiration to prevent permanent impairment or damage, endangered critical organs (intracerebral, intraocular, intraspinal, pericardial, or retroperitoneal), was life threatening, or caused death.
Eleven episodes of major bleeding were assigned to the enoxaparin group; however, the classification criteria of major bleeding in these patients was prolonged hospitalization in six and rehospitalization in three, and none needed treatment for bleeding. No major bleeding events were assigned to the compression device group. This difference in the rate of major bleeding was inconsistent with the reported total number of blood units used, number of transfusions, mean number of units transfused per group, mean number of units transfused per patient, mean bleeding index, number of patients with a bleeding index of =2, number of patients with a change of =20 g/L, and mean change in hemoglobin. No episode of bleeding that endangered critical organs, was life threatening, or caused death was seen in either group.
Although most studies with an increase in the rate of major bleeding also show an increase in the rate of minor bleeding, in the present study the rate of minor bleeding was similar in both groups (37% and 42% in the compression device and enoxaparin groups, respectively; p = 0.319). The authors also reported that the rate of major bleeding found in the enoxaparin group is similar to the 5.3% rate reported after the pooling of the results of five studies that utilized low-molecular-weight heparin after total hip arthroplasty2. However, these five studies used different definitions of major bleeding and were done in the 1990s.
Recent pooled analyses of studies of total hip arthroplasty with new oral antithrombotic drugs, which used enoxaparin for comparison, review boards, and objective criteria of major bleeding, showed major bleeding rates of <1.5%3,4. Moreover, previous studies comparing mechanical compression devices with enoxaparin in hip replacement found no significant difference in the rate of major bleeding5,6. Furthermore, the reported rate of major bleeding in the placebo group in previous studies on the use of enoxaparin after total hip arthroplasty was 4%, showing that patients who did not receive any antithrombotic drug also can bleed because of the surgical procedure itself7,8.
We believe that the elevated rate of major bleeding in the enoxaparin group reported by Colwell et al. is overestimated and could be due to the open-label nature of the study; the absence of independent, blinded evaluators; and the use of subjective qualifying criteria to define major bleeding.