2016/10/27 by Khanh To Duc, Duc, Khanh To, Monica Chiogna +3
Health Professions · Mathematics · #62C99 #62P10 #FOS: Computer and information sciences #Medical Coding and Health Information #Methodology (stat.ME) #Statistical Methods and Inference #Statistical Methods in Clinical Trials
paper · pdf · doi:10.48550/arxiv.1610.08805
openalex publication_date 2016/10/27 · openalex created_date 2022/10/01 · openalex updated_date 2026/07/28
The volume under the receiver operating characteristic surface (VUS) is\nuseful for measuring the overall accuracy of a diagnostic test when the\npossible disease status belongs to one of three ordered categories. In medical\nstudies, the VUS of a new test is typically estimated through a sample of\nmeasurements obtained by some suitable sample of patients. However, in many\ncases, only a subset of such patients has the true disease status assessed by a\ngold standard test. In this paper, for a continuous-scale diagnostic test, we\npropose four estimators of the VUS which accommodate for nonignorable\nmissingness of the disease status. The estimators are based on a parametric\nmodel which jointly describes both the disease and the verification process.\nIdentifiability of the model is discussed. Consistency and asymptotic normality\nof the proposed estimators are shown, and variance estimation is discussed. The\nfinite-sample behavior is investigated by means of simulation experiments. An\nillustration is provided.\n