A Comparison of Capture-recapture Modeling to Estimate the Number of Patients with Psoriasis in Trang Province

Annual Meeting in Mathematics 2023

Abstract

This study aims to estimate the number of patients with psoriasis in Trang Province between 2015 and 2018. The population size estimator based on Poisson distribution called the MLEPoi was used as a basic model for homogeneity population. However, the heterogeneity often occurs in capture-recapture experiment. The population size estimators based on the Poisson mixture model; the MLEGeo, LCMP, TG, Chao, Zelterman and LB estimators were selected for heterogeneous population. By using the ratio plot of the Poisson model for investigating a suitable model for psoriasis data, the results suggested that the Poisson mixture model performs better than the Poisson model. Therefore, the LCMP and MLEGeo estimators provided the best accuracy with excellent goodness-of-fit over competitors. The TG estimator showed promise as an alternative choices to estimate population size in all cases, wile the LB estimator provided the worst result with violated assumption parameters. Using the most appropriate population size estimator determined the average number of hidden patients with psoriasis in Trang Province at 536 persons per year. Combination of reported and unreported cases to the true number of patients, an estimate of patients with psoriasis was 1,104 (95%: 804–1,045) people per year.