N Kaboré1, G Marnat2, F Rouanet3, X Barreau2, E Verpillot4, P Menegon2, I Maachi5, J Berge2, I Sibon6, A Bénard7. 1. Pôle de santé publique, service d'information médicale, USMR & CIC-EC 14-01, CHU de Bordeaux, 33000 Bordeaux, France; Inserm, Bordeaux Population Health Research Center, team EMOS, UMR 1219, université Bordeaux, 33000 Bordeaux, France. 2. Pôle imagerie médicale, service de radiologie et de neuro-imagerie diagnostique et thérapeutique, CHU de Bordeaux, 33000 Bordeaux, France. 3. Pôle neurosciences cliniques, unité neurovasculaire, CHU de Bordeaux, 33000 Bordeaux, France. 4. Inserm, Bordeaux Population Health Research Center, team EMOS, UMR 1219, université Bordeaux, 33000 Bordeaux, France. 5. Pôle produits de santé, pharmacie clinique dispositifs médicaux, CHU de Bordeaux, 33000 Bordeaux, France. 6. Pôle neurosciences cliniques, unité neurovasculaire, CHU de Bordeaux, 33000 Bordeaux, France; INCIA, CNRS UMR 5287, université de Bordeaux, 33000 Bordeaux, France. 7. Pôle de santé publique, service d'information médicale, USMR & CIC-EC 14-01, CHU de Bordeaux, 33000 Bordeaux, France; Inserm, Bordeaux Population Health Research Center, team EMOS, UMR 1219, université Bordeaux, 33000 Bordeaux, France. Electronic address: antoine.benard@u-bordeaux.fr.
Abstract
BACKGROUND AND PURPOSE: Recent studies demonstrated the benefit of mechanical thrombectomy (MT) plus intravenous tissue-type plasminogen activator (IV-tPA) (MT-IV-tPA) in acute ischemic stroke. This study aimed to estimate the cost-utility of MT-IV-tPA compared with IV-tPA alone from the perspective of the French National Health Insurance. METHODS: We developed a decision tree for the first 3 months after stroke onset and a Markov model until 10 years post-stroke. The health states of the Markov model were according to the modified Rankin Scale (mRS): independent (mRS=0-2), dependent (mRS=3-5), dead (mRS=6). Recurrent stroke was the fourth health stage of our model. We conducted systematic literature reviews and meta-analyses to estimate the cost and utility of each health state, and the transition probabilities between health states. A microcosting study was conducted to estimate the cost of MT. We estimated the incremental cost-effectiveness ratio of MT-IV-tPA and conducted a probabilistic analysis in order to estimate the probability that MT-IV-tPA is cost-effective compared to IV-tPA, the expected value of perfect information (EVPI), and the expected value of partial perfect information (EVPPI), given the uncertainty surrounding the value of our model's parameters. RESULTS: The total mean (standard deviation (SD) cost of MT was €6708.9 (2357.0). The incremental cost-effectiveness ratio (ICER) of the strategy using IV-tPA combined to MT costs was €14,715 per QALY gained as compared to a strategy using IV-tPA alone. The probabilistic analysis showed that the probability of MT-IV-TPA being cost-effective was 85.4% at threshold willingness-to-pay of €30,000 per QALY gained, reaching 98% at €50,000 per QALY gained. CONCLUSION: Although there is no universally accepted willingness-to-pay threshold in France, our analysis suggest that MT combined to IV-tPA can be considered a cost-effective treatment compared with IV-tPA alone.
BACKGROUND AND PURPOSE: Recent studies demonstrated the benefit of mechanical thrombectomy (MT) plus intravenous tissue-type plasminogen activator (IV-tPA) (MT-IV-tPA) in acute ischemic stroke. This study aimed to estimate the cost-utility of MT-IV-tPA compared with IV-tPA alone from the perspective of the French National Health Insurance. METHODS: We developed a decision tree for the first 3 months after stroke onset and a Markov model until 10 years post-stroke. The health states of the Markov model were according to the modified Rankin Scale (mRS): independent (mRS=0-2), dependent (mRS=3-5), dead (mRS=6). Recurrent stroke was the fourth health stage of our model. We conducted systematic literature reviews and meta-analyses to estimate the cost and utility of each health state, and the transition probabilities between health states. A microcosting study was conducted to estimate the cost of MT. We estimated the incremental cost-effectiveness ratio of MT-IV-tPA and conducted a probabilistic analysis in order to estimate the probability that MT-IV-tPA is cost-effective compared to IV-tPA, the expected value of perfect information (EVPI), and the expected value of partial perfect information (EVPPI), given the uncertainty surrounding the value of our model's parameters. RESULTS: The total mean (standard deviation (SD) cost of MT was €6708.9 (2357.0). The incremental cost-effectiveness ratio (ICER) of the strategy using IV-tPA combined to MT costs was €14,715 per QALY gained as compared to a strategy using IV-tPA alone. The probabilistic analysis showed that the probability of MT-IV-TPA being cost-effective was 85.4% at threshold willingness-to-pay of €30,000 per QALY gained, reaching 98% at €50,000 per QALY gained. CONCLUSION: Although there is no universally accepted willingness-to-pay threshold in France, our analysis suggest that MT combined to IV-tPA can be considered a cost-effective treatment compared with IV-tPA alone.
Keywords:
Acute ischemic stroke; Cost-effectiveness; Cost-utility; Expected value of partial perfect information; Expected value of perfect information; Mechanical thrombectomy