Development of an Automated ICD Coding System for Outpatients at Songklanagarind Hospital: A Retrospective Study
Development of a Modified Work Ability Index for Nurses
DOI:
https://doi.org/10.31584/psumj.2026281538Keywords:
automated ICD, knowledge base, outpatients, rule-basedAbstract
Objective: To develop and evaluate the performance of an automatic disease and procedure coding system for outpatients at Songklanagarind Hospital using rule-based techniques. This system aims to address the problem of payment rejections from health funds due to errors in disease and procedure coding.
Material and Methods: This retrospective study utilized 49,497 outpatient visit summaries from Songklanagarind Hospital in 2023 that did not meet the payment conditions set by health funds, as evaluated by the hospital information system (HIS). Data were divided into a training dataset (24,748 records) and an evaluation dataset (24,749 records). The automatic coding system was developed using rule-based techniques, comprising disease code databases, procedure code databases, a rule repository, and a matching rules engine. System performance was assessed using Spearman’s rank correlation coefficient, confusion matrix, accuracy, sensitivity, false positive rate (FPR), and area under the curve (AUC).
Results: The system achieved an accuracy of 84.42%, a sensitivity of 65.53%, and an AUC of 81.74%. Spearman’s rho was 0.8517, indicating high concordance between codes assigned by the automatic system and those assigned by experts. The system was able to assign 57,870 codes out of a total of 117,295 codes, To ensure the record is not rejected or subjected to reimbursement delays exceeding 50% of the total medical expenses (specifically for cases that do not meet the payment criteria of the health insurance fund, as assessed by the hospital information system [HIS]).
Conclusion: The rule-based automated disease and procedure coding system demonstrates high performance, effectively reducing workload, increasing efficiency, and minimizing coding errors. It is suitable as an assistive tool for coding outpatient cases, particularly when processing large volumes of data.
References
Sitaru S, Nhan F, Gasteiger C, Rueckert D, Biedermann T, Zink A. Digitalising the past decades: automated ICD-10 coding of unstructured free text dermatological diagnoses. BMC Health Serv Res 2024;24:1297.
Chen PF, Wang SM, Liao WC, Kuo LC, Chen KC, Lin YC, et al. Automatic ICD-10 Coding and Training System: Deep Neural Network Based on Supervised Learning. JMIR Medical Informatics 2021;9:e23230.
Venkatesh KP, Raza MM, Kvedar JC. Automating the overburdened clinical coding system: challenges and next steps. NPJ Digit Med 2023;.6:16.
Chen PF, Chen KC, Liao WC, Lai F, He TL, Lin SC, et al. Automatic international classification of diseases coding system: deep contextualized language model with rule based approaches. JMIR Med Inform 2022;10:e37557.
Kaur R, Ginige JA, Obst O. A Systematic literature review of automated ICD coding and classification systems using discharge summaries [homepage on the Internet]. arXiv; 2021 [cited 2025 Mar 25]. Available from: http://arxiv.org/abs/2107.10652
Diao X, Huo Y, Zhao S, Yuan J, Cui M, Wang Y, et al. Automated ICD coding for primary diagnosis via clinically interpretable machine learning. Int J Med Inform 2021;153:104543.
Oberste L, Finze N, Hoffmann P, Heinzl A. Supporting the billing process in outpatient medical care: automated medical coding through machine learning. ECIS 2022 Research Papers [homepage on the Internet]. 2022; [cited 2025 Jun 5]. Available from: https://aisel.aisnet.org/ecis2022_rp/136
Gruson D, Magalhaes T, Ruszanov A, Granaldi C, Bernardini S, Buttigieg SC. Hyperautomation in healthcare: perspectives from a joint IFCC – EHMA session. Electron J Int Fed Clin Chem Lab Med 2023;34:284–6.
Hazarika I. Artificial intelligence: opportunities and implications for the health workforce. Int Health 2020;12:241–5.
Samah T, Samar M. Investigating the key trends in applying artificial intelligence to health technologies: a scoping review. PLoS One 2025;20:e0322197.
Zhao Z, Lu W, Peng X, Xing L, Zhang W, Zheng C. Automated ICD coding via contrastive learning with back reference and synonym knowledge for smart self-diagnosis applications. IEEE transactions on consumer electronics 2024;70:6042–53.
Zhou T, Cao P, Chen Y, Liu K, Zhao J, Niu K, et al. Automatic ICD coding via interactive shared representation networks with self-distillation mechanism. In: Zong C, Xia F, Li W, Navigli R, editors. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) [homepage on the Internet]. Online: Association for computational linguistics; 2021 [cited 2025 Mar 25]. Available from: https://aclanthology.org/2021.acl-long.463/
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Author and Journal
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
