Assessment of tuberculosis spatial hotspot areas in Antananarivo, Madagascar, by combining spatial analysis and genotyping

Abstract

Background: Tuberculosis (TB) remains a public health problem in Madagascar. A crucial element of TB control is the development of an easy and rapid method for the orientation of TB control strategies in the country. Our main objective was to develop a TB spatial hotspot identification method by combining spatial analysis and TB genotyping method in Antananarivo.

Methods: Sputa of new pulmonary TB cases from 20 TB diagnosis and treatment centers (DTCs) in Antananarivo were collected from August 2013 to May 2014 for culture. Mycobacterium tuberculosis complex (MTBC) clinical isolates were typed by spoligotyping on a Luminex® 200 platform. All TB patients were respectively localized according to their neighborhood residence and the spatial distribution of all pulmonary TB patients and patients with genotypic clustered isolates were scanned respectively by the Kulldorff spatial scanning method for identification of significant spatial clustering. Areas exhibiting spatial clustering of patients with genotypic clustered isolates were considered as hotspot TB areas for transmission.

Results: Overall, 467 new cases were included in the study, and 394 spoligotypes were obtained (84.4%). New TB cases were distributed in 133 of the 192 Fokontany (administrative neighborhoods) of Antananarivo (1 to 15 clinical patients per Fokontany) and patients with genotypic clustered isolates were distributed in 127 of the 192 Fokontany (1 to 13 per Fokontany). A single spatial focal point of epidemics was detected when ignoring genotypic data (p = 0.039). One Fokontany of this focal point and three additional ones were detected to be spatially clustered when taking genotypes into account (p < 0.05). These four areas were declared potential TB transmission hotspots in Antananarivo and will be considered as priority targets for surveillance in the future.

Conclusion: This method, combining spatial analysis and TB genotyping will now be used for further focused clinical and epidemiological studies in Madagascar and will allow better TB control strategies by public health authorities.

Keywords: Antananarivo; Genotyping; Geographic Information System; Mycobacterium tuberculosis; Spatial cluster.

Fig. 1

Structure of the urban district of Antananarivo (UDA). a describes the distribution of population density per Fokontany and the localization of the principal public markets in the UDA. The poorest Fokontany are hatched. b describes the distribution of the most populous Fokontany. c describes the distribution of TB cases recruited for the study. TB diagnosis and treatment centers included in the study where represented in green or blue point

Fig. 2

Quantitative data on the studied sample population. Out of a population of 1.1 million people investigated during 8 months, and given the previous estimated TB incidence rate in Antananarivo (141 cases/100000 inhabitants/year), we reached a 25% exhaustivity in our recruitment and a 86% of genotyping rate success once geographically inadequate cases were excluded

Fig. 3

a Distribution of spoligotyping-defined lineages obtained on 394 spoligotypes: spoligotyping nomenclature according to Brudey et al. [5], followed by genome-based (L1 to L4) lineage and sublineage nomenclature according to Coll et al. [24]. b Spatial distribution of M. tuberculosis spoligotyping-defined lineages within the 192 Fokontany of Antananarivo. (size of circle is proportional to cases number, cf. Figure)

Fig. 4

Spatial signatures of TB identified by the Kulldorff spatial scan method. a Spatial clustering of TB cases and patients with genotypic clustered isolate. b distribution of isolates families in each spatial clustering of patients with genotypic clustered isolates