2018/05/18 by Andrea Gabrio, Gabrio, Andrea, Rachael Hunter +5
Economics, Econometrics and Finance · Mathematics · Medicine · #Advanced Causal Inference Techniques #Applications (stat.AP) #Computer science #Data mining #Econometrics #Economic and Environmental Valuation #Economics #Estimation #FOS: Computer and information sciences #Health Systems, Economic Evaluations, Quality of Life #Longitudinal data #Machine learning #Mathematics #Medicine #Missing data #Random effects model #Resource (disambiguation) #Statistics #stat.AP
paper · pdf · doi:10.48550/arxiv.1805.07149
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2018/05/18 · arxiv created 2020/05/22 · arxiv updated 2020/05/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Health economic evaluations based on patient-level data collected alongside\nclinical trials~(e.g. health related quality of life and resource use measures)\nare an important component of the process which informs resource allocation\ndecisions. Almost inevitably, the analysis is complicated by the fact that some\nindividuals drop out from the study, which causes their data to be unobserved\nat some time point. Current practice performs the evaluation by handling the\nmissing data at the level of aggregated variables (e.g. QALYs), which are\nobtained by combining the economic data over the duration of the study, and are\noften conducted under a missing at random (MAR) assumption. However, this\napproach may lead to incorrect inferences since it ignores the longitudinal\nnature of the data and may end up discarding a considerable amount of\nobservations from the analysis. We propose the use of joint longitudinal models\nto extend standard cost-effectiveness analysis methods by taking into account\nthe longitudinal structure and incorporate all available data to improve the\nestimation of the targeted quantities under MAR. Our approach is compared to\npopular missingness approaches in trial-based analyses, motivated by an\nexploratory simulation study, and applied to data from two real case studies.\n