A Drift-Aware and Fairness-Aware Explainable AI Framework for Major Pathogen Transitions

• Fernando H. M de Oliveira UPE
• Cleyton M. de O. Rodrigues UPE

DOI: https://doi.org/10.5753/eniac.2026.30585

Resumo

Distinguishing SARS-CoV-2 from seasonal influenza is challenging due to overlapping symptoms and shifting disease prevalence. To address this problem, we propose a drift-aware and fairness-aware explainable AI framework for reliable clinical triage under non-stationary conditions. Using 21 low-cost triage features and Brazil’s SIVEP-Gripe surveillance data (2023–2025), we evaluated XGBoost-based adaptation strategies during a major etiological inversion. Multi-task Learning with Label Shift prior correction achieved the best performance while improving demographic equity. Combined with calibration and temporal SHAP explanations, the framework provides a transparent, equitable, and cost-effective triage solution for dynamic public health settings.

Referências

Alemi, F., Guralnik, E., Vang, J., Wojtusiak, J., Peterson, R., Roess, A., and Jain, P. (2023). Guidelines for triage of covid-19 patients presenting with multisystemic symptoms. Quality Management in Healthcare, 32(Supplement 1):S3–S10.

Chen, T. and Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 785–794.

Costa, L. F. L., de Mesquita Silva Montenegro, M., Rabello Neto, D. L., de Oliveira, A. T. R., Trindade, J. E. O., Adair, T., and Marinho, M. F. (2020). Estimating completeness of national and subnational death reporting in Brazil: application of record linkage methods. Population Health Metrics, 18(1):22.

de Oliveira, F. H. M. and Rodrigues, C. M. O. (2026). A data-driven and explainable machine learning approach to predict prolonged hospitalization in brazilian sars patients. Revista de Engenharia e Pesquisa Aplicada, 11(1):41–50.

Duckworth, C., Chmiel, F. P., Burns, D. K., Zlatev, Z. D., White, N. M., Daniels, T. W., Kiuber, M., and Boniface, M. J. (2021). Using explainable machine learning to characterise data drift and detect emergent health risks for emergency department admissions during COVID-19. Scientific Reports, 11(1):23017.

Finlayson, S. G., Subbaswamy, A., Singh, K., Bowers, J., Kupke, A., Zittrain, J., Kohane, I. S., and Saria, S. (2021). The clinician and dataset shift in artificial intelligence. New England Journal of Medicine, 385(3):283–286.

Han, Y., Guo, J., Li, X., and Zhong, Z. (2025). Differences in clinical characteristics between coronavirus disease 2019 (COVID-19) and influenza: a systematic review and meta-analysis. NPJ Primary Care Respiratory Medicine, 35(1):8.

Lundberg, S. M. and Lee, S. I. (2017). A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30.

Matthews, B. (1975). Comparison of the predicted and observed secondary structure of T4 phage lysozyme. Biochimica et Biophysica Acta (BBA) - Protein Structure, 405(2):442–451.

Rufibach, K. (2010). Use of brier score to assess binary predictions. Journal of clinical epidemiology, 63(8):938–939.

Viana dos S. Santana, I., Sobrinho, Á., Dias da Silva, L., and Perkusich, A. (2023). Machine learning for COVID-19 and influenza classification during coexisting outbreaks. Applied Sciences, 13(20):11518.

Webster, K. E., Webster, K., Parkhouse, T., Dawson, S., Jones, H. E., Jones, H., Brown, E. L., Brown, E., Hay, A. D., Hay, A., et al. (2024). Diagnostic accuracy of point-of-care tests for acute respiratory infection: a systematic review of reviews. Health Technology Assessment, pages 1–75.

Yanamala, N., Krishna, N. H., Hathaway, Q. A., Radhakrishnan, A., Sunkara, S., Patel, H., Farjo, P., Patel, B., and Sengupta, P. P. (2021). A vital sign-based prediction algorithm for differentiating COVID-19 versus seasonal influenza in hospitalized patients. NPJ Digital Medicine, 4(1):95.

Zhou, X., Wang, Z., Li, S., Liu, T., Wang, X., Xia, J., and Zhao, Y. (2021). Machine learning-based decision model to distinguish between COVID-19 and influenza: a retrospective, two-centered, diagnostic study. Risk Management and Healthcare Policy, pages 595–604.

Zhu, Q. (2020). On the performance of Matthews correlation coefficient (MCC) for imbalanced dataset. Pattern Recognition Letters, 136:71–80.

Zoabi, Y., Deri-Rozov, S., and Shomron, N. (2021). Machine learning-based prediction of COVID-19 diagnosis based on symptoms. NPJ Digital Medicine, 4(1):3.