Social network analysis in tuberculosis control among the Aboriginal population of Manitoba

Tuberculosis (TB) in the province of Manitoba (population 1.15 million) is a multi-faceted problem, with trends that parallel the current national trends. Canadian-born people with treaty status and foreign-born individuals make up the majority of new TB cases and have sustained the steady level of TB incidence in Manitoba in the last decade. Therefore, enhanced TB control measures are required for these population subgroups, if the incidence of this important disease is to be further reduced.; The present study is composed of three parts. The first part involves a description of TB in Manitoba between 1992 and 1999. The study found that Canadian-born people with treaty status form the largest single group of all TB cases in Manitoba (44%), although they represent only 8.9% of the provincial population. A single TB strain, FP1, dominates all TB cases with isolates in Manitoba (25.8%). Canadian-born people with treaty status represent 75% of all FP1 cases. Certain Northern First Nation communities (reserves) have TB incidence as high as 496 per 100,000 person-years, and 64% of Canadian born with treaty status TB cases originated from only eight communities.; The second part involved a pilot study of the feasibility of using Social Network Analysis (SNA) in TB epidemiology and control in Manitoba. Contact tracing investigation data kept at the Manitoba TB Registry office was employed. We confirmed that it is feasible to use SNA using the Manitoba TB data, and found that TB outbreak networks in Manitoba have varying degrees of compactness, with boundaries that extend beyond the geopolitical boundaries of the communities under investigation. The pilot study findings raised important questions for further study, in particular the question as to whether TB outbreak networks reflect the social structure of the affected community.; The third part was a case-control study in which SNA was used in conjunction with primary interview data and contact investigation data from one high TB incidence Northern Manitoba Aboriginal community (Community 1). The main objectives of this part were to determine the feasibility and utility of using primary interview data for constructing and analyzing the TB transmission network in a high incidence community, and to determine whether or not the TB network is a reflection of the affected community's social structure. Eighty-three individuals were successfully recruited for the study in this remote community (62% of the target sample). The case-control study confirmed that the TB network reflected the underlying social structure of the community. The TB network of Community 1 was found to be a scale free network and very compact, with an obvious centrally located and densely connected group of network members. The study demonstrated that as k increases in the Seidman k-core collapse sequence, the probability of active or latent tuberculosis infection increases. Those in the network who were exposed to 7 or more TB cases had a 100% probability of being TB cases themselves. Multivariate analysis confirmed that the TB group members (in comparison to the control and contact groups) lived in houses which were significantly more crowded (mean of 1.6 people per room), and were socially connected to a greater number of other TB patients (to 4.0 other TB cases vs. 0.6 for controls). The study enabled the creation of profiles of the TB, contact and control (general population) groups, respectively.; SNA was successfully employed in this study to visualize TB networks, to identify characteristics of key individuals in TB outbreaks, and to delineate outbreak boundaries. It is concluded that SNA is a useful epidemiological tool for TB control. SNA methods can improve contact tracing performance and potentially improve TB control and prevention programs.