Spatiotemporal Analysis of the Association Between Community Violence and Tuberculosis and Treatment Outcomes in Fortaleza, Brazil, 2014-2023.

BACKGROUND

Violence may be associated with tuberculosis (TB) burden through multiple pathways. We examined associations between homicide rates (a proxy for neighborhood violence) and TB-related outcomes in 114 bairros in Fortaleza, Brazil, from 2014 to 2023.

METHODS

We analyzed homicide rates, incidence of newly reported TB, and rates of TB with unsuccessful treatment outcomes. We assessed spatial autocorrelation and clusters using Moran's I and space-time scan statistics. Geographically weighted regression (GWR) and Bayesian spatial models assessed spatial and spatiotemporal associations.

RESULTS

Annual homicide rates ranged from 26.8-81.1 per 100,000. Newly reported TB incidence was 50-60 per 100,000 during 2015-2021 and 66.8 in 2023, while rate of TB with unsuccessful outcomes increased from 10.8 to 17.5. GWR models showed positive associations between cumulative homicide rates and cumulative incidence of newly reported TB (β>0.5, p<0.05) and rates of TB with unsuccessful outcomes (0<β<0.5, p<0.05) across all bairros. Bayesian models showed associations per one-SD increase in homicide rate with newly reported TB (RR=1.05, 95% CrI:1.02-1.08) and TB with unsuccessful outcomes (RR=1.11, 95% CrI:1.05-1.18).

CONCLUSIONS

Higher homicide rates were spatially and spatiotemporally associated with higher TB burden. Violence may serve as an indicator of neighborhoods with elevated TB burden.