Decision Framework Focused on Missing-Data Alarms in Healthcare

Authors

DOI:

https://doi.org/10.3991/ijoe.v22i07.61659

Keywords:

missing data, risk-targeted calibration, EVSI, decision support, glucose alarms, healthcare time series

Abstract


In healthcare, missing data is common. Missing information can arise for various reasons, such as timing and synchronization issues, incorrect reporting, and connection loss or incorrect data-flow configuration. Time plays a crucial role in longitudinal clinical data. Many studies have focused on minimizing the root mean squared error (RMSE) of imputed missing data. However, in clinical practice, the primary goal is to make safe decisions by minimizing the risk of missing key events. The goal is not to reconstruct accurate missing data but to determine if the existing data is sufficient to support the detection of critical events. The predictive model combines a probabilistic event definition, empirical risk, calibrated thresholds for risk-targeted threshold calibration, a lightweight counterfactual glucose sampler, and the expected value of sample information (EVSI). The EVSI policy does not always request a confirmatory measurement. It is used only when the expected reduction in decision burden is higher than the additional delay. Results show an improved balance of risk and burden when compared to both non-query scenarios and simpler baseline querying strategies. The contribution of this work is not a new state-of-the-art data imputation system but a clinically interpretable decision layer for alarm management under missing data scenarios.

Author Biographies

Elma Zanaj, Polytechnic University of Tirana, Tirana, Albania

Elma Zanaj received M.S. degree in Telecommunications Engineering from Polytechnic University of Tirana in 2003, and Ph.D. degree in Electronics, Informatics and Telecommunications Engineering from Marche Polytechnic University, Ancona, Italy, in 2008. She is currently a Full Professor at the Polytechnic University of Tirana, Albania, with research focusing on WSNs applications, IoT and the use of AI in healthcare. (E-mail: ezanaj@fti.edu.al).

Lorena Balliu, Polytechnic University of Tirana, Tirana, Albania

Lorena Balliu is a full-time Assistant Lecturer at the Department of Computer Science Fundamentals, Polytechnic University of Tirana. She is currently pursuing a Ph.D. at the Department of Electronics and Telecommunications. Her research interests include Artificial Intelligence, Machine Learning algorithms, and their applications in telemedicine and data analysis, supported by both academic and private sector experience. (E-mail: lorena.balliu@fti.edu.al).

Anita Xhemali, Polytechnic University of Tirana, Tirana, Albania

Anita Xhemali obtained her M.Sc. degree in Computer Science from the Polytechnic University of Tirana in 2005. She is currently a Ph.D. candidate at the Faculty of Information Technology. She is currently a full-time Lecturer at the Faculty of Electrical Engineering. Her research explores the integration of fuzzy logic systems with intelligent IoT architecture and advanced healthcare developing efficient computational approaches. (E-mail: anita.xhemali@fti.edu.al).

Gledis Basha, Polytechnic University of Tirana, Tirana, Albania

Gledis Basha obtained his B.Sc. (2012) and M.Sc. (2014) degrees from the Faculty of Information Technology, Polytechnic University of Tirana (UPT). He is currently a full-time Assistant Lecturer and a Ph.D. candidate at the Faculty of Information Technology. His research focuses on information technology systems and healthcare technologies, particularly their practical applications in engineering and industry. (E-mail: gbasha@fti.edu.al).

References

[1] Du, W., Côté, D., & Liu, Y. (2023). SAITS: Self-Attention-based Imputation for Time Series. Expert Systems with Applications, 219, 119619. https://doi.org/10.1016/j.eswa.2023.119619

[2] Le, L. P., Nguyen Thi, X.-H., Nguyen, T., Riegler, M. A., Halvorsen, P., & Nguyen, B. T. (2024). Missing data imputation for noisy time-series data and applications in healthcare. arXiv. https://doi.org/10.48550/arXiv.2412.11164

[3] Cao, W., Wang, D., Li, J., Zhou, H., Li, L., & Li, Y. (2018). BRITS: Bidirectional Recurrent Imputation for Time Series. Advances in Neural Information Processing Systems, 31. https://doi.org/10.48550/arXiv.1805.10572

