DRG grouping by machine learning: from expert-oriented to data-based method

Liu et al. BMC Medical Informatics and Decision Making https://doi.org/10.1186/s12911-021-01676-7 (2021) 21:312 Open Access RESEARCH DRG grouping by machine learning: from expert‑oriented to data‑based method Xiaoting Liu1,2†, Chenhao Fang3†, Chao Wu1, Jianxing Yu1,4* and Qi Zhao1 Abstract Background: Diagnosis-related groups (DRGs) are a payment system that could effectively solve the problem of excessive increases in healthcare costs which are applied as a principal measure in the healthcare reform in China. However, expert-oriented DRG grouping is a black box with the drawbacks of upcoding and high cost. Methods: This study proposes a method of data-based grouping, designed and updated by machine learning algorithms, which could be trained by real cases, or even simulated cases. It inherits the decision-making rules from the expert-oriented grouping and improves performance by incorporating continuous updates at low cost. Five typical classification algorithms were assessed and some suggestions were made for algorithm choice. The kappa coefficients were reported to evaluate the performance of grouping. Results: Based on tenfold cross-validation, experiments showed that data-based grouping had a similar classification performance to the expert-oriented grouping when choosing suitable algorithms. The groupings trained by simulated cases had less accuracy when they were tested by the real cases rather than simulated cases, but the kappa coefficients of the best model were still higher than 0.6. When the grouping was tested in a new DRGs system, the average kappa coefficients were significantly improved from 0.1534 to 0.6435 by the update; and with enough computation resources, the update process could be completed in a very short time. Conclusions: As a new potential option, the data-based grouping meets the requirements of the DRGs system and has the advantages of high transparency and low cost in the design and update process. Keywords: Diagnosis-related groups (DRGs), Grouping, Machine learning, China, Healthcare Introduction In the most recent healthcare reform, China has made substantial progress in improving equal access to care and enhancing financial protection. However, gaps remain in efficiency in the delivery and control of health expenditures [1]. With the enhancement and standardisation of medical information systems and clinical pathways, the Chinese government has paid closer attention to payment *Correspondence: † Xiaoting Liu and Chenhao Fang are co-first authors of this paper. They contributed the paper equally 1 School of Public Affairs, Zhejiang University, Zijingang Campus, Hangzhou 310058, Zhejiang Province, China Full list of author information is available at the end of the article reform and enhanced supervision of the quality of medical care in the new round of healthcare reform, hoping to curb soaring medical expenditures [2, 3]. One of the core measures is provider payment reform, in which diagnosis-related groups (DRGs) payment is perceived as a valuable alternative to the conventional fee-for-service (FFS) payment method. In 2009, the Chinese government announced the initiation of the prospective DRG-based payment reform. Until 2016, two national DRG groupings, CN-DRGs and C-DRGs, were developed and tested in Sanming, Shenzhen and Karamay, and nearly twenty of the thirty-two provinces in mainland China implemented the simplified DRGs. Originating from Yale University and first implemented in the United States in 1983 [4], DRGs is a payment © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Liu et al. BMC Medical Informatics and Decision Making (2021) 21:312 Page 2 of 11 system that can gather patients with similar clinical symptoms and similar resource consumption patterns into the same group. The medical expenses that patients and medical insurance need to pay are only related to the results of grouping [5, 6]. In the DRGs system, excessive drugs and treatment provided by hospitals will not be paid for, which improves healthcare quality and stabilises costs [7]. Aiming at allowing for more ‘outside’ control on hospital expenditure, several pieces of common grouping software have been developed to standardise and facilitate hospital payments in China. However, as in many other countries, the basic DRGs structure has undergone numerous revisions since its creation, leading to a less stable, more complex, and often confusing process [8]. The grouping, an exhaustive patient case classification system, is the core design characteristic of a DRG-based payment system [9]. Treatment trajectory encoding information about a patient and their clinical treatment is put through a large formal decision tree—the grouping, which consists of thousands of decision rules, each evaluating to either true or false. By traversing these decision rules, a care product is defined and determined [10]. As the grouping is a black box, the decision-making rules of which are not disclosed to the public, its algorithmic nature makes reimbursing a highly technical endeavour. Due to the complexity and lack of transparency of the grouping software, on the one hand, it might spark a public debate about whether providers and professionals might use the system to further their interest [10]; on the other hand, as clinicians have stated, the grouping software has rendered the payment process too complex and error-sensitive, leading to remuneration errors and subsequent loss of hospital income [11]. Moreover, with the purpose of cost control, DRGs payment is usually supported by the mechanism of Global Budget and a maximum growth percentage for hospital care in the government’s pilot practices of DRGs [12]. Thus, in the special context of healthcare reform in China, the grouping software of DRGs is embedded within political concerns and measurements. Though it is necessary to keep the care code correct by updating it, most of the pilot reforms have a (...truncated)