Ru Han1,2, Clément François3,4, Mondher Toumi3,4. 1. University of Aix-Marseille, Marseille, France. rha@creativ-ceutical.com. 2. Creativ-Ceutical, 215, rue de Faubourg St-Honoré, 75008, Paris, France. rha@creativ-ceutical.com. 3. University of Aix-Marseille, Marseille, France. 4. Creativ-Ceutical, 215, rue de Faubourg St-Honoré, 75008, Paris, France.
Abstract
BACKGROUND: Health state utility values (HSUVs) identified from utility elicitation studies are widely used in pharmacoeconomic evaluations for chronic hepatitis C (CHC) and are particularly instrumental in health technology assessment (HTA) evaluations such as those from the National Institute for Health and Care Excellence (NICE). OBJECTIVE: The aim of this study was to identify HSUVs used in cost-utility analyses (CUAs) for CHC in Europe and to evaluate the impact of HSUV selection on cost-effectiveness results in terms of the incremental cost per quality-adjusted life-year (QALY) gained (ICER). METHODS: A systematic search of pharmacoeconomic evaluations for CHC was updated in the MEDLINE and EMBASE databases for the periods 2012-2017 and 2017-2020. Data on health states, HSUVs, and utility elicitation studies were extracted. The difference in HSUVs of the same health state in different CUAs, and the difference between HSUVs of one health state and of the interlink health state in the same CUAs, were calculated. A quality assessment was performed to evaluate the selection of HSUVs in CUAs. Sets of HSUVs identified were used in a reconstructed CUA model to assess the impact on the ICER. RESULTS: Twenty-six CUAs conducted in European countries and referring to 17 utility elicitation studies were included. The difference in HSUVs of the same health state in different CUAs ranged from 0.021 (liver transplant) to 0.468 (decompensated cirrhosis). The difference between HSUVs of one health state and of the interlink health state of the next disease severity level was calculated between the health states of F0-F1/mild and F2-F3/moderate (n = 11, 0.040-0.110), F2-F3/moderate and F4/compensated cirrhosis (n = 18, 0.027-0.130), compensated cirrhosis and decompensated cirrhosis (n = 22, 0.020-0.100), decompensated cirrhosis and hepatocellular carcinoma (n = 24, 0.000-0.200), hepatocellular carcinoma and liver transplant in the first year (n = 17, - 0.329 to 0.170) and liver transplant in the first and subsequent years (n = 17, - 0.340 to 0.000). The utility elicitation study selected by most CUAs (n = 11) was recommended as the source of HSUVs, at least for the CUAs conducted in the UK, based on the results of quality assessment. Seven sets of HSUVs were generated to fit the reconstructed model and changed the results of the incremental analysis from being cost effective to not being cost effective (ICER ranging from £2460 to £24,954 per QALY gained), and to being dominated in the UK setting. CONCLUSIONS: The CUAs for CHC were found to apply to various HSUVs from different utility elicitation studies in the same health state. This variability in HSUVs has the potential to significantly affect ICER and ICER-based reimbursement decisions. A rigorous selection of HSUVs in CUAs to inform healthcare resource allocation is suggested for future studies of CUAs and for guideline development.
BACKGROUND: Health state utility values (HSUVs) identified from utility elicitation studies are widely used in pharmacoeconomic evaluations for chronic hepatitis C (CHC) and are particularly instrumental in health technology assessment (HTA) evaluations such as those from the National Institute for Health and Care Excellence (NICE). OBJECTIVE: The aim of this study was to identify HSUVs used in cost-utility analyses (CUAs) for CHC in Europe and to evaluate the impact of HSUV selection on cost-effectiveness results in terms of the incremental cost per quality-adjusted life-year (QALY) gained (ICER). METHODS: A systematic search of pharmacoeconomic evaluations for CHC was updated in the MEDLINE and EMBASE databases for the periods 2012-2017 and 2017-2020. Data on health states, HSUVs, and utility elicitation studies were extracted. The difference in HSUVs of the same health state in different CUAs, and the difference between HSUVs of one health state and of the interlink health state in the same CUAs, were calculated. A quality assessment was performed to evaluate the selection of HSUVs in CUAs. Sets of HSUVs identified were used in a reconstructed CUA model to assess the impact on the ICER. RESULTS: Twenty-six CUAs conducted in European countries and referring to 17 utility elicitation studies were included. The difference in HSUVs of the same health state in different CUAs ranged from 0.021 (liver transplant) to 0.468 (decompensated cirrhosis). The difference between HSUVs of one health state and of the interlink health state of the next disease severity level was calculated between the health states of F0-F1/mild and F2-F3/moderate (n = 11, 0.040-0.110), F2-F3/moderate and F4/compensated cirrhosis (n = 18, 0.027-0.130), compensated cirrhosis and decompensated cirrhosis (n = 22, 0.020-0.100), decompensated cirrhosis and hepatocellular carcinoma (n = 24, 0.000-0.200), hepatocellular carcinoma and liver transplant in the first year (n = 17, - 0.329 to 0.170) and liver transplant in the first and subsequent years (n = 17, - 0.340 to 0.000). The utility elicitation study selected by most CUAs (n = 11) was recommended as the source of HSUVs, at least for the CUAs conducted in the UK, based on the results of quality assessment. Seven sets of HSUVs were generated to fit the reconstructed model and changed the results of the incremental analysis from being cost effective to not being cost effective (ICER ranging from £2460 to £24,954 per QALY gained), and to being dominated in the UK setting. CONCLUSIONS: The CUAs for CHC were found to apply to various HSUVs from different utility elicitation studies in the same health state. This variability in HSUVs has the potential to significantly affect ICER and ICER-based reimbursement decisions. A rigorous selection of HSUVs in CUAs to inform healthcare resource allocation is suggested for future studies of CUAs and for guideline development.
Authors: Scott A McDonald; Amirah Azzeri; Fatiha Hana Shabaruddin; Maznah Dahlui; Soek S Tan; Adeeba Kamarulzaman; Rosmawati Mohamed Journal: Appl Health Econ Health Policy Date: 2018-12 Impact factor: 2.561
Authors: Rodolfo Castro; Louise Crathorne; Hugo Perazzo; Julio Silva; Chris Cooper; Jo Varley-Campbell; Daniel Savignon Marinho; Marcela Haasova; Valdilea G Veloso; Rob Anderson; Chris Hyde Journal: BMC Med Res Methodol Date: 2018-06-13 Impact factor: 4.615