Bayesian spatio-temporal models for mapping TB mortality risk and its relationship with social inequities in a region from Brazilian Legal Amazon

Ecological-level factors associated with tuberculosis incidence and mortality: A systematic review and meta-analysis

Globally, tuberculosis (TB) is the leading infectious cause of morbidity and mortality, with the risk of infection affected by both individual and ecological-level factors. While systematic reviews on individual-level factors exist, there are currently limited studies examining ecological-level factors associated with TB incidence and mortality. This study was conducted to identify ecological factors associated with TB incidence and mortality. A systematic search for analytical studies reporting ecological factors associated with TB incidence or mortality was conducted across electronic databases such as PubMed, Embase, Scopus, and Web of Science, from each database's inception to October 30, 2023. A narrative synthesis of evidence on factors associated with TB incidence and mortality from all included studies, alongside random-effects meta-analysis where applicable, estimated the effects of each factor on TB incidence. A total of 52 articles were included in the analysis, and one study analysed two outcomes, giving 53 studies. Narrative synthesis revealed predominantly positive associations between TB incidence and factors such as temperature (10/18 studies), precipitation (4/6), nitrogen dioxide (6/9), poverty (4/4), immigrant population (3/4), urban population (3/8), and male population (2/4). Conversely, air pressure (3/5), sunshine duration (3/8), altitude (2/4), gross domestic product (4/9), wealth index (2/8), and TB treatment success rate (2/2) mostly showed negative associations. Particulate matter (1/1), social deprivation (1/1), and population density (1/1) were positively associated with TB mortality, while household income (2/2) exhibited a negative association. In the meta-analysis, higher relative humidity (%) (relative risk (RR) = 1.45, 95%CI:1.12, 1.77), greater rainfall (mm) (RR = 1.56, 95%CI: 1.11, 2.02), elevated sulphur dioxide (μg m-3) (RR = 1.04, 95% CI:1.01, 1.08), increased fine particulate matter concentration (PM2.5) (μg/ m3) (RR = 1.33, 95% CI: 1.18, 1.49), and higher population density (people/km2) (RR = 1.01,95%CI:1.01-1.02) were associated with increased TB incidence. Conversely, higher average wind speed (m/s) (RR = 0.89, 95%CI: 0.82,0.96) was associated with decreased TB incidence. TB incidence and mortality rates were significantly associated with various climatic, socioeconomic, and air quality-related factors. Intersectoral collaboration across health, environment, housing, social welfare and economic sectors is imperative for developing integrated approaches that address the risk factors associated with TB incidence and mortality.

A comparative analysis of classical and machine learning methods for forecasting TB/HIV co-infection

TB/HIV coinfection poses a complex public health challenge. Accurate forecasting of future trends is essential for efficient resource allocation and intervention strategy development. This study compares classical statistical and machine learning models to predict TB/HIV coinfection cases stratified by gender and the general populations. We analyzed time series data using exponential smoothing and ARIMA to establish the baseline trend and seasonality. Subsequently, machine learning models (SVR, XGBoost, LSTM, CNN, GRU, CNN-GRU, and CNN-LSTM) were employed to capture the complex dynamics and inherent non-linearities of TB/HIV coinfection data. Performance metrics (MSE, MAE, sMAPE) and the Diebold-Mariano test were used to evaluate the model performance. Results revealed that Deep Learning models, particularly Bidirectional LSTM and CNN-LSTM, significantly outperformed classical methods. This demonstrates the effectiveness of Deep Learning for modeling TB/HIV coinfection time series and generating more accurate forecasts.