Anita Lal1, Mohammadreza Mohebi2, Rohan Sweeney3, Marjory Moodie4, Anna Peeters5, Rob Carter6. 1. Deakin Health Economics, Centre for Population Health Research, Deakin University, Geelong, Victoria, Australia; Global Obesity Centre (GLOBE), Centre for Population Health Research, Deakin University, Geelong, Victoria, Australia. Electronic address: anital@deakin.edu.au. 2. Biostatics Unit, Faculty of Health, Deakin University, Geelong, Victoria, Australia. 3. Centre for Health Economics, Monash Business School, Monash University, Melbourne, Victoria, Australia. 4. Deakin Health Economics, Centre for Population Health Research, Deakin University, Geelong, Victoria, Australia; Global Obesity Centre (GLOBE), Centre for Population Health Research, Deakin University, Geelong, Victoria, Australia. 5. Global Obesity Centre (GLOBE), Centre for Population Health Research, Deakin University, Geelong, Victoria, Australia. 6. Deakin Health Economics, Centre for Population Health Research, Deakin University, Geelong, Victoria, Australia.
Abstract
BACKGROUND: There is an implicit equity approach in cost-effectiveness analysis that values health gains of socioeconomic position groups equally. An alternative approach is to integrate equity by weighting quality-adjusted life-years according to the socioeconomic position group. OBJECTIVES: To use two approaches to derive equity weights for use in cost-effectiveness analysis in Australia, in contexts in which the use of the traditional nonweighted quality-adjusted life-years could increase health inequalities between already disadvantaged groups. METHODS: Equity weights derived using epidemiological data used burden of disease and mortality data by Socio-Economic Indexes for Areas quintiles from the Australian Institute of Health and Welfare. Two ratios were calculated comparing quintile 1 (lowest) to the total Australian population, and comparing quintile 1 to quintile 5 (highest). Preference-based weights were derived using a discrete choice experiment survey (n = 710). Respondents chose between two programs, with varying gains in life expectancy going to a low- or a high-income group. A probit model incorporating nominal values of the difference in life expectancy was estimated to calculate the equity weights. RESULTS: The epidemiological weights ranged from 1.2 to 1.5, with larger weights when quintile 5 was the denominator. The preference-based weights ranged from 1.3 (95% confidence interval 1.2-1.4) to 1.8 (95% confidence interval 1.6-2.0), with a tendency for increasing weights as the gains to the low-income group increased. CONCLUSIONS: Both methods derived plausible and consistent weights. Using weights of different magnitudes in sensitivity analysis would allow the appropriate weight to be considered by decision makers and stakeholders to reflect policy objectives.
BACKGROUND: There is an implicit equity approach in cost-effectiveness analysis that values health gains of socioeconomic position groups equally. An alternative approach is to integrate equity by weighting quality-adjusted life-years according to the socioeconomic position group. OBJECTIVES: To use two approaches to derive equity weights for use in cost-effectiveness analysis in Australia, in contexts in which the use of the traditional nonweighted quality-adjusted life-years could increase health inequalities between already disadvantaged groups. METHODS: Equity weights derived using epidemiological data used burden of disease and mortality data by Socio-Economic Indexes for Areas quintiles from the Australian Institute of Health and Welfare. Two ratios were calculated comparing quintile 1 (lowest) to the total Australian population, and comparing quintile 1 to quintile 5 (highest). Preference-based weights were derived using a discrete choice experiment survey (n = 710). Respondents chose between two programs, with varying gains in life expectancy going to a low- or a high-income group. A probit model incorporating nominal values of the difference in life expectancy was estimated to calculate the equity weights. RESULTS: The epidemiological weights ranged from 1.2 to 1.5, with larger weights when quintile 5 was the denominator. The preference-based weights ranged from 1.3 (95% confidence interval 1.2-1.4) to 1.8 (95% confidence interval 1.6-2.0), with a tendency for increasing weights as the gains to the low-income group increased. CONCLUSIONS: Both methods derived plausible and consistent weights. Using weights of different magnitudes in sensitivity analysis would allow the appropriate weight to be considered by decision makers and stakeholders to reflect policy objectives.
Authors: Kiffer G Card; Marina Adshade; Robert S Hogg; Jody Jollimore; Nathan J Lachowsky Journal: BMC Public Health Date: 2022-06-13 Impact factor: 4.135