Christine Mpundu-Kaambwa1, Gang Chen2, Remo Russo3,4, Katherine Stevens5, Karin Dam Petersen6, Julie Ratcliffe7. 1. Institute for Choice, University of South Australia, Business School, Adelaide, South Australia, Australia. mpucc001@mymail.unisa.edu.au. 2. Centre for Health Economics, Monash Business School, Monash University, Melbourne, Victoria, Australia. 3. Faculty of Health Sciences, School of Medicine, Flinders University, Adelaide, South Australia, Australia. 4. Department of Paediatric Rehabilitation, Women's and Children's Hospital, Adelaide, South Australia, Australia. 5. Health Economics and Decision Science, University of Sheffield, Sheffield, United Kingdom. 6. Department of Business and Management, Faculty of Social Sciences, Aalborg University, Aalborg East, Denmark. 7. Flinders Health Economics Group, Flinders University, Adelaide, South Australia, Australia.
Abstract
BACKGROUND: The Pediatric Quality of Life Inventory™ 4.0 Short Form 15 Generic Core Scales (hereafter the PedsQL) and the Child Health Utility-9 Dimensions (CHU9D) are two generic instruments designed to measure health-related quality of life in children and adolescents in the general population and paediatric patient groups living with specific health conditions. Although the PedsQL is widely used among paediatric patient populations, presently it is not possible to directly use the scores from the instrument to calculate quality-adjusted life-years (QALYs) for application in economic evaluation because it produces summary scores which are not preference-based. OBJECTIVE: This paper examines different econometric mapping techniques for estimating CHU9D utility scores from the PedsQL for the purpose of calculating QALYs for cost-utility analysis. METHODS: The PedsQL and the CHU9D were completed by a community sample of 755 Australian adolescents aged 15-17 years. Seven regression models were estimated: ordinary least squares estimator, generalised linear model, robust MM estimator, multivariate factorial polynomial estimator, beta-binomial estimator, finite mixture model and multinomial logistic model. The mean absolute error (MAE) and the mean squared error (MSE) were used to assess predictive ability of the models. RESULTS: The MM estimator with stepwise-selected PedsQL dimension scores as explanatory variables had the best predictive accuracy using MAE and the equivalent beta-binomial model had the best predictive accuracy using MSE. CONCLUSIONS: Our mapping algorithm facilitates the estimation of health-state utilities for use within economic evaluations where only PedsQL data is available and is suitable for use in community-based adolescents aged 15-17 years. Applicability of the algorithm in younger populations should be assessed in further research.
BACKGROUND: The Pediatric Quality of Life Inventory™ 4.0 Short Form 15 Generic Core Scales (hereafter the PedsQL) and the Child Health Utility-9 Dimensions (CHU9D) are two generic instruments designed to measure health-related quality of life in children and adolescents in the general population and paediatric patient groups living with specific health conditions. Although the PedsQL is widely used among paediatric patient populations, presently it is not possible to directly use the scores from the instrument to calculate quality-adjusted life-years (QALYs) for application in economic evaluation because it produces summary scores which are not preference-based. OBJECTIVE: This paper examines different econometric mapping techniques for estimating CHU9D utility scores from the PedsQL for the purpose of calculating QALYs for cost-utility analysis. METHODS: The PedsQL and the CHU9D were completed by a community sample of 755 Australian adolescents aged 15-17 years. Seven regression models were estimated: ordinary least squares estimator, generalised linear model, robust MM estimator, multivariate factorial polynomial estimator, beta-binomial estimator, finite mixture model and multinomial logistic model. The mean absolute error (MAE) and the mean squared error (MSE) were used to assess predictive ability of the models. RESULTS: The MM estimator with stepwise-selected PedsQL dimension scores as explanatory variables had the best predictive accuracy using MAE and the equivalent beta-binomial model had the best predictive accuracy using MSE. CONCLUSIONS: Our mapping algorithm facilitates the estimation of health-state utilities for use within economic evaluations where only PedsQL data is available and is suitable for use in community-based adolescents aged 15-17 years. Applicability of the algorithm in younger populations should be assessed in further research.
Authors: Rafael A Pinedo-Villanueva; David Turner; Andrew Judge; James P Raftery; Nigel K Arden Journal: Qual Life Res Date: 2012-04-22 Impact factor: 4.147
Authors: M Boers; A C Verhoeven; H M Markusse; M A van de Laar; R Westhovens; J C van Denderen; D van Zeben; B A Dijkmans; A J Peeters; P Jacobs; H R van den Brink; H J Schouten; D M van der Heijde; A Boonen; S van der Linden Journal: Lancet Date: 1997-08-02 Impact factor: 79.321
Authors: Tosin Lambe; Emma Frew; Natalie J Ives; Rebecca L Woolley; Carole Cummins; Elizabeth A Brettell; Emma N Barsoum; Nicholas J A Webb Journal: Pharmacoeconomics Date: 2018-04 Impact factor: 4.981