Application of cumulative ratio logit model and partial proportional odds model in analyzing factors influencing hospitalization costs for lung cancer surgery patients
Abstract：Objective To illuminate the application of cumulative odds logit model and partial proportional odds model in ordinal categorical data.Methods The cumulative odds logit model and partial proportional odds model were used to fit the relationship between hospital costs and relative factors of lung cancer surgery. The goodness of fit of the two models was compared by the log likelihood value (-2ln L) of the models.Results When the independent variables did not fully satisfy the proportional odds assumption (gender χ2=15.888, P＜0.001; surgical procedure χ2=35.874，P＜0.001), the fitting results of the two methods were different. At different segmentation points, there was a significant difference in the relationship between hospitalization costs and surgical procedures. The partial proportional odds model (-2ln L=5 797.112) was better than the cumulative odds logit model (-2ln L=5 852.420). The difference of the log likelihood values was statistically significant (P<0.05).Conclusion In order to analyze the ordinal categorical data, the proportional odds assumption should be verified firstly. If all the independent variables satisfy the condition, the cumulative odds logit model is selected, otherwise the partial proportional odds model is used to study the data.
曾祥嫚. 应用累积比数logit模型和偏比例优势模型分析肺癌手术患者住院费用影响因素[J]. 中国医院统计, 2024, 31(1): 47-49.
Zeng Xiangman. Application of cumulative ratio logit model and partial proportional odds model in analyzing factors influencing hospitalization costs for lung cancer surgery patients . journal1, 2024, 31(1): 47-49.
［3］LALL R, CAMPBELL M J, WALTERS S J, et al. A review of ordinal regression models applied on health-related quality of life assessments［J］. Stat Methods Med Res, 2002, 11(1):49-67. DOI: 10.1191/0962280202sm271ra.
［4］CHARLSON M E, POMPEI P, ALES K L, et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation［J］. J Chronic Dis, 1987, 40(5):373-383. DOI: 10.1016/0021-9681(87)90171-8.
