Statistical Models Addressing Acute Respiratory Diseases: A Scoping Review

This review highlights the application of statistical models in addressing respiratory diseases worldwide. A scoping review was conducted following Joanna Briggs Institute methodology, with searches in Medline (PubMed), Web of Science, LILACS, BVS, CINAHL, EBSCO, and EMBASE between August and September 2024, updated in June 2025. Descriptors and synonyms were drawn from DeCS and MeSH, with no language restrictions: respiratory tract diseases, statistical models, time factors, and health management. Included materials encompassed articles, theses, dissertations, and technical reports published between 2014 and 2025, while protocols, editorials, systematic reviews, and studies unrelated to the research question were excluded. Eligible studies applied statistical models to healthcare demand for acute respiratory diseases. The final sample comprised 29 studies, grouped into three categories: (1) demand forecasting, emphasizing SEIR models for estimating pandemic evolution, predicting cases and deaths, and guiding resource allocation; (2) epidemiological analysis, using compartmental, time series, and Bayesian models to infer transmission patterns and evaluate control measures; and (3) risk and prognosis, employing machine learning models for clinical prediction and individual risk stratification. Findings highlight the importance of combining methods, leveraging multiple data sources, and adapting models to local contexts. Integrating traditional and modern models using local data is essential for effective surveillance, prognosis, and decision-making in dynamic scenarios.