Analytical approach:
A decision analytic model was developed in order to determine the cost-effectiveness of the two prophylactic therapies. The model was populated with published evidence. The time horizon adopted in the analysis was not stated. The authors stated that two different perspectives, that of the third-party payer and that of the institution, were used in the analysis.
Effectiveness data:
The clinical data were derived from published studies. Data on clinical effectiveness (reduction of VTE) from one key clinical trial showed similar efficacy for heparin and enoxaparin , thus the authors assumed equal efficacy for the two treatments under study. Data on HIT and HITT were identified from a review of the literature (MEDLINE), details of which were reported. The authors selected the most appropriate estimates from among those available, using median values when several estimates were available. Life expectancy data were obtained from the National Center for Health Statistics. The key clinical data were the risk of HIT and HITT with the two alternative drugs.
Monetary benefit and utility valuations:
Utility estimates were obtained from a registry at the Harvard Center for Risk Analysis. Details of these estimates were not given.
Measure of benefit:
The summary benefit measure was the quality-adjusted life-years (QALYs). These were derived using the decision model. The use of discounting was not reported.
Cost data:
The categories of costs included in the analysis were visits to primary or secondary care physicians, drugs, laboratory tests and hospital stay. The reimbursement to institution and physicians as a result of HIT and HITT was considered from the payer perspective (Medicare), while the cost of medications and the potential loss of income for the hospital from additional days of hospitalisation for HIT and HITT were included in the institutional analysis. Direct medical costs were based on Medicare reimbursement rates and were presented as macro-categories. Potential charges associated with the development of HIT or HITT were considered and added to usual reimbursement rates. Costs were in US dollars ($) and the prices referred to 2004 and 2005. The use of discounting was not reported.
Analysis of uncertainty:
A deterministic sensitivity analysis was undertaken in order to ensure the validity of results over a range of estimates used in the model, such as rates of HIT, rates of thrombosis, mortality, reimbursement rates and medication costs. Alternative values were based on authors’ opinions as well as published data.