COMPARISON OF MULTIPLE ARTIFICIAL INTELLIGENCE-MODELS’ PREDICTIVE POWER IN DETECTING DELAYED ANALGESIA IN EMERGENCY DEPARTMENT PATIENTS
DOI:
https://doi.org/10.69656/pjp.v22i1.1931Keywords:
Acute pain management, Artificial Intelligence, Delayed analgesia, Emergency medicine, Machine Learning, Random Forest, TriageAbstract
Background: Artificial intelligence (AI) has significant potential to enhance risk assessment by identifying patients at higher risk of delayed analgesia. The goal of this work was to create and validate AI models that predict the probability of delayed analgesia, and compare the predictive power of multiple AI models in detecting delayed analgesia in emergency department (ED) patients and avoiding longer patient stays. Methods: From Dec 2024 to Jan 2025, 300 adult patients with moderate to severe pain were studied in the Emergency Department of an academic facility teaching hospital. This retrospective observational study was collected and analysed retrospectively, with age, gender, triage category, triage pain score, and presentation during peak hours serving as input features. Five machine learning models were constructed and compared for their accuracy to forecast delayed analgesia. Important predictors were identified using SHAP (SHapley Additive exPlanations) values for the AI model with the highest accuracy. Results: Random Forest and J48 achieved 77% accuracy, with Random Forest having greater recall for delayed cases (Precision=0.71, Recall=0.84) for anticipating delayed analgesia. Naive Bayes and Logistic Regression had low recall for delayed cases, while MLP Neural Network demonstrated moderate predictive usefulness. Random Forest model had the best performance with the highest AUC [0.83 (95% CI: 0.75–0.90)] on ROC analysis. Conclusion: AI models were successfully implemented, Random Forest outperforming the others, for early identification of patients at risk of delayed analgesia. AI models can decrease unnecessary delays and improve pain management in ED.
Pak J Physiol 2026;22(1):52–7, DOI: https://doi.org/10.69656/pjp.v22i1.1931
Downloads
References
Paul Hunt JF. Management of acute pain in adults in the Emergency Department: Summary of recommendations. Royal College of Emergency Medicine; 2024.
Pines JM, Shofer FS, Isserman JA, Abbuhl SB, Mills AM. The effect of emergency department crowding on analgesia in patients with back pain in two hospitals. Acad Emerg Med 2010;17(3):276–83. doi:10.1111/j.1553-2712.2009.00676.x
Arendts G, Fry M. Factors associated with delay to opiate analgesia in Emergency Departments. J Pain 2006;7(9):682–6. doi:10.1016/j.jpain.2006.03.003.
Gabriel RA, Simpson S, Zhong W, Burton BN, Mehdipour S, Said ET. A neural network model using pain score patterns to predict the need for outpatient opioid refills following ambulatory surgery: algorithm development and validation. JMIR Perioper Med 2023;6:e40455. https://periop.jmir.org/2023/1/e40455/
Nair AA, Velagapudi MA, Lang JA, Behara L, Venigandla R, Velagapudi N, et al. Machine learning approach to predict postoperative opioid requirements in ambulatory surgery patients. PLoS One 2020;15(7):e0236833.
