完全随机缺失条件下分类随机变量数据缺失插补方法的比较研究

目的探讨完全随机缺失条件下分类随机变量数据缺失对研究结果的影响,对各方法插补效果进行评价。方法基于上海地区35岁及以上吸烟人群吸烟与肺癌死亡关系的完整数据集,在5%、10%、20%及30%缺失率下,模拟有序分类变量(吸烟年数分组syfz)缺失和二分类变量(性别sex)缺失,重复模拟100次。采用删除法、众数插补法、多重插补-logistic回归法(MI/logistic)及多重插补-判别分析法(ML/discrim)对分类变量数据缺失进行处理。对插补效果从插补正确率及插补后模型参数的变化两个方面进行评价。结果有序分类变量缺失:各缺失率下,MI/logistic插补的正确率最高,MI/logistic和MI/discrim插补后模型参数的偏差均较小,对于吸烟年数sy以分组形式syfz纳入模型数据缺失导致模型参数的相对偏差更小,对syfz插补后模型参数相对偏差也小于连续变量sy插补后模型参数相对偏差。二分类变量缺失:各缺失率下,众数插补的正确率最高,删除法处理缺失数据后模型参数的偏差最小。结论连续变量缺失对模型结果的影响大于分类变量缺失,对于有数据缺失的连续变量可将其离散化,以分类变量的形式进行分析。缺失数据插补模型的拟合效果会直接影响插补效果,当模型拟合效果较差时可能会带来更大的偏差。     

完全随机缺失条件下连续型随机变量数据缺失插补方法的比较研究

目的探讨完全随机缺失条件下连续型随机变量数据缺失对研究结果的影响,对各方法插补效果进行比较。方法基于上海地区35岁及以上吸烟人群吸烟与肺癌死亡关系的完整数据集,在5%、10%、20%及30%缺失率下,模拟单变量(吸烟年数sy)缺失,采用了7种方法处理单变量缺失;模拟多变量(吸烟年数sy和每天吸烟支数smd)缺失,采用了4种方法处理多变量缺失。对插补效果从缺失变量均值的变化、插补精确性及插补后模型参数的变化三个方面进行评价。结果单变量缺失:各缺失率下,回归插补sy均值的偏差最小,MI/REG、MI/PMM和MI/MCMC插补后模型参数的偏差均较小,删除法sy均值与模型参数的偏差均最大。多变量缺失:各缺失率下,回归插补sy均值的偏差最小,删除法最大;条件均值插补smd均值的偏差最小,MI/MCMC最大;条件均值插补模型参数的偏差最小,MI/MCMC最大。结论用不同指标对各方法插补效果进行评价会得出不同的结果,应根据统计分析的目的和关注点选择最合适的缺失数据处理方法。总体来看,插补法处理缺失数据的效果优于删除法,缺失率越高,优势越显著。     

队列研究中纵向缺失数据填补方法的模拟研究

目的 数据缺失是队列研究中几乎无法避免的问题。本文旨在通过模拟研究,比较当前常见的8种缺失数据处理方法在纵向缺失数据中的填补效果,为纵向缺失数据的处理提供有价值的参考。方法 模拟研究基于R语言编程实现,通过Monte Carlo方法产生纵向缺失数据,通过比较不同填补方法的平均绝对偏差、平均相对偏差和回归分析的Ⅰ类错误,评价不同填补方法对于纵向缺失数据的填补效果及对后续多因素分析的影响。结果 均值填补、k近邻填补（KNN）、回归填补和随机森林的填补效果接近,且表现稳定;多重插补和热卡填充次于以上填补方法;K均值聚类和EM算法填补效果最差,表现也最不稳定。均值填补、EM算法、随机森林、KNN和回归填补可较好地控制Ⅰ类错误,多重插补、热卡填充和K均值聚类不能有效控制Ⅰ类错误。结论 对于纵向缺失数据,在随机缺失机制下,均值填补、KNN、回归填补和随机森林均可作为较好的填补方法,当缺失比例不太大时,多重插补和热卡填充也表现较好,不推荐K均值聚类和EM算法。     

