Influence of Sex and Sex-Based Disparities on Prevalent Tuberculosis, Vietnam, 2017-2018

Abstract

To assess sex disparities in tuberculosis in Vietnam, we conducted a nested, case-control study based on a 2017 tuberculosis prevalence survey. We defined the case group as all survey participants with laboratory-confirmed tuberculosis and the control group as a randomly selected group of participants with no tuberculosis. We used structural equation modeling to describe pathways from sex to tuberculosis according to an a priori conceptual framework. Our analysis included 1,319 participants, of whom 250 were case-patients. We found that sex was directly associated with tuberculosis prevalence (adjusted odds ratio for men compared with women 3.0 [95% CI 1.7-5.0]) and indirectly associated through other domains. The strong sex difference in tuberculosis prevalence is explained by a complex interplay of factors relating to behavioral and environmental risks, access to healthcare, and clinical manifestations. However, after controlling for all those factors, a direct sex effect remains that might be caused by biological factors.

Keywords: Vietnam; bacteria; epidemiology; nested case-control study; sex-based disparities; tuberculosis and other mycobacteria.

Figure 1

Conceptual framework of the structural equation model to describe pathways between tuberculosis prevalence, sex, and the associated domains for case–control analysis of tuberculosis prevalence, Vietnam, 2017–2018.

Figure 2

Summary of the data flow for case–control analysis of tuberculosis prevalence, Vietnam, 2017–2018.

Figure 3

Structural equation model of the relationships between domains and Xpert-positive (Xpert MTB/Rif; Cepeheid, https://www.cepheid.com) tuberculosis prevalence for case–control analysis of tuberculosis prevalence, Vietnam, 2017–2018. For significant associations, the arrow thickness corresponds to the effect size. Each outcome was adjusted for age, area, and region. See Table 3 for the full estimation results. Model results were weighted using sampling and lost-to-follow-up weights. Bootstrapped area under the curve (1,000 replications) was 0.90 (95% CI 0.89–0.92).