Strategies for the Management of Suspected Heparin-Induced Thrombocytopenia: A Cost-Effectiveness Analysis

Background: Heparin-induced thrombocytopenia (HIT) is a rare but serious complication of heparin therapy. The diagnosis of HIT is difficult because its signs are non-specific and the heparin-platelet factor 4 (PF4) antibody test used to confirm the diagnosis is imprecise. Drugs used to treat HIT are costly and may carry an increased risk of bleeding. Objective: To evaluate the cost effectiveness, from a societal perspective, of four treatment approaches for patients with suspected HIT within a US critical care setting. Methods: A decision-tree was constructed for the management of a hypothetical cohort of critical care patients with possible evidence of HIT. The management strategies were: (i) no antibody testing, continue heparin (No Test and Wait); (ii) antibody testing, continue heparin while test results pending (Test and Wait); (iii) antibody testing and switch to a direct thrombin inhibitor (DTI) while test results pending (Test and Switch); and (iv) no antibody testing but switch to a DTI (No Test and Switch). We used argatroban as the DTI in our analysis. The outcomes were direct medical costs ($US; 2004 values), QALYs and incremental cost-effectiveness ratios (ICERs). Results: Assuming an HIT prevalence of 3%, relative to less costly strategies, ordering an antibody test and switching patients to argatroban if the result was positive (Test and Wait) had an ICER of $US163_396/QALY. Pre-emptive switching to argatroban without antibody testing (No Test and Switch) was the most effective strategy but had an ICER of >$US1 million/QALY relative to the Test and Switch option. These results were highly sensitive to HIT prevalence among patients presenting with thrombocytopenia. Assuming a willingness to pay of $US50_000 per QALY, the Test and Wait strategy became cost effective when the prior probability of HIT was 8%. At a prior probability of 12%, Test and Switch was cost effective, and at probabilities of HIT in the 60-75% range, No Test and Switch was cost effective. In two-way analysis, the probability of developing a thrombotic event was a key driver of treatment choice at specific HIT probabilities. Conclusions: Testing for HIT in all typical critical care patients with thrombocytopenia is unlikely to represent a cost-effective management strategy. With increasing probability of HIT, strategies that include testing and a more rapid switch to a DTI appear more desirable. Accurate clinical judgment of the prior probability of HIT has a critical influence on the cost-effective management of HIT.