药物经济学评价中删失成本数据的处理方法

通过对药物经济学评价中删失成本数据的类型介绍,总结不同类型删失成本数据的识别和处理方法,为实施药物经济学评价时处理删失成本数据提供方法学上的参考。     

R语言survHE程序包在卫生经济学评价中的应用

R语言中survHE程序包是一款专门用于卫生经济学评价中进行生存分析的程序包,该程序包可以使用频率学和贝叶斯方法拟合各种生存参数模型,并且通过一组特定函数将卫生经济学评估中涉及生存分析的工作流程系统化和简单化,例如:(1) digitise和make.ipd函数可以将从文献中提取的KM数据重构为个体数据集;(2) fit.models函数可以对重构的个体数据集进行参数生存分析并对相应的模型进行评估;(3) make.surv函数可以对生存曲线进行概率敏感性分析。文章结合实例重点展示了在卫生经济学评价中使用survHE程序包进行生存分析的完整操作过程。     

多重填补法和多水平模型在纵向随访数据中的应用

目的 探讨在纵向随访数据中如何处理缺失值和相关性,充分利用所收集到的数据来反映研究总体。方法 先模拟产生纵向完整数据集和缺失数据集,然后用多重填补法（multiple imputation methods,MI）和多水平模型（multilevel model,MLM）来处理,再用随机区组方差分析比较各组的差异,最后用实例验证。结果 不同缺失类型和不同缺失比例的数据集所得结果一致:基于MI的MLM所得的偏差比MLM小,且随着填补次数的增多而有所减小;偏差随着缺失率的增大而增加,样本量大的结果更稳定。实例分析也验证了模拟的结果。结论 用多重填补法和多水平模型共同处理纵向随访数据可以提高结果的准确性和精确性。     

含缺失值时间序列的ARMA模型拟合

目的 给出一种有效的处理含缺失值时间序列的方法，完成缺失值的内插及ＡＲＭＡ模型的参数估计。方法 用状态空间的Ｍａｒｋｏｖ表达描述时间序列，进而采用Ｋａｌｍａｎ滤波技术。结果 实例分析表明，不仅可以完成缺失值的有效内插，模型拟合效果及预测结虹也甚为柒满意。结论 用基于状态空间表达的Ｋａｌｍａｎ滤波技术，可以实现平稳可逆时间序列中缺失值的内插及ＡＲＭＡ模型拟合。     

多水平统计模型在4×4交叉设计不完整数据的生物等效性中的应用

目的针对多阶段交叉设计中的不完整数据,分别采用多水平模型RIGLS算法、MOM以及REML算法分析,并与完整数据的结果进行比较。方法针对BE研究中4×4交叉设计,分别运用FDA指导原则中推荐的矩法(Method of Moments,MOM)、限制性极大似然法(Restricted Maximum Likelihood Method,REML)以及基于限制性迭代广义最小二乘法(Restricted Iterative Generalized Least Squares,RIGLS)估计的多水平模型,采用实例与模拟结合的方式进行分析,探讨不同方法在有随机缺失的不完整数据中的结果比较。结果 (1)多水平模型的首要优势在于考虑了数据误差的层次性,将传统模型中的误差随机项分解到与数据层次结构对应的各个水平上。(2)多水平模型可通过在个体水平拟合随机效应,直接估计个体-药物交互作用的方差σ2D,克服传统方法因间接估计而引起的偏倚。(3)多水平模型其算法不要求所有的观测个体有相同的观测次数,可充分利用含有缺失值的信息,对于缺失数据和非均衡设计具有较好普适性。结论本研究将生物等效性与多水平建模有机结合,为非均衡、高变异、小样本、有缺失的生物等效性评价开拓了新的思路,提供了新的方法。     

卫生经济学评价报告标准2022解读

卫生经济学评价相关研究数量与日俱增。卫生经济学评价报告标准2022(CHEERS 2022)共包含28个条目清单, 在CHEERS 2013声明的基础上, CHEERS 2022新增了包括卫生经济学分析计划、模型共享, 以及社区、患者和公众等利益相关方的研究参与等内容, 兼顾了卫生经济学评价的未来学科发展方向。CHEERS 2022不仅有助于全球研究者采用统一的标准规范报告卫生经济学评价研究结果, 为同行评议专家、编辑及读者提供有用的审查工具, 还可支持卫生技术评估机构建立规范的评价报告标准。本文旨在简要介绍CHEERS 2022, 解读部分条目, 结合传染病流行病学领域的卫生经济学评价实例进行分析, 以期为研究者规范报告卫生经济学评价研究提供参考。     

数据缺失机制对逐步回归变量筛选的影响

目的 评价各种数据缺失机制对逐步回归变量筛选结果的影响.方法 通过模拟产生不同缺失机制和缺失类型的数据,用筛选到的真实变量的个数和损失函数大小作为指标,评价其对逐步同归的影响.结果 完整数据情况下的筛选表现优于各缺失机制卜表现；缺失类型比缺失机制对筛选结果的影响更为明显.结论用逐步回归对含缺失值的数据进行变量筛选时,需要关注缺失机制和缺失类型.     

