摘要
目的基于两参数项目反应模型计算乳腺癌各维度治疗质量,利用变异系数加权组合赋权模型进行维度赋权,加权计算乳腺癌维度融合的综合治疗质量,为改善治疗质量提供依据。方法选择不同维度乳腺癌治疗质量的评价指标,收集医院乳腺癌患者的病历数据,利用两参数logistic项目反应模型计算各维度治疗质量;利用序关系分析法(garbage-first,G1法)计算主观权重、余弦夹角法计算客观权重、将主客观权重纳入变异系数加权组合赋权模型获得组合权重;加权计算维度融合后的综合治疗质量并进行评价。结果选择三个维度的9个乳腺癌治疗质量评价指标,三个维度分别为功能评价、治疗方法及术前检查,收集9家医院3669例患者的完整病历信息。计算获得医院的各维度治疗质量:功能评价维度最高得分为1.0317,治疗方法维度最高得分为0.9232,术前检查维度最高得分为1.2001。G1法算得三个维度的主观权重分别为0.5364、0.2980和0.1656;余弦夹角法算得客观权重分别为:0.5220、0.3554和0.5868。利用变异系数加权组合赋权模型获得各维度组合权重:第9家医院的功能评价维度最大组合权重为0.5293,第5家医院的治疗方法维度最大组合权重为0.3209,第1家医院的术前检查维度最大组合权重为0.3857。加权计算融合后的综合治疗质量:排名前三的医院分别为第1家,第7家和第9家医院,综合得分分别为1.04220,1.0371和1.0157。结论研究提示组合权重法能在主客观权重出现极值时获得更稳定的权重,使融合后的治疗质量排名更合理;功能评价维度对综合治疗质量的影响较大,提高功能评价维度的使用率有利于改善乳腺癌的治疗质量。
Objective The quality of care of each dimension of breast cancer was calculated based on the two-parameter item response model,and the weighted combination weighting model of coefficient of variation was used to calculate the quality of care of breast cancer dimension fusion,which provided the basis for improving the treatment quality.Methods The evaluation indexes of treatment quality of breast cancer in different dimensions were selected,the medical records of breast cancer patients in hospitals were collected,and the quality of care of each dimension was calculated by two-parameter logistic item response model.The G1 method was used to calculate the subjective weight,the cosine angle method was used to calculate the objective weight,and the subjective and objective weight were incorporated into the weighted combination weighting model of coefficient of variation to obtain the combination weight.The comprehensive treatment quality after dimension fusion was weighted and evaluated.Results Nine evaluation indexes of breast cancer treatment quality were selected from three dimensions,including functional evaluation,treatment methods and preoperative examination.The complete medical records of 3669 patients from 9 hospitals were collected.The highest score of the functional evaluation dimension was 1.0317,the highest score of the treatment method dimension was 0.9232 and the highest score of preoperative examination dimension was 1.2001.The subjective weights of the three dimensions calculated by G1 method were 0.5364,0.2980 and 0.1656 respectively.Objective weight was calculated by Cosine Angle method:the highest objective weight of functional evaluation dimension was 0.5220,the highest objective weight of treatment method dimension was 0.3554,and the highest objective weight of preoperative examination dimension was 0.5868.The weighted combination weighting model of coefficient of variation was used to obtain the combination weight of each dimension:the maximum combination weight of function evaluation dimension of the 9th hospital was 0.5293,the maximum combination weight of treatment dimension of the 5th hospital was 0.3209,and the maximum combination weight of preoperative examination dimension of the 1st hospital was 0.3857.Weighted comprehensive quality of care after fusion:the top three hospitals were the first,seventh and ninth hospitals,with combined scores of 1.04220,1.0371 and 1.0157,respectively.Conclusion The study suggests that the combined weight method can obtain more stable weight when the subjective and objective weight appears extreme value,and make the treatment quality ranking after fusion more reasonable.Functional evaluation dimension has a great influence on the quality of comprehensive treatment,and improving the utilization rate of functional evaluation dimension is beneficial to improve the quality of breast cancer treatment.
作者
安宁
栗景坤
刘美娜
An Ning;Li Jingkun;Liu Meina(Department of Medical Statistics,Harbin Medical University(150081),Harbin)
出处
《中国卫生统计》
CSCD
北大核心
2023年第5期646-650,共5页
Chinese Journal of Health Statistics
基金
国家自然科学基金资助项目(82173614)。
关键词
治疗质量评价
项目反应模型
维度融合
Evaluation of quality of care
Item response model
Dimensions of fusion
