Determinants, risk factors and spatial analysis of multi-drug resistant pulmonary tuberculosis in Jodhpur, India

Abstract

This study was planned to estimate the proportion of confirmed multi-drug resistance pulmonary tuberculosis (TB) cases out of the presumptive cases referred to DTC (District Tuberculosis Center) Jodhpur for diagnosis; to identify clinical and socio-demographic risk factors associated with the multidrug-resistant pulmonary TB and to assess the spatial distribution to find out clustering and pattern in the distribution of pulmonary TB with the help of Geographic Information System (GIS). In the Jodhpur district, 150 confirmed pulmonary multi-drug resistant tuberculosis (MDR-TB) cases, diagnosed by probe-based molecular drug susceptibility testing method and categorized as MDR in DTC's register (District Tuberculosis Center), were taken. Simultaneously, 300 control of confirmed non-MDR or drug-sensitive pulmonary TB patients were taken. Statistical analysis was done with logistic regression. In addition, for spatial analysis, secondary data from 2013-17 was analyzed using Global Moran's I and Getis and Ordi (Gi*) statistics. In 2012-18, a total of 12563 CBNAAT (Cartridge-based nucleic acid amplification test) were performed. 2898 (23%) showed M. TB positive but rifampicin sensitive, and 590 (4.7%) showed rifampicin resistant. Independent risk factors for MDR TB were ≤60 years age (AOR 3.0, CI 1.3-7.1); male gender (AOR 3.4, CI 1.8-6.7); overcrowding (AOR 1.6, CI 1.0-2.7); using chulha (smoke appliance) for cooking (AOR 2.5, CI 1.2-4.9), past TB treatment (AOR 5.7, CI 2.9-11.3) and past contact with MDR patient (AOR 10.7, CI 3.7-31.2). All four urban TUs (Tuberculosis Units) had the highest proportion of drug-resistant pulmonary TB. There was no statistically significant clustering, and the pattern of cases was primarily random. Most of the hotspots generated were present near the administrative boundaries of TUs, and the new ones mostly appeared in the area near the previous hotspots. A random pattern seen in cluster analysis supports the universal drug testing policy of India. Hotspot analysis helps cross administrative border initiatives with targeted active case finding and proper follow-up.