AI-based ICD coding and classification approaches using discharge summaries: A systematic literature review. (1st March 2023)

• Record Type: Journal Article
• Authors: Kaur, Rajvir; Ginige, Jeewani Anupama; Obst, Oliver
• Is Part Of: Expert systems with applications. Volume 213:Part B(2023)
• Publication Date: 2023-03-01
• Subjects: Computer assisted clinical coding -- Clinical classification and coding -- Discharge summaries -- Natural Language Processing -- Machine learning -- Deep learning; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2022.118997
• Languages: English
• ISSNs: 0957-4174

Abstract

Abstract: The assignment of codes to free-text clinical narratives have long been recognised to be beneficial for secondary uses such as funding, insurance claim processing and research. The current scenario of assigning clinical codes is a manual process which is very expensive, time-consuming and error prone. In recent years, many researchers have studied the use of Natural Language Processing (NLP), related machine learning and deep learning methods and techniques to resolve the problem of manual coding of clinical narratives and to assist human coders to assign clinical codes more accurately and efficiently. The main objective of this systematic literature review is to provide a comprehensive overview of automated clinical coding systems that utilise appropriate NLP, machine learning and deep learning methods and techniques to assign the International Classification of Diseases (ICD) codes to discharge summaries. We have followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and conducted a comprehensive search of publications from January, 2010 to December 2021 in four high quality academic databases: PubMed, ScienceDirect, Association for Computing Machinery (ACM) Digital Library, and the Association for Computational Linguistics (ACL) Anthology. We reviewed 6128 publications; 42 met the inclusion criteria. This review identified: 6 datasets having discharge summaries (2 publicly available, 4 acquired from hospitals); 14 NLP techniques along with some other data extraction processes, different feature extraction and embedding techniques. The review also shows that there is a significant increase in the use of deep learning models compared to machine learning. To measure the performance of classification methods, different evaluation metrics are used. Efforts are still required to improve ICD code prediction accuracy, availability of large-scale de-identified clinical corpora with the latest version of the classification system. This can be a platform to guide and share knowledge with the less experienced coders and researchers. Highlights: This review focuses on automated ICD code assignment using discharge summaries. We consider 42 studies published between January 2010 and December 2021. Selected studies use Natural Language Processing and Machine Learning approaches. Comparison of studies shows a significant increase in use of deep learning models. We highlight limitations in existing studies and discuss open research challenges.