Billingsley Kaambwa1, Gang Chen2, Julie Ratcliffe3, Angelo Iezzi2, Aimee Maxwell2, Jeff Richardson2. 1. Flinders Health Economics Group, Flinders University, A Block, Repatriation General Hospital, 202-16 Daws Road, Daw Park, Adelaide, SA, 5041, Australia. billingsley.kaambwa@flinders.edu.au. 2. Centre for Health Economics, Building 75, 15 Innovation Walk, Monash University, Clayton, VIC, 3800, Australia. 3. Flinders Health Economics Group, Flinders University, A Block, Repatriation General Hospital, 202-16 Daws Road, Daw Park, Adelaide, SA, 5041, Australia.
Abstract
PURPOSE: Economic evaluation of health services commonly requires information regarding health-state utilities. Sometimes this information is not available but non-utility measures of quality of life may have been collected from which the required utilities can be estimated. This paper examines the possibility of mapping a non-utility-based outcome, the Sydney Asthma Quality of Life Questionnaire (AQLQ-S), onto five multi-attribute utility instruments: Assessment of Quality of Life 8 Dimensions (AQoL-8D), EuroQoL 5 Dimensions 5-Level (EQ-5D-5L), Health Utilities Index Mark 3 (HUI3), 15 Dimensions (15D), and the Short-Form 6 Dimensions (SF-6D). METHODS: Data for 856 individuals with asthma were obtained from a large Multi-Instrument Comparison (MIC) survey. Four statistical techniques were employed to estimate utilities from the AQLQ-S. The predictive accuracy of 180 regression models was assessed using six criteria: mean absolute error (MAE), root mean squared error (RMSE), correlation, distribution of predicted utilities, distribution of residuals, and proportion of predictions with absolute errors <0.0.5. Validation of initial 'primary' models was carried out on a random sample of the MIC data. RESULTS: Best results were obtained with non-linear models that included a quadratic term for the AQLQ-S score along with demographic variables. The four statistical techniques predicted models that performed differently when assessed by the six criteria; however, the best results, for both the estimation and validation samples, were obtained using a generalised linear model (GLM estimator). CONCLUSIONS: It is possible to predict valid utilities from the AQLQ-S using regression methods. We recommend GLM models for this exercise.
PURPOSE: Economic evaluation of health services commonly requires information regarding health-state utilities. Sometimes this information is not available but non-utility measures of quality of life may have been collected from which the required utilities can be estimated. This paper examines the possibility of mapping a non-utility-based outcome, the Sydney Asthma Quality of Life Questionnaire (AQLQ-S), onto five multi-attribute utility instruments: Assessment of Quality of Life 8 Dimensions (AQoL-8D), EuroQoL 5 Dimensions 5-Level (EQ-5D-5L), Health Utilities Index Mark 3 (HUI3), 15 Dimensions (15D), and the Short-Form 6 Dimensions (SF-6D). METHODS: Data for 856 individuals with asthma were obtained from a large Multi-Instrument Comparison (MIC) survey. Four statistical techniques were employed to estimate utilities from the AQLQ-S. The predictive accuracy of 180 regression models was assessed using six criteria: mean absolute error (MAE), root mean squared error (RMSE), correlation, distribution of predicted utilities, distribution of residuals, and proportion of predictions with absolute errors <0.0.5. Validation of initial 'primary' models was carried out on a random sample of the MIC data. RESULTS: Best results were obtained with non-linear models that included a quadratic term for the AQLQ-S score along with demographic variables. The four statistical techniques predicted models that performed differently when assessed by the six criteria; however, the best results, for both the estimation and validation samples, were obtained using a generalised linear model (GLM estimator). CONCLUSIONS: It is possible to predict valid utilities from the AQLQ-S using regression methods. We recommend GLM models for this exercise.
Authors: David Feeny; William Furlong; George W Torrance; Charles H Goldsmith; Zenglong Zhu; Sonja DePauw; Margaret Denton; Michael Boyle Journal: Med Care Date: 2002-02 Impact factor: 2.983
Authors: Nicole K Eberhart; Cathy D Sherbourne; Maria Orlando Edelen; Brian D Stucky; Nancy L Sin; Marielena Lara Journal: Qual Life Res Date: 2013-09-24 Impact factor: 4.147
Authors: M F Janssen; A Simon Pickard; Dominik Golicki; Claire Gudex; Maciej Niewada; Luciana Scalone; Paul Swinburn; Jan Busschbach Journal: Qual Life Res Date: 2012-11-25 Impact factor: 4.147
Authors: Mónica Hernández Alava; Allan Wailoo; Stephen Pudney; Laura Gray; Andrea Manca Journal: Health Technol Assess Date: 2020-06 Impact factor: 4.014