[4] Tashiro, Y., Song, J., Song, Y., & Ermon, S. (2021). CSDI: Conditional Score-based Diffusion Models for Probabilistic Time Series Imputation. Advances in Neural Information Processing Systems, 34. https://doi.org/10.48550/arXiv.2107.03502

[5] Qian, L., et al. (2025). CSAI: Conditional Self-Attention Imputation for Healthcare Time-series. IEEE Journal of Biomedical and Health Informatics. https://doi.org/10.1109/JBHI.2025.3648181

[6] Liu, M., et al. (2023). Handling Missing Values in Healthcare Data: A Systematic Review of Deep Learning-based Imputation Techniques. Artificial Intelligence in Medicine, 142, 102587. https://doi.org/10.1016/j.artmed.2023.102587

[7] Du, W., et al. (2024). TSI-Bench: Benchmarking Time Series Imputation. arXiv. https://doi.org/10.48550/arXiv.2406.12747

[8] Qian, L., et al. (2024). Beyond Random Missingness: Clinically Rethinking for Healthcare Time Series Imputation. arXiv. https://doi.org/10.48550/arXiv.2405.17508

[9] Wang, Z., Yang, L., Sun, L., Wen, Q., & Wang, Y. (2024). Task-oriented Time Series Imputation Evaluation via Generalized Representers. Advances in Neural Information Processing Systems, 37, pp. 137403–137431. https://doi.org/10.52202/079017-4365

[10] Raj, J. A., Qian, L., & Ibrahim, Z. (2024). Modular Deep Learning for Multivariate Time-Series: Decoupling Imputation and Downstream Tasks. arXiv. https://doi.org/10.48550/arXiv.2411.03941

[11] Poleto, F. Z., Singer, J. M., & Paulino, C. D. (2011). Missing data mechanisms and their implications on the analysis of categorical data. Statistics and Computing, 21(1), pp. 31–43. https://doi.org/10.1007/s11222-009-9143-x

[12] Alharthi, A. M., et al. (2022). Improving Penalized Logistic Regression Model with Missing Values in High-Dimensional Data. International Journal of Online and Biomedical Engineering (iJOE), 18(02), pp. 40–54. https://doi.org/10.3991/ijoe.v18i02.25047

[13] Angelopoulos, A. N., Bates, S., Candès, E. J., Jordan, M. I., & Lei, L. (2025). Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control. Annals of Applied Statistics, 19(2), pp. 1641–1662. https://doi.org/10.1214/24-AOAS1998

[14] Angelopoulos, A. N., Bates, S., Fisch, A., Lei, L., & Schuster, T. (2024). Conformal Risk Control. International Conference on Learning Representations. https://doi.org/10.48550/arXiv.2208.02814

[15] Qin, Y., van der Schaar, M., & Lee, C. (2023). Risk-Averse Active Sensing for Timely Outcome Prediction under Cost Pressure. Advances in Neural Information Processing Systems, 36, pp. 6397–6411. https://doi.org/10.52202/075280-0279

[16] Tran, C., & Fioretto, F. (2023). Data Minimization at Inference Time. Advances in Neural Information Processing Systems, 36. https://doi.org/10.48550/arXiv.2305.17593

[17] Steuten, L. M. G., et al. (2013). A Systematic and Critical Review of the Evolving Methods and Applications of Value of Information in Academia and Practice. PharmacoEconomics, 31(1), pp. 25–48. https://doi.org/10.1007/s40273-012-0008-3

[18] Heath, A., Kunst, N., & Jackson, C. (Eds.). (2024). Value of Information for Healthcare Decision-Making. Chapman and Hall/CRC. https://doi.org/10.1201/9781003156109

[19] Mohammed, B. G., & Hasan, D. S. (2023). Smart Healthcare Monitoring System Using IoT. International Journal of Interactive Mobile Technologies (iJIM), 17(01), pp. 141–152. https://doi.org/10.3991/ijim.v17i01.34675

[20] Ocares-Cunyarachi, L., & Andrade-Arenas, L. (2023). Design of a Mobile App to Monitor and Control in Real Time Type 2 Diabetes Mellitus in Peru. International Journal of Interactive Mobile Technologies (iJIM), 17(10), pp. 176–192. https://doi.org/10.3991/ijim.v17i10.38207

