Multiple Decrement Modeling of BPJS Health Transitions Under Air Pollution Exposure in DKI Jakarta

Abstract

Air pollution represents a significant environmental health burden in Indonesia. Despite this, the quantitative relationship between ambient pollutant concentrations and population-level health state transitions within Indonesia’s national health insurance system remains uncharacterized. This study estimates cause-specific forces of decrement for health state transitions within the BPJS Kesehatan administrative claims data from DKI Jakarta spanning 2022 to 2024, and examines the association between ambient air pollutant concentrations and estimated transition intensities through exploratory LASSO regression. The analytical dataset comprises 454,001 FKTP and 147,379 FKRTL episodes, stratified by five-year age bands and calendar quarter. Multiple decrements across both care settings are modeled, with the FKTP-to-FKRTL referral (rujuk lanjut) decrement serving as the primary focus. NO₂ emerged as a candidate predictor in exploratory screening, appearing with a positive coefficient in 6 of 17 age bands, with model-implied elasticities ranging from 0.063 to 0.544. Scenario analysis showed that a hypothetical 20% increase in ambient NO₂ corresponded to fitted referral intensity increases of between 1.3 and 10.9 percent across retained age bands.