Do we Know Enough Yet? A Cumulative Cost-Effectiveness Analysis of Mechanical Thrombectomy

OBJECTIVES: Cumulative meta-analysis is often used to investigate the progression of an evidence-base over time. When trials contribute numerous parameters to a cost-effectiveness model, subtle changes can have implications for the estimate of cost-effectiveness. This study incorporated cumulative meta-analysis results into an economic model of mechanical thrombectomy to investigate the impact on cost-effectiveness of a rapidly expanding evidence base. METHODS: Data from nine published RCTs were sequentially added to inform parameter values regarding treatment efficacy and device usage in an economic model estimating estimated the cost-effectiveness of adding mechanical thrombectomy to established standard-of-care. The incremental cost-effectiveness ratio (ICER) and probability of cost-effectiveness relative to the willingness-to-pay threshold (WTP) of e20,000 per QALY were analysed. RESULTS: On the basis of the first published trial, the intervention was not cost-effective at e 20,000 per QALY. After the second trial published, the ICER was 613,077 per QALY. The ICER remained below 6 20,000 per QALY until the publication of the seventh trial, after which the ICER was above the WTP threshold. For an 18 month period between the publication of the second and seventh trials, the decision would have been clear relative to the WTP threshold. For the 10 months since the publication of the seventh trial, the decision is less clear. CONCLUSIONS: Assuming a stable estimate of cost-effectiveness due to an apparently stable estimate of clinical efficacy may not be appropriate, particularly where trials contribute multiple parameters to a model. Trials may be powered to detect effect in a primary endpoint, but not in secondary endpoints that may be quite influential in an economic model. Cost-effectiveness analysis also presents an estimate at a point in time - appreciating the evolving evidence base is important to understand the extent to which the cost-effectiveness may change with additional data from ongoing trials.