The competition and cooperation mechanism of policy goals in China's diagnosis-related group: a quantitative analysis based on the BERTopic and Lotka-Volterra models.

Abstract

Background: Aiming at the practical problems such as upcoding and decomposing hospitalizations emerging in China's Diagnosis-Related Group (DRG) reform, existing studies mostly attribute such issues to aspects like behavioral motivations, which fail to explain the root causes of the problems.

Methods: This paper takes policy goals as the research starting point and constructs a three-dimensional analytical framework of "identifying goals-quantifying relationships-explaining mechanisms". Based on 377 DRG policy texts covering the central, provincial and municipal levels from 2009 to 2025, this study employs the BERTopic model to identify the policy goals of three phases in turn, visualizes the evolutionary characteristics of the goals through Jaccard similarity analysis, and introduces the Lotka-Volterra model to quantify the competitive and cooperative coefficients among the six major policy goals in the national promotion phase, further classifying their relationships into three types: cooperation, predation and competition.

Results: The study finds that the policy goal system of China's DRG reform presents an evolutionary law from fragmented advancement to systematic improvement, with healthcare insurance fund management and controlling healthcare fees as the core thread running through all three phases. The six major policy goals in the national promotion phase are improving primary healthcare levels (G1), standardizing diagnosis and treatment behaviors (G2), controlling healthcare fees (G3), raising the use efficiency of fund (G4), enhancing service quality (G5), and reflecting the technical value of healthcare personnel (G6), among which there exist 6 groups of cooperative relationships, 5 groups of predatory relationships and 4 groups of competitive relationships. Specifically, G5 and G6 exhibit a prominent synergistic effect, while G5 is subject to the triple squeeze from G2, G3 and G4, and G1 is in a competitive relationship with G2, G3 and G4 respectively.

Conclusions: From the perspective of policy design, this study reveals that the structural contradictions among DRG policy goals may induce healthcare institutions and healthcare personnel to adopt strategic behaviors under the pressure of multiple goals. This research provides a theoretical basis for optimizing DRG policy design and promoting goal coordination, and also offers reference significance for DRG reforms worldwide.