Long-term impact of Covid-19-related nonpharmaceutical interventions on tuberculosis

An interrupted time series analysis using Bayesian method

Bibliographic Data

ID19552145
AuthorsYongbin Wang (0000-0001-7854-7020, Xinxiang Medical University), Yue Xi (0000-0001-6150-4382, Xinxiang Medical University), Yanyan Li (0009-0000-0010-2654, Xinxiang Medical University), Peiping Zhou (0009-0009-2369-0319, Xinxiang Medical University), Chunjie Xu (0000-0002-7666-5573, Chinese Academy of Medical Sciences & Peking Union Medical College)
Year2025
Volume15
Pages04012-04012
Publication date2025-01-24
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenueJournal of Global Health (JOURNAL)
Journal identifiersISSN: 2047-2978 • E-ISSN: 2047-2986
PublisherInternational Society of Global Health (PUBLISHER • GB)
DOI10.7189/jogh.15.04012
PMID39849949
OpenAlexW4406808002
LanguageEN
Citations received2
References cited34

Background: The implementation of non-pharmaceutical interventions (NPIs) during the COVID-19 pandemic may inadvertently influence the epidemiology of tuberculosis (TB). (TB). However, few studies have explored how NPIs impact the long-term epidemiological trends of TB. We aimed to estimate the impact of NPIs implemented against COVID-19 on the medium- and long-term TB epidemics and to forecast the epidemiological trend of TB in Henan. Methods: We first collected monthly TB case data from January 2013 to September 2022, after which we used the data from January 2013 to December 2021 as a training data set to fit the Bayesian structural time series (BSTS) model and the remaining data as a testing data set to validate the model's predictive accuracy. We then conducted an intervention analysis using the BSTS model to evaluate the impact of the COVID-19 pandemic on TB epidemics and to project trends for the upcoming years. Results: A total of 590 455 TB cases were notified from January 2013 to September 2022, resulting in an annual incidence rate of 57.4 cases per 100 000 population, with a monthly average of 5047 cases (5.35 cases per 100 000 population). The trend in TB incidence showed a significant decrease during the study period, with an annual average percentage change of -7.3% (95% confidence interval (CI) = -8.4, -6.1). The BSTS model indicated an average monthly reduction of 25% (95% CI = 17, 32) in TB case notifications from January 2020 to December 2021 due to COVID-19 (probability of causal effect = 99.80%, P = 0.002). The mean absolute percentage error in the forecast set was 14.86%, indicating relatively high predictive accuracy of the model. Furthermore, TB cases were projected to total 43 584 (95% CI = 29 471, 57 291) from October 2022 to December 2023, indicating a continued downward trend. Conclusions: COVID-19 has had medium- and long-term impacts on TB epidemics, while the overall trend of TB incidence in Henan is generally declining. The BSTS model can be an effective option for accurately predicting the epidemic patterns of TB, and its results can provide valuable technical support for the development of prevention and control strategies

2019-20 coronavirus outbreak · Bayesian probability · Biology · Interrupted time series · Interrupted Time Series Analysis · Pandemic · Pathology · Psychological intervention · Statistics · Time series · Tuberculosis · Computer Science · Healthcare Facilities Design and Sustainability · Immune responses and vaccinations · Mathematics · Medicine · Tuberculosis Research and Epidemiology · Internal Medicine · Virology

Tools

Open DOIOpen Access