USING VALUE OF INFORMATION ANALYSIS TO ESTABLISH THE POTENTIAL BENEFIT OF FURTHER COMPARATIVE EFFECTIVENESS RESEARCH: THE CASE OF CLOPIDOGREL IN THE UK

Purpose: Patients with Non‑ST‑elevation- acute coronary syndrome (NSTE-ACS) are at increased mortality risk due to ischaemic vascular events. Based on a clinical trial with a duration of one year, early treatment with clopidogrel combined with aspirin has been established to be cost-effective by an economic model. Yet it is unclear whether shorter durations of clopidogrel are more cost-effective in selected risk groups and whether, therefore, a study of comparative effectiveness for shorter durations is warranted. Aim of this study is to formally assess the potential value and feasibility of this clinical trial for the optimal duration of clopidogrel treatment.
Methods: The study is based on an update of a previous economic evaluation and a value of information (VOI) analysis for the NHS R&D HTA Programme. Consideration was also given to the potential impact that the introduction of a generic version of clopidogrel may have on the VOI results.

Results: A policy of 12-months clopidogrel for patients with NSTE-ACS appears cost-effective in both in average patients and in the subgroup of higher-risk patients, compared to shorter term durations. However, for lower-risk patients treatment with clopidogrel beyond 3-months does not appear to be cost-effective as long as clopidogrel is on patent. At a generic price of 25% of its current price, one year of clopidogrel gets cost-effective across all patient groups.
Estimates of EVPI were markedly higher for the combined analysis of all patients and for analysis of high-risk patients alone, compared to those for lower risk patients. In the lower risk group, total EVPI ranged from £7.91 million to £20.38 million at a threshold of £30,000 per QALY. After the entry of generic clopidogrel when the trial is likely to be able to report, EVPI ranged from £10.8 million to £11.9 million. The partial EVPI calculations demonstrated that approximately 40-45% of this value related to the treatment effectiveness parameters for clopidogrel (i.e. those for which an RCT would be required).

Conclusions: In lower risk groups, for which shorter durations of clopidogrel appear more cost-effective, it seems unlikely that an adequately powered RCT would be considered to provide value for money due to the significant cost that would be required to undertake such a study and the cost of the uncertainty that such a trial might resolve.