Predicting EuroQoL 5 Dimensions 5 Levels (EQ-5D-5L) Utilities from Older People's Quality of Life Brief Questionnaire (OPQoL-Brief) Scores

作     者：Kaambwa, Billingsley Ratcliffe, Julie 

作者机构：Flinders Univ S Australia Sch Med Hlth Econ Unit Hlth Sci BldgBedford Pk CampusSturt Rd Bedford Pk SA 5042 Australia Univ South Australia Inst Choice Sch Business Way Lee BldgCity West Campus Adelaide SA Australia 

出 版 物：《PATIENT-PATIENT CENTERED OUTCOMES RESEARCH》 (患者)

年 卷 期：2018年第11卷第1期

页      面：39-54页

核心收录：

学科分类：1204[管理学-公共管理] 10[医学] 

基　　金：Australian Research Council [LP110200079] ACH Group Catholic Community Services Hammondcare Helping Hand Resthaven Flinders Telehealth in the Home (FTH) Australian Government Australian Research Council [LP110200079] Funding Source: Australian Research Council 

主　　题：Legal Services Health-Related Quality Of Life load capacity based algorithm old animals Public utilities Hash based algorithms Questionnaires Persons EuroQOL Economic Evaluation evolutionary algorithm 

摘      要：Economic evaluation of healthcare treatment and services targeted at older people requires measurement of utility-based quality-of-life outcomes but it is not always possible to collect such outcome data. It may, however, be possible to estimate these outcomes using non-utility measures of quality of life where the latter have been collected. The objective of this study was to develop a regression-based algorithm to map a non-utility-based outcome, the Older People s Quality of Life brief questionnaire (OPQoL-brief), onto a utility-based outcome, the EuroQoL 5 Dimensions 5 Levels (EQ-5D-5L). The estimation sample comprised 330 community-based Australian older people ( 65 years), while the validation sample consisted of 293 older people from a separate study. Six regression techniques were employed to estimate utilities from OPQoL-brief. The predictive accuracy of 54 regression models (six regression techniques x nine model specifications) 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 0.05. The 54 regression models predicted EQ-5D-5L utilities that performed differently when assessed by the six criteria. However, best results were obtained from an ordinary least squares (OLS) model where all 13 OPQoL-brief items were included as continuous variables (OLS 4). RMSE and MAE estimates for this model (0.2201 and 0.1638, respectively) were within the range of published estimates. It is possible to predict valid utilities from OPQoL-brief using regression methods. We recommend OLS model (4) for this exercise.