疾病治疗质量纵向综合评价模型的构建及在CIS中的应用

Construction of Longitudinal Comprehensive Evaluation Model of Disease Treatment Quality and Its Application in CIS

目的构建疾病治疗质量纵向综合评价模型,应用模型评价急性缺血性脑卒中(cerebral ischemic stroke,CIS)的治疗质量。方法潜变量测量模型联合多项式和多水平回归,建立疾病治疗质量纵向综合评价模型,利用马尔科夫蒙特卡洛(Markov Chain Monte Carlo,MCMC)方法进行参数估计。模拟研究确定模型的适用条件,通过均方根误差(root mean square error,RMSE)和秩相关系数评价模型参数估计的准确性,并与潜变量增长曲线模型(latent growth curve model,LGCM)进行效果比较。结果构建可用于评价疾病治疗质量纵向变化、趋势变化及协变量影响的综合模型。模拟研究结果:评价对象数对模型效果影响较小,观测时间数6个、评价指标数20个时,模型可获得评价对象不同时间点治疗质量及变化趋势。治疗质量估计值与真值的秩相关系数均值为0.8128,高于LGCM的0.7610,协变量回归系数估计值与真值的RMSE均值为0.3756,低于LGCM的0.448;观测时间数及评价指标数越多,纵向综合评价模型效果越准确,且始终高于LGCM。应用结果:2011年到2016年,医院CIS平均治疗质量综合得分增加了0.9256,二级和三级医院分别增加了0.9432和1.321,教学和非教学医院分别增加了1.4904和1.1155。结论纵向综合评价模型可以全面合理评价疾病治疗质量及其纵向变化,分析评价对象的差异对变化趋势的影响。我国CIS治疗质量总体向好,变化趋势受医院等级及教学状态影响,研究结果可为后续有针对性的治疗质量干预措施制定及治疗质量的持续改善提供参考。

Objective Construct a longitudinal comprehensive evaluation model of disease treatment quality to evaluate the quality of treatment for cerebral ischemic stroke(CIS).Methods The latent variable measurement model was combined with polynomial and multilevel regression to establish a longitudinal comprehensive evaluation model of disease treatment quality,the MCMC was used for parameter estimation.Simulation studies were conducted to determine the applicable conditions of the model,the accuracy of model parameter estimation was evaluated by root mean square error and rank correlation coefficient,the results were compared with the latent growth curve model(LGCM).Results A comprehensive model that can be used to evaluate the longitudinal change,trend change and the impact of covariates in the quality of disease care was constructed.The number of evaluation objects has a little effect on the model effect.When the number of observation time was 6 and the number of evaluation indicators was 20,the model could obtain the treatment quality and change trend of the evaluation object at different time points.The mean rank correlation coefficient between the estimated value of treatment quality and the true value was 0.8128,which was higher than that of LGCM(0.7610).The mean RMSE between the estimated value of the covariate regression coefficient and the true value was 0.3756,which was lower than that of LGCM(0.448).The more observation time and evaluation indexes,the more accurate the longitudinal comprehensive evaluation model was,and it was always higher than LGCM.From 2011 to 2016,the average CIS quality of care comprehensive score of hospitals increased by 0.9256,secondary and tertiary hospitals increased by 0.9432 and 1.321,teaching and non-teaching hospitals increased by 1.4904 and 1.1155.Conclusion The longitudinal comprehensive evaluation model can comprehensively and reasonably evaluate the quality of disease treatment and its longitudinal changes,and analyze the impact of differences in evaluation objects on the change trend.The quality of CIS treatment in China is generally good,and the changing trend is affected by the level of hospitals and teaching status.The results of this study can provide reference for the subsequent development of targeted treatment quality intervention measures and continuous improvement of treatment quality.