Matsangidou M, Liampas A, Pittara M, Pattichi CS, Zis P. Machine learning in pain medicine: An up-to-date systematic review. Pain Ther 2021;10(2):1067–84. doi:10.1007/s40122-021-00324-2
Lv S, Sun N, Hao C, Li J, Li Y. Development and validation of machine learning models for predicting post-cesarean pain and individualized pain management strategies: a multicenter study. BMC Anesthesiol 2025;25(1):170. doi:10.1186/s12871-025-03034-w
Tan CW, Koh JZ, Jin H, Han NR, Cheng SM, Ta AWA, et al. Machine learning approach to predict postoperative pain after spinal morphine administration during caesarean delivery. Heliyon 2024;10(23):e40602. doi:10.1016/j.heliyon.2024.e40602
Emam OS, Eldaly AS, Avila FR, Torres-Guzman RA, Maita KC, Garcia JP, et al. Machine learning algorithms predict long-term postoperative opioid misuse: A systematic review. Am Surg 2024;90(1):140–51. doi:10.1177/00031348231198112
Okada Y, Ning Y, Ong MEH. Explainable artificial intelligence in emergency medicine: an overview. Clin Exp Emerg Med 2023;10(4):354–62. doi:10.15441/ceem.23.145
Rampanjato RM, Florence M, Patrick NC, Finucane BT. Factors influencing pain management by nurses in emergency departments in Central Africa. Emerg Med J 2007;24(7):475–6. doi:10.1136/emj.2006.045815
Wang L, Song C, Bai Y, Huang X, Shi H, Pan J. Practice and reflection on the management mode of pain quality control in emergency pre-check and triage. Ann Palliat Med 2020;9(4):1879–85. doi:10.21037/apm-20-1108
Chen YW, Lee JH, Chiang CY, Yeh YN, Lin JC, Tsai MJ. Factors associated with delayed order-to-administration time in the emergency department: a retrospective analysis. BMC Emerg Med 2025;25(1):74. doi:10.1186/s12873-025-01229-5
Lautenbacher S, Peters JH, Heesen M, Scheel J, Kunz M. Age changes in pain perception: A systematic-review and meta-analysis of age effects on pain and tolerance thresholds. Neurosci Biobehav Rev 2017;75:104–13. doi:10.1016/j.neubiorev.2017.01.039
Downie WW, Leatham PA, Rhind VM, Wright V, Branco JA, Anderson JA. Studies with pain rating scales. Ann Rheum Dis 1978;37(4):378–81. doi:10.1136/ard.37.4.378
Medicine E. Index, Version 4. AHRQ Pub. No. 05-0046-2 May 2005; 2005. http://www.ahrq.gov/research/esi/
Ganjali R, Golmakani R, Ebrahimi M, Eslami S, Bolvardi E. Accuracy of the emergency department triage system using the emergency severity index for predicting patient outcome; A single center experience. Bull Emerg Trauma 2020;8(2):115–20. doi:10.30476/BEAT.2020.46452
Corp. I. IBM SPSS Statistics for Windows, Version 25.0.; 2017.
Foundation PS. Python Language Reference, version 3.8. https://www.python.org/
Llc G. Google Colaboratory. https://colab.research.google.com/?
Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP. SMOTE: Synthetic Minority Over-sampling Technique. J Artif Intell Res 2002;16:321–57. doi:10.1613/JAIR.953
Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, et al. Scikit-learn: Machine learning in python. J Mach Learn Res 2011(12):2825–30. https://www.jmlr.org/papers/volume12/ pedregosa11a/pedregosa11a.pdf?source=post_page
Lematre G, Nogueira F, Aridas CK. Imbalanced-learn: A python toolbox to tackle the curse of imbalanced datasets in machine learning. J Mach Learn Res 2017;18:1–5. https://www.jmlr.org/ papers/volume18/16-365/16-365.pdf
Lundberg S, Lee SI. A unified approach to interpreting model predictions. arXiv; 2017. URL: https://arxiv.org/abs/1705.07874 doi:10.48550/ARXIV.1705.07874
Bloom B, Fritz CL,?Gupta S, Pott J, Sjene I, Astin-Chamberlain R, et al. Older age and risk for delayed abdominal pain care in the emergency department? Eur J Emerg Med 2024;31(5):332–8.
Platts-Mills TF, Hunold KM, Weaver MA, Dickey RM, Fernandez AR, Fillingim RB, et al. Pain treatment for older adults during prehospital emergency care: Variations by patient gender and pain severity. J Pain 2013;14(9):966–74. doi:10.1016/j.jpain.2013.03.014
Anderson AB, Grazal CF, Balazs GC, Potter BK, Dickens JF, Forsberg JA. Can predictive modeling tools identify patients at high risk of prolonged opioid use after ACL reconstruction? Clin Orthop Relat Res 2020;478(7):00-1618. doi:10.1097/ CORR.0000000000001251
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Ayesha Saeed, Muhammad Moeed Azwar Bhatti, Zubia Razzaq, Zainub Saeed
This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.
The author(s) retain the copyrights and allow their publication in Pakistan Journal of Physiology, Pak J Physiol, PJP to be FREE for research and academic purposes. It can be downloaded and stored, printed, presented, projected, cited and quoted with full reference of, and acknowledgement to the author(s) and the PJP. The contents are published with an international CC-BY-ND-4.0 License.