卫生技术经济评价方法在唐氏综合征产前筛查方案选择中的应用探讨

目的:探讨技术经济方法在卫生技术经济评价中的应用。方法:以唐氏综合征两种筛查方案的实际数据为案例,应用技术经济分析方法,进行卫生技术经济评价分析。结果:在卫生技术评价中,结合具体数据的技术特征,应用微观经济学平均成本、边际成本,以及总收益和总成本函数曲线进行必要的拟合,可给出相对精确的最大效率和最大效益的判断。现行卫生技术经济评价中的增量分析存在一定的误区。结论:在技术自身灵敏度变化或诊疗技术组合要素变化的条件下,当其技术效率呈边际递减规律时,增量成本效果比的计算应采用正统方式,且其值的判别,并非愈小愈好。     

Economic evaluation of health promotion: Friend or foe?

It is commonly believed that economic evaluation is hostile to health promotion and that the requirement for health programs to be cost effective will result in a biased allocation of funds in favour of programs that can demonstrate short-run benefits as defined by inadequate outcome measures. The paper is concerned with the validity of this perception. It is argued that economic evaluation has the potential for treating health promotion activities on an equal basis with other health interventions. The major obstacle to this does not arise from the theory of economic evaluation, which is discussed, but from a lack of information about outcomes. Without this information any evaluation - economic or otherwise - is flawed.
Three problems relating to the economic evaluation of health promotion activities are considered. These are:
* the discounting of future health benefits;
* the potential for economic evaluation to be counter-productive if applied to 'immature' projects; and
* the practical problems encountered in the measurement of the outcomes of health promotion programs.
A four-fold classification of possible outcomes is suggested which is based upon a distinction between disease cure, individual health promotion, community welfare and systemic change designed to promote either individual health or social well-being. The capacity of economics to incorporate these objectives is discussed.     

Economic Evaluation in Health Care: The Point of View of Informed Physicians

OBJECTIVES: We investigated health professionals with a solid background in health-care management and economics to get their opinion and attitude on the use of economic evaluation at the policy, organizational, and professional levels of decision-making. METHODS: A 12-item questionnaire was sent to 374 Italian health-care professionals who received training in economic evaluation of health-care programs in the last 10 years at the Bocconi School of Management, Milan, Italy. RESULTS: The response rate was 46.8% (175 questionnaires). All respondents stated that the basics of economic evaluation analysis must be part of the overall knowledge of health-care professionals. The usefulness of economic evaluation for professional activities was rated 3.83 (scale 1-5). Respondents stated that economic evaluation is used more for managerial decisions than for clinical ones (mean 2.89 vs. 2.74, P = 0.09). "Decisions are taken according to a short-term perspective" was the most frequently reported barrier for the actual use of economic evaluation studies, particularly by managers (76.7%). "More training in health economics" was indicated as the incentive to expand its use by the majority of both clinicians and managers (64.6%). Significantly more managers than clinicians (74.4% vs. 54.1%, P = 0.005) considered that the maximum benefits of economic evaluation are reaped at organizational level. CONCLUSIONS: Informed Italian health professionals have a positive attitude toward the principles and the techniques of economic evaluation. They show appreciation of their potential role and report making some use of them in actual decision making.     

