Small area estimation for morbidity rate prediction

Abstract

Complete and up-to-date data is important for development. However, its provision is costly. Survey is more efficient data collection solution, but the available sample is sometimes unable to produce direct estimate with sufficient precision. Small area estimation is one of the solutions by increasing the effectiveness of the sample. Indicators need to be presented with good precision so that policies are right on target. One of them is the health indicator which is expressed by morbidity rate. Morbidity rate is the percentage of population with health complaints that interfere with activities. Precision morbidity data is needed as criterion for determining non-communicable disease prevention at regional level as mandated by Law No. 17 of 2023. However, direct estimation from the survey shows that there are still districts/cities with Relative Standard Error (RSE) more than 25% like in Papua Island. Even though, especially Papua Province has an increasing morbidity rate. For this reason, in this study, indirect estimation of SAE Hierarchical Bayes (HB) Beta-Logistic was carried out to obtain estimate with good precision (RSE less than 25%). The results emphasize that SAE HB Beta-Logistic provide the estimate of morbidity rate precisely than the direct estimates for all districts/cities on the Papua Island.