Objective: The paper explores the similarities and differences of different case complexity measurement models based on internal medicine diagnosis groups of pulmonary malignant tumor to provide a reference for further optimizing DRG grouping. Methods: Taking 2131 cases of pulmonary malignant tumor internal medicine diagnosis groups from a tertiary hospital in Luoyang City of Henan Province from 2019 to 2022 as the research object, the paper adopts the MCC/CC tables, PCCL model and ECC model to calculate the case complexity of the research object, constructs groups through decision trees, and compares and analyzes the effectiveness of the groups. Results: According to the MCC/CC table, the internal medicine diagnosis groups of pulmonary malignant tumor was divided into 3 groups, the CV was 0.51, 0.57 and 0.62, and the RIV was 0.02. According to the PCCL model, they were divided into 5 groups, the CV was 0.57, 0.57, 0.50, 0.76 and 0.70, and the RIV was 0.14. According to the ECC model, they were divided into 3 groups, the CV was 0.61, 0.63 and 0.49, and the RIV was 0.03. According to the results of non-parametric test, there were significant differences in hospitalization expenses among the basic groups. Conclusion: Based on the grouping performance, the MCC/CC table and ECC model are better than the PCCL model, and the grouping effect of the MCC/CC table and ECC model under different ADRG needs to be further compared and studied.