Interpretable Early Detection of Adverse Drug Reactions: Integrating Robust Anomaly Scoring with Temporal Lag Analysis and Causal Verification

Article Sidebar

PDF
Published: 2026-02-11
Keywords:
adverse drug reaction detection, explainable artificial intelligence, causal inference, pharmacovigilance, temporal analysis

Main Article Content

Yisi Liu
Business Data Analytics & Human Resources Management, Loyola University Chicago, Illinois, USA

Abstract

Adverse drug reactions impose substantial clinical burdens, yet conventional pharmacovigilance approaches suffer from prolonged detection latencies and elevated false positive rates. This paper presents an integrated framework combining robust statistical anomaly scoring, temporal lag pattern mining, and causal inference methodologies to enable early ADR identification while maintaining interpretability. The three-stage pipeline employs median absolute deviation-based robust scoring, applies Cumulative Sum Control Charts for temporal pattern analysis, and utilizes propensity score matching with Bradford Hill criteria for causal verification. SHAP-based feature attribution generates clinically actionable evidence chains. Validation on FAERS and VigiBase datasets demonstrates 87% AUC with 73% sensitivity at 89% specificity, achieving a median time-to-signal of 4.2 months. The framework exhibits robust generalization across demographic subgroups, establishing a paradigm for trustworthy AI deployment in pharmacovigilance applications.

Article Details

Issue

Vol. 1 No. 1 (2025): 2025 International Conference on Intelligent Computing and Automated Systems (ICAS 2025)

Section

Articles

How to Cite

Interpretable Early Detection of Adverse Drug Reactions: Integrating Robust Anomaly Scoring with Temporal Lag Analysis and Causal Verification. (2026). Journal of Science, Innovation & Social Impact, 1(1), 471-484. https://sagespress.com/index.php/JSISI/article/view/75

References

1. Y. Zhao, Y. Yu, H. Wang, Y. Li, Y. Deng, G. Jiang, and Y. Luo, "Machine learning in causal inference: Application in pharmacovigilance," Drug Safety, vol. 45, no. 5, pp. 459-476, 2022. doi: 10.1007/s40264-022-01155-6

2. S. Lee, S. Kim, J. Lee, J. Y. Kim, M. H. Song, and S. Lee, "Explainable artificial intelligence for patient safety: A review of application in pharmacovigilance," IEEE Access, vol. 11, pp. 50830-50840, 2023.

3. Z. Dong and F. Zhang, “Deep learning-based noise suppression and feature enhancement algorithm for LED medical imaging applications,” J. Sci., Innov. Soc. Impact, vol. 1, no. 1, pp. 9–18, 2025.

4. S. Singh, R. Kumar, S. Payra, and S. K. Singh, "Artificial intelligence and machine learning in pharmacological research: Bridging the gap between data and drug discovery," Cureus, vol. 15, no. 8, 2023. doi: 10.7759/cureus.44359

5. H. Shin, and S. Lee, "An OMOP-CDM based pharmacovigilance data-processing pipeline (PDP) providing active surveillance for ADR signal detection from real-world data sources," BMC Medical Informatics and Decision Making, vol. 21, no. 1, p. 159, 2021. doi: 10.1186/s12911-021-01520-y

6. R. I. Silva, J. Maro, and M. Kulldorff, "Exact sequential test for clinical trials and post-market drug and vaccine safety surveillance with Poisson and binary data," Statistics in Medicine, vol. 40, no. 22, pp. 4890-4913, 2021.

7. Y. Zhu, R. A. Hubbard, J. Chubak, J. Roy, and N. Mitra, "Core concepts in pharmacoepidemiology: Violations of the positivity assumption in the causal analysis of observational data: Consequences and statistical approaches," Pharmacoepidemiology and Drug Safety, vol. 30, no. 11, pp. 1471-1485, 2021. doi: 10.1002/pds.5338

8. F. Di Martino, and F. Delmastro, "Explainable AI for clinical and remote health applications: A survey on tabular and time series data," Artificial Intelligence Review, vol. 56, no. 6, pp. 5261-5315, 2023. doi: 10.1007/s10462-022-10304-3

9. E. Bresso, P. Monnin, C. Bousquet, and F. Calvier, "É," , Ndiaye, N. C., Petitpain, N., ... & Coulet, A. (2021). Investigating ADR mechanisms with explainable AI: A feasibility study with knowledge graph mining. BMC Medical Informatics and Decision Making, vol. 21, no. 1, p. 171, 2021.

10. V. Dabas, A. Thomas, P. Khatri, F. Iandolo, and A. Usai, "Decrypting disruptive technologies: Review and research agenda of explainable AI as a game changer," In 2023 IEEE International Conference on Technology Management, Operations and Decisions (ICTMOD), November, 2023, pp. 1-6. doi: 10.1109/ictmod59086.2023.10438156

11. G. Abgrall, A. L. Holder, Z. Chelly Dagdia, K. Zeitouni, and X. Monnet, "Should AI models be explainable to clinicians? Critical Care, 28(1), 301," 2024.

12. Z. Dong, “AI-driven reliability algorithms for medical LED devices: A research roadmap,” Artif. Intell. Mach. Learn. Rev., vol. 5, no. 2, pp. 54–63, 2024.

13. Y. Wang, J. Ma, S. Ma, J. Wang, and J. Li, "Causal evaluation of post-marketing drugs for drug-induced liver injury from electronic health records," In 2023 45th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), July, 2023, pp. 1-4. doi: 10.1109/embc40787.2023.10340721

14. G. Candore, K. Hedenmalm, J. Slattery, A. Cave, X. Kurz, and P. Arlett, "Can we rely on results from IQVIA medical research data UK converted to the observational medical outcome partnership common data model? A validation study based on prescribing codeine in children," Clinical Pharmacology & Therapeutics, vol. 107, no. 4, pp. 915-925, 2020. doi: 10.1002/cpt.1785

15. A. Suh, G. Appleby, E. W. Anderson, L. Finelli, and D. Cashman, "Communicating performance of regression models using visualization in pharmacovigilance," In 2021 IEEE Workshop on Visual Analytics in Healthcare (VAHC), October, 2021, pp. 6-13. doi: 10.1109/vahc53616.2021.00006

16. H. Sholehrasa, X. Xu, D. Caragea, J. E. Riviere, and M. Jaberi-Douraki, "Predictive modeling and explainable AI for veterinary safety profiles, residue assessment, and health outcomes using real-world data and physicochemical properties," arXiv preprint arXiv:2510.01520, 2025.

17. A. Kumar, J. P. Singh, and A. K. Singh, "Explainable BERT-LSTM stacking for sentiment analysis of COVID-19 vaccination," IEEE Transactions on Computational Social Systems, 2023.