Specific urban units identified in tuberculosis epidemic using a geographical detector in Guangzhou, China

Abstract

Background A remarkable drop in tuberculosis (TB) incidence has been achieved in China, although in 2019 it was still considered the second most communicable disease. However, TB’s spatial features and risk factors in urban areas remain poorly understood. This study aims to identify the spatial differentiations and potential influencing factors of TB in highly urbanized regions on a fine scale. Methods This study included 18 socioeconomic and environmental variables in the four central districts of Guangzhou, China. TB case data obtained from the Guangzhou Institute of Tuberculosis Control and Prevention. Before using Pearson correlation and a geographical detector (GD) to identify potential influencing factors, we conducted a global spatial autocorrelation analysis to select an appropriate spatial scales. Results Owing to its strong spatial autocorrelation (Moran’s I = 0.33, Z = 4.71), the 2 km × 2 km grid was selected as the spatial scale. At this level, TB incidence was closely associated with most socioeconomic variables (0.31 < r < 0.76, P < 0.01). Of five environmental factors, only the concentration of fine particulate matter displayed significant correlation (r = 0.21, P < 0.05). Similarly, in terms of q values derived from the GD, socioeconomic variables had stronger explanatory abilities (0.08 < q < 0.57) for the spatial differentiation of the 2017 incidence of TB than environmental variables (0.06 < q < 0.27). Moreover, a much larger proportion (0.16 < q < 0.89) of the spatial differentiation was interpreted by pairwise interactions, especially those (0.60 < q < 0.89) related to the 2016 incidence of TB, officially appointed medical institutions, bus stops, and road density. Conclusions The spatial heterogeneity of the 2017 incidence of TB in the study area was considerably influenced by several socioeconomic and environmental factors and their pairwise interactions on a fine scale. We suggest that more attention should be paid to the units with pairwise interacting factors in Guangzhou. Our study provides helpful clues for local authorities implementing more effective intervention measures to reduce TB incidence in China’s municipal areas, which are featured by both a high degree of urbanization and a high incidence of TB. Graphical Abstract

Diseases spectrum in the field of spatiotemporal patterns mining of infectious diseases epidemics: A bibliometric and content analysis

“…In particular, these spatial grids have been gradually considered as the final units where prevention and control measures can produce practical effects in urban regions ( 89 , 97 ). The appropriate grid scale can be chosen by comparing the ability (as reflected by some indicators, such as Moran's I) of a range of regular grids at different scales to characterize the spatiotemporal characteristics of infectious disease epidemics ( 51 ). However, if the resolution of the raw data is limited (e.g., at the county scale), when translating the findings into control measures at a finer scale (e.g., at the township scale), it must be carefully investigated if the findings had similar characteristics at a finer scale.…”

The association between climate variables and tuberculosis in Kolaka District, Southeast Sulawesi Province, Indonesia, 2013–2020: a Bayesian autoregressive model

“…Areas with high rainfall tend to have a higher population density, and high population density can lead to poor environmental conditions. 16,58 Poor environmental conditions, such as high density, lack of ventilation, and poor sanitation, can worsen the spread of tuberculosis. People living in areas with high rainfall may be forced to reside in unhealthy environments, such as overcrowded settlements in urban areas that lack adequate sanitation facilities, which can increase the risk of TB transmission.…”

Investigating Spatial Patterns of Pulmonary Tuberculosis and Main Related Factors in Bandar Lampung, Indonesia Using Geographically Weighted Poisson Regression

“…In the case of spatial epidemiology, future studies can explore spatial clustering methods, e.g., spatial autocorrelation [63], global Moran's I statistics, Kulldorff's scan statistic [64,65], Getis-Ord Gi* [66,67], the generalized linear regression model, and the generalized additive model [68] to analyze the spatial distribution of pulmonary TB. In addition, combining several geospatial techniques with epidemiologically related cases can provide further insight [69].…”