Rising Tuberculosis Presentations in Ontario, Canada Are Not Explained by Community SARS-CoV-2 Activity: A Population-Based Time-Series Study
Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Background Tuberculosis (TB) notifications have increased in several high-income countries since 2020. Proposed explanations include pandemic-related disruption of treatment for latent tuberculosis, changes in population composition, and immune dysfunction following SARS-CoV-2 infection. We assessed whether TB presentations increased during the pandemic era and then tested whether community SARS-CoV-2 activity was followed by increased mycobacterial disease. Methods We conducted a population-based ecological time-series study of first TB- and nontuberculous mycobacterial (NTM)-coded hospital, emergency-department and ambulatory-clinic presentations in the Greater Toronto–Hamilton Area Plus (GTHA+), Ontario (2024 population approximately 11.2 million), from April 2011 through March 2024. NTM served as a comparator because it shares diagnostic pathways with TB but differs epidemiologically. Negative-binomial projection and full-series models estimated pandemic-era changes under three alternative trend assumptions. Distributed lag non-linear models estimated cumulative rate ratios for standardized community viral activity over 0–12 weeks; sensitivity analyses extended the maximum lag to 26 and 52 weeks. Influenza and respiratory syncytial virus (RSV) were prespecified comparators; data for five additional viral exposures were obtained after the primary analysis was completed and analysed as exploratory comparisons. Results We identified 3,985 first TB-coded and 5,214 first NTM-coded acute-care presentations. Across six models, estimated pandemic-era TB increases ranged from 3% to 22%. No model suggested an increased in NTM. SARS-CoV-2 activity was not positively associated with TB (cumulative rate ratio per 1-SD increase over 0–12 weeks 0.909, 95% CI 0.822–1.006) or NTM (0.953, 0.868–1.047). No positive associations were identified in subgroup analyses or with maximum lags of 26 or 52 weeks. Influenza was not associated with either outcome. RSV was associated with TB and NTM; only the RSV–NTM association met the Bonferroni threshold among the prespecified comparisons, and no exploratory associations met the experiment-wide threshold. Conclusions We identified a modest pandemic-era increase in TB-coded acute-care presentations in Ontario, Canada, although its estimated size depended on assumptions about underlying trends. NTM-coded presentations did not increase. Community SARS-CoV-2 activity was not positively associated with subsequent TB or NTM presentations at lags of up to one year and did not provide a temporal explanation for the population-level TB increase. Individual-level effects cannot be excluded.
Source: medRxiv
Institutions: University of Toronto, Public Health Agency of Canada
Language: English
Subfield: Infectious Diseases
Field: Medicine
Domain: Health Sciences
Open Access status: green
Award: G5581477886
OpenAlex ID: W7234977930
