I n the pharmacoeconomic evaluation of enoxaparin versus un- fractionated heparin (UFH) for the prevention of venous thrombosis in medical patients, Leykum et al. 1 concluded that lowmolecular-weight heparin (LMWH) offered greater cost utility. This important benefit was based on a single advantage of enoxaparin: a purported lower tendency to cause heparin-induced thrombocytopenia (HIT)/T compared to unfractionated heparin. The authors suggest that this premise is "well supported by published data."
Unfortunately, Leykum and colleagues base their analysis on published information from surgical patients and extrapolate that data to the medically ill. They incorrectly suggest that no studies of HIT frequency in venous thromboembolism (VTE) prevention related to LMWH or UFH in medical patients exist. Girolami et al. prospectively evaluated the incidence of HIT in 306 hospitalized medical patients who received UFH for VTE prevention. 2 HIT developed in 1.4%. A similar incidence of HIT (0.8%) was noted by Prandoni et al. in 376 hospitalized medical patients administered low-dose LMWH for VTE prevention. 3 Expanding the discussion to include patients receiving systemic anticoagulation with either UFH or LMWH reveals even more data suggesting equivalent safety profiles. A randomized study of UFH or LMWH evaluated the frequency of HIT in 1137 patients with symptomatic deep vein thrombosis confirmed by phlebography. 4 Of the 375 patients treated with UFH, 2 developed HIT (0.53%). The same frequency of HIT (0.53%) was observed in 374 patients treated with long-term LMWH (28 days). No patient treated with short-term LMWH developed HIT. Interestingly, Girolami et al. did not identify HIT in 238 patients who were systemically anticoagulated with UFH. 2 Another prospective study, by Prandoni et al., reported that 0.8% of 728 patients developed HIT during systemic anticoagulation with LMWH. 3 The concept of equivalent rates of HIT induction is further supported by 8 other studies reviewed by Locke et al. in their evaluation of this adverse drug reaction in nonsurgical patients (see Table 1). 5 The American College of Chest Physicians consensus statement on antithrombotic therapy queried whether preferential use of LMWH is supported by the literature as a strategy to prevent HIT. 6 They acknowledged that although LMWHs seem to be safer for orthopedic patients due to a lower incidence of HIT, data from nonorthopedic surgery studies are not definitive. As a result, it did not recommend preferential use of LMWH rather than unfractionated heparin in hospitalized medical patients.
In summary, evidence suggests that HIT occurs infrequently in the medical patient population regardless of whether UFH or LMWH is used. Clearly more data need to be developed using consistent methodology before we can definitively characterize the relative risks of HIT associated with UFH and LMWH. In the meantime, Leykum's cost utility analysis should be recalculated using data that are specific to this patient group.