卫生经济学评价方法在临床路径中的适用性、现状及应用思路

了解卫生经济学评价方法及其在临床路径中的适用性,正确认识和应用卫生经济学评价方法,对科学建立和客观评价临床路径具有重要意义。文章通过对卫生经济学评价的概念、4种方法的特征和适用性以及应用现状的比较分析,提出了临床路径卫生经济学评价的应用思路。     

Secondary Data Analysis of National Surveys in Japan Toward Improving Population Health

Secondary data analysis of national health surveys of the general population is a standard methodology for health metrics and evaluation; it is used to monitor trends in population health over time and benchmark the performance of health systems. In Japan, the government has established electronic databases of individual records from national surveys of the population’s health. However, the number of publications based on these datasets is small considering the scale and coverage of the surveys. There appear to be two major obstacles to the secondary use of Japanese national health survey data: strict data access control under the Statistics Act and an inadequate interdisciplinary research environment for resolving methodological difficulties encountered when dealing with secondary data. The usefulness of secondary analysis of survey data is evident with examples from the author’s previous studies based on vital records and the National Health and Nutrition Surveys, which showed that (i) tobacco smoking and high blood pressure are the major risk factors for adult mortality from non-communicable diseases in Japan; (ii) the decrease in mean blood pressure in Japan from the late 1980s to the early 2000s was partly attributable to the increased use of antihypertensive medication and reduced dietary salt intake; and (iii) progress in treatment coverage and control of high blood pressure is slower in Japan than in the United States and Britain. National health surveys in Japan are an invaluable asset, and findings from secondary analyses of these surveys would provide important suggestions for improving health in people around the world.Key words: secondary data analysis, national health survey, population health, Japan     

Occupational Safety and Health Interventions to Reduce Musculoskeletal Symptoms in the Health Care Sector

Introduction Health care work is dangerous and multiple interventions have been tested to reduce the occupational hazards. Methods A systematic review of the literature used a best evidence synthesis approach to address the general question “Do occupational safety and health interventions in health care settings have an effect on musculoskeletal health status?” This was followed by an evaluation of the effectiveness of specific interventions. Results The initial search identified 8,465 articles, for the period 1980–2006, which were reduced to 16 studies based on content and quality. A moderate level of evidence was observed for the general question. Moderate evidence was observed for: (1) exercise interventions and (2) multi-component patient handling interventions. An updated search for the period 2006–2009 added three studies and a moderate level of evidence now indicates: (1) patient handling training alone and (2) cognitive behavior training alone have no effect on musculoskeletal health. Few high quality studies were found that examined the effects of interventions in health care settings on musculoskeletal health. Conclusions The findings here echo previous systematic reviews supporting exercise as providing positive health benefits and training alone as not being effective. Given the moderate level of evidence, exercise interventions and multi-component patient handling interventions (MCPHI) were recommended as practices to consider. A multi-component intervention includes a policy that defines an organizational commitment to reducing injuries associated with patient handling, purchase of appropriate lift or transfer equipment to reduce biomechanical hazards and a broad-based ergonomics training program that includes safe patient handling and/or equipment usage. The review demonstrates MCPHI can be evaluated if the term multi-component is clearly defined and consistently applied.     

Multiple imputation for body mass index: lessons from the Australian Longitudinal Study on Women's Health

In large epidemiological studies missing data can be a problem, especially if information is sought on a sensitive topic or when a composite measure is calculated from several variables each affected by missing values. Multiple imputation is the method of choice for 'filling in' missing data based on associations among variables. Using an example about body mass index from the Australian Longitudinal Study on Women's Health, we identify a subset of variables that are particularly useful for imputing values for the target variables. Then we illustrate two uses of multiple imputation. The first is to examine and correct for bias when data are not missing completely at random. The second is to impute missing values for an important covariate; in this case omission from the imputation process of variables to be used in the analysis may introduce bias. We conclude with several recommendations for handling issues of missing data.     

Principled Missing Data Treatments

We review a number of issues regarding missing data treatments for intervention and prevention researchers. Many of the common missing data practices in prevention research are still, unfortunately, ill-advised (e.g., use of listwise and pairwise deletion, insufficient use of auxiliary variables). Our goal is to promote better practice in the handling of missing data. We review the current state of missing data methodology and recent missing data reporting in prevention research. We describe antiquated, ad hoc missing data treatments and discuss their limitations. We discuss two modern, principled missing data treatments: multiple imputation and full information maximum likelihood, and we offer practical tips on how to best employ these methods in prevention research. The principled missing data treatments that we discuss are couched in terms of how they improve causal and statistical inference in the prevention sciences. Our recommendations are firmly grounded in missing data theory and well-validated statistical principles for handling the missing data issues that are ubiquitous in biosocial and prevention research. We augment our broad survey of missing data analysis with references to more exhaustive resources.