A MULTILEVEL ANALYSIS: MODEL OF TUBERCULOSIS SUSPECT IN ACCESS TO HEALTH SERVICES IN INDONESIA

At present, three countries, namely India, China, and Indonesia, contribute the largest tuberculosis (TB) cases in the world. The World Health Organization has set a target of minimum 70% case detection rate (CDR). However, until 2004, the CDR in Indonesia was only 51.8%. Although TB drugs are provided for fiee and trained personnel are ready in appointed health facilities to provide DOTS, CDR has been low due to low access to the facilities. The speciJic objectives of this study were to develope equation modeling of individual, household, and district level variables aflecting the access. This research used cross sectional data of BES 11 evaluation and BPS data. Multilevel statistical analyses of GLLAMM logistic regression intercept were carried out to attain the study objectives. The study found that variables associated with access to TB centers at individual level are knowledge offiee TB programs (OR = 2.2; 95% CI 1.97-2.81), health insurance (OR = 2.9; 95% CI 1.15-7.75), advice (OR = 3.5; 95% CI 1.69-7.14), perceived moderate costs of health centers (OR = 1.54; 95% CI 1.11 -4.79), perceived cheap transporation cost (OR = 7.8; 95% CI 2.5 7-23.43), and perceived nearness to health facilities (OR = 22.5; 95% CI 3.4-1 48.4). At the household level, are: time travel to health facilities and household income per capita, and at the district level, these factors are the proportion of individual below poverty (OR of second quarter = 0.49; 95% CI 0.22-0.83) and the ratio of population to health facilities (OR of second quarter = 0.43; 95% CI 0.12-0.66).