Graph theoretical analysis of tuberculosis transmission in Western Province, Sri Lanka with integrated forecasting and TOPSIS

Abstract

Tuberculosis (TB), caused by Mycobacterium tuberculosis, remains a persistent global health challenge, particularly prevalent in low- and middle-income countries due to socioeconomic disparities and limited healthcare access. This study focuses on the Western Province of Sri Lanka, characterized by dense population centres, extensive trade, tourism, and significant urban-rural mobility, all contributing factors to the rapid spread of TB. The research aims to refine TB control strategies through several key objectives. First, the study forecasts TB cases for 2022 across genders, age groups, and types of TB using data up to 2021. Despite lacking comprehensive data beyond 2022, the methodological framework and insights offer valuable guidance for future research and public health strategies. This limitation underscores the need for continued data collection and analysis to refine predictive models and enhance TB control measures. Second, the research involves constructing a comprehensive patient data table integrating demographic details, TB types, and contact information of TB patients in the Western Province. Third, the study uses patient contact data to create a TB contact network model representing disease transmission dynamics. Principles of graph theory, particularly those within social network analysis, are then applied to analyze the network, focusing on metrics such as closeness and betweenness centrality to identify influential nodes and structural characteristics like modularity and average clustering coefficient to understand community formations. Fourth, the study implements the TOPSIS method to rank nodes within the TB contact network based on their potential to spread TB, aiding in identifying critical nodes for targeted intervention strategies. Data collection included compiling TB annual reports from 2012 to 2021 from the National Programme for Tuberculosis Control and Chest Diseases (NPTCCD) website, and statistical forecasts for 2022 were generated using ARIMA models in R, projecting TB cases by gender, age groups, and TB type. Descriptive statistics highlighted demographic trends, with male patients and the 55-64 age group projected to have higher TB incidence rates, primarily pulmonary TB cases. Network analysis revealed a low average clustering coefficient (0.002) and a high modularity value (0.856), indicating sparse local connections and strong community structures within the TB contact network. Comparative analysis pre- and post-TOPSIS implementation demonstrated an enhanced ability to pinpoint influential spreaders, enhancing the precision of targeted intervention strategies. This study underscores the importance of tailored medical and community-based interventions in the Western Province to effectively mitigate TB transmission, emphasizing optimizing TB control measures and improving treatment outcomes.