Analysis on the Implementation of DRG Payment in a Hospital Using K-means++ Clustering Combined with the Elbow Method
Objectives This study aimed to analyze the case grouping settlement situation after a Three A and Tertiary General Hospital in Beijing which officially implemented the China Healthcare Security Diagnosis-Related Group(CHS-DRG)payment for one year,to provide quantitative data support for the detailed management of hospital DRGs.Methods Utilizing the K-means++ clustering combined with the Elbow method,the study selected cluster features such as DRG group number,average length of stay,and the proportion of cost expenses,analyzing the hospital service situation of 12,634 cases across 17 clinical departments from March 15th,2022,to March 15th,2023.Principal component analysis was used to visualize the clustering results.Results The Three A and Tertiary General Hospital actually paid a total surplus of 33,354 thousand yuan for 12 634 cases from March 15th,2022 to March 15th,2023,accounting for 13.2%of the total amount.A total of 444 DRG groups were enrolled,of which 314 were surplus DRG groups and 111 were deficit DRG groups.There were 17 clinical departments,including 15 surplus departments,which were grouped into 3 cluster groups.Among the top five departments in cluster group 1 including cardiology,urology,orthopedics,general surgery and oncology and gastroenterology contributed 25,879 thousand yuan,accounting for 77.59%of the total surplus.At the same time,there were 364 DRG groups in these five departments,accounting for 81.98%of the total number of groups in the hospital.The total number of cases was 6571,accounting for 52.01%of the total number of cases.There were losses in two clinical departments out of three clusters,which accounted for 8%of the total surplus.Conclusions By clustering and grouping the one-year DRG payment data of the Three A and Tertiary General Hospital,the surplus and deficit departments were screened out,and the profit and loss data of key DRG disease groups in the hospital were compared to analyze the differences between the surplus departments and deficit departments in the number of DRG groups,the average length of stay,the proportion of drug consumption and the proportion of medical expenses,etc.,and the development direction and cost structure optimization suggestions for each cluster group of departments were put forward.It provided data support and improvement direction for improving the fine management level of a hospital.
NETL
NSTL
万方数据
