Spatial and temporal analysis of xpert confimed MDR-TB in the North-West province of South Africa from 2011-2015

Abstract

The burden of multi-drug resistant Tuberculosis (MDR-TB) cases is a reflection of the continuous and recent transmission of infection within the population. This study therefore aimed to describe the spatiotemporal distribution of Xpert tested Rifampicin (RIF) resistant cases using the Gene Xpert CT probe dropout/delay (a test to diagnose the presence and the type of RIF resistance from specific Mycobacterium tuberculosis strain), and also evaluate the demographic and socio-economic factors that are associated with the distribution of MDR-TB in the North West province from 2011-2015. Spatial and spatiotemporal autocorrelation were explored using the Global Moran’s I, local Anselin Moran’s I and Kulldorff’s scan statistics. A multivariate zero-truncated Poisson model was used to generate estimates of parameters for demographic and socioeconomic factors. A total of 2949 RIF resistant cases, reported by the National Health Laboratory Services (NHLS) testing sites from 2011-2015 was added to the analysis. The cumulative incidence rate of MDR-TB was 139.4 cases/100,000 inhabitants. The highest incidence rate of 223.7 cases/100,000 inhabitants was seen in the City of Matlosana municipality. There was no evidence of temporal clustering but spatial and spatiotemporal clustering of MDR-TB cases were mostly observed in the South Western to the South Eastern region of the province where potential overcrowding from seasonal migration and economic activities like mining are prevalent. Mycobacterium tuberculosis (MTB) strains with genotypes causing the probe dropout/delay of 17 (RR=5.00, P-value<0.001), 20 (RR=5.22, P-value<0.001) and 24 (RR=5.3, P-value<0.001) were mostly transmitted within these MDR-TB hotspots in 2012. The Municipal risk for more MDR-TB cases, was positively associated with, MTB bacterial load (IRR=1.15, 95% CI: 1.05-1.26), youth male to female ratio (IRR=1.59, 95% CI: 1.19-2.14) and male proportion (IRR=1.09, 95% CI: 1.01-1.16) and negatively associated with the number of households (IRR=0.87, 95% CI: 0.79-0.94, P-value: 0.001). There is evidence of hotspots in the South Western and South Eastern regions of North West Province resulting from direct transmission of specific MTB strains between clustered facilities during the study period. Also, beyond the direct transmission, other ecological factors such as increasing median bacterial load and increased male to female ratio can increase the municipality risk of MDR-TB. These hotspots should be prioritized in the implementation of strategic MDR-TB control programs