[21] Al-Zoubi, A. Y., Tahat, A., Wahsheh, R. A., Taha, M., Al-Tarawneh, L., & Hasan, O. (2022). A Bachelor Degree Program in IoT Engineering: Accreditation Constraints and Market Demand. International Journal of Engineering Pedagogy (iJEP), 12(4), pp. 17–34. https://doi.org/10.3991/ijep.v12i4.31429

[22] Zanaj, E., Balliu, L., Basha, G., & Gjata, E. (2024). Studying the Behavior of a Modified Deep Learning Model for Disease Detection Through X-ray Chest Images. International Journal of Advanced Computer Science and Applications (IJACSA), 15(5). https://doi.org/10.14569/IJACSA.2024.0150585

[23] Balliu, L., et al. (2024). Enhancing heart disease prediction accuracy by comparing classification models employing varied feature selection techniques. Serbian Journal of Electrical Engineering, 21(3), pp. 375–390. https://doi.org/10.2298/SJEE2403375B

[24] Xhemali, A., Zanaj, E., Basha, G., & Balliu, L. (2025). Enhancing Energy Efficiency Prediction in Assisted Living Through GA-FIS, PSO-FIS, and NSGA-II-FIS: A Comparative Evaluation. Automation, Robotics & Communications for Industry 4.0/5.0, ARCI 2025, p. 62.

[25] Harrell, F. E. (2015). Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis (2nd ed.). Springer. https://doi.org/10.1007/978-3-319-19425-7

[26] Hakkal, S., et al. (2024). XGBoost to Enhance Learner Performance Prediction. Computers and Education: Artificial Intelligence, 7, 100254. https://doi.org/10.1016/j.caeai.2024.100254

[27] Kassem, K., et al. (2024). An innovative artificial intelligence-based method to compress complex models into explainable, model-agnostic and reduced decision support systems with application to healthcare (NEAR). Artificial Intelligence in Medicine, 151, 102841. https://doi.org/10.1016/j.artmed.2024.102841

[28] Sadatsafavi, M., et al. (2025). Expected Value of Sample Information Calculations for Risk Prediction Model Validation. Medical Decision Making, 45(3), pp. 232–244. https://doi.org/10.1177/0272989X251314010

[29] Heath, A., Manolopoulou, I., & Baio, G. (2018). Efficient Monte Carlo Estimation of the Expected Value of Sample Information Using Moment Matching. Medical Decision Making, 38(2), pp. 163–173. https://doi.org/10.1177/0272989X17738515

[30] Ades, A. E., Lu, G., & Claxton, K. (2004). Expected value of sample information calculations in medical decision modeling. Medical Decision Making, 24(2), pp. 207–227. https://doi.org/10.1177/0272989X04263162

[31] Hidalgo, J. I., Alvarado, J., Botella, M., Aramendi, A., Velasco, J. M., & Garnica, O. (2024). HUPA-UCM Diabetes Dataset [Data set]. Mendeley Data, V1. https://doi.org/10.17632/3hbcscwz44.1

[32] Okoye, K., & Hosseini, S. (2024). Wilcoxon Statistics in R: Signed-Rank Test and Rank-Sum Test. In R Programming: Statistical Data Analysis in Research (pp. 279–303). Springer Nature Singapore. https://doi.org/10.1007/978-981-97-3385-9_13

[33] Gilad-Bachrach, R., Navot, A., & Tishby, N. (2005). Query by Committee Made Real. Advances in Neural Information Processing Systems, 18, pp. 443–450. https://doi.org/10.5555/2976248.2976304

Cover iJOE – Vol 22, No 07 (2026)

Downloads

Published

2026-07-16

How to Cite

Zanaj, E., Balliu, L., Xhemali, A., & Basha, G. (2026). Decision Framework Focused on Missing-Data Alarms in Healthcare. International Journal of Online and Biomedical Engineering (iJOE), 22(07), pp. 207–223. https://doi.org/10.3991/ijoe.v22i07.61659

Issue

Vol. 22 No. 07 (2026)

Section

Papers

License

Copyright (c) 2026 Elma Zanaj, Lorena Balliu, Anita Xhemali, Gledis Basha

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.