常用正交表的构造原理及SAS实现

目的介绍常用正交表的构造原理,并应用SAS宏语句编制通用程序模块生成各种常用的正交表。方法应用哈达玛矩阵法、正交拉丁方法和并列法构造不同类型的常用正交表。结果采用哈达玛矩阵直积法、特征函数法、正交拉丁方法和并列法可以分别构造试验次数N=2s型、N≠2s且N=2(p+1)型(p为素数且p≡1(mod4))的2水平正交表、Lt2(tm)型正交表和混合型正交表。结论不同类型的正交表构造原理方法不同,应用SAS软件可生成各种常用的正交表,方便实用。     

临床试验中定性指标统计分析报表的SAS宏实现

<正>SAS软件是业界公认的一款统计分析软件,在临床试验的统计分析中应用十分广泛。应用其强大的宏程序功能,可以高效快捷地实现报表的自动化输出,使统计结果直接生成Word统计报表,避免了以往手工粘贴统计结果出现错误的问题~([1-2])。根据统计分析报告中常用的格式,本文将编写一套较为通用的定性指标(结果变量有序或者无序)的统计分析SAS宏程序。资料与统计方法根据统计学原理,常用频数、百分比/构成比等指     

疫情Word报表的SAS编程实现

目的进一步提高疫情分析工作效率和工作质量。方法：根据实际工作特点,调试现有的疫情数据库疫情统计报表程序。结果通过编写SAS程序和利用结果传递系统(ODS),使疫情数据库的统计结果直接生成Word统计报表,避免了以往手工粘贴统计结果到Word表格的麻烦,使得多项统计内容一次生成。结论编写的程序调试简便,减少了工作环节,方便于疫情统计分析的实际应用。     

单因素多水平临床试验定量指标统计分析报表的SAS宏实现

正SAS软件输出的统计结果多而复杂,常需要人为将统计结果复制粘贴到统计报告相应位置上,这一过程耗时耗力且容易出错。为此,可根据需要编制相应的SAS宏程序,直接生成针对各种类型资料的统计分析报表。以往对于成组计量或计数资料的统计分析报表的SAS宏程序已经有了一定的探讨~([1-2]),本文主要介绍单因素多水平临床试验定量指标统计分析报表自动实现的SAS宏程序。资料与统计方法     

SAS/IntrNet功能简介及初步应用

目的介绍统计分析软件(statistical analysis system,SAS)网络化模块(SAS/IntrNet)的基本功能并应用于医学实际。方法应用SAS/IntrNet远程调用SAS程序进行在线分析,从客户端获取参数通过网络传递给服务器端的SAS程序进行分析,并反馈结果给客户端。结果利用SAS/IntrNet程序处理流程和相关配置文件,成功地实现了在线膳食暴露评估分析实例,使得用户可以在不安装SAS软件的情况下方便地通过网页进行在线SAS统计分析。结论利用SAS/IntrNet模块实现SAS网络化应用,可使不甚了解SAS编程的使用者通过友好的访问界面即可实现SAS的在线统计分析,大大提高了SAS系统信息共享的能力。     

基于SAS软件的地市级医院健康体检数据预处理方法探索

目的 系统分析当前健康体检数据的数据特征，利用Excel和SAS软件宏过程实现数据预处理。方法 利用某地市级三甲医院2017年10月至2020年12月健康体检数据平台中的健康体检数据，通过数据梳理总结当前体检数据的特征，制定相应的预处理规则，并基于Excel和SAS软件提出具体数据预处理方案、操作流程及宏代码。结果 通过Excel和SAS软件进行了健康体检数据的批量列名转换，使其符合SAS软件变量名命名规则；实现了多个不同结构的数据集合并而不出现截断值，保证了数据库的完整性；通过删除缺失变量和观察、合并重复变量和识别重复观察等过程，最终结合人工识别完成了体检数据预处理，形成了可供研究者进一步使用的健康体检数据库。在处理过程中编写了SAS宏过程，实现了数据预处理代码模块化。结论 通过Excel和SAS软件可以实现健康体检数据高效预处理、提高了数据质量、增加了数据可利用性，为数据库的利用和分析奠定基础，为健康体检数据的多中心研究应用的实现提供可能，具有一定的应用推广价值。     

《医学统计方法与SAS应用技巧》一书出版发行

该书介绍了医学统计的基本方法和国际上通用的统计分析系统(简称SAS)软件包的应用技巧。重点介绍医学试验设计和方差分析以及如何简易、成功地使用SAS软件分析医学试验数据。内容不仅包括医学试验设计,常用和多元统计分析方法的概况,具体应用实例,相应的SAS引导程序     

连续变量诊断试验数据的ROC分析

目的 介绍一种连续诊断变量的ROC回归模型，以及在独立和相关结构下的参数估计方法，给出参数误差估计的Bootstrap方法。方法 应用SAS软件中的GENMOD过程和SAS语言编写的程序实现上述过程，并通过实例说明其应用效果。结果 利用ROC曲线方程，可以扣除协变量对诊断试验结果评价的影响，并能够计算出在不同协变量取值下的ROC曲线下面积，提供更为丰富和可靠的信息。结论 文中给出的ROC回归模型可以有效地用于连续变量诊断试验数据的ROC分析。     

应用SAS软件的动态数据交换技术自动生成统计报告

一般而言，SAS的输出结果要经过编辑才能成为符合统计学要求的统计图表，而制作统计报告的工作量极大，编辑过程也很易出错。如何利用SAS直接产生用户要求格式的统计报告，是值得研究的问题。dynamic data exchange(DDE)，即动态数据交换技术。SAS的DDE技术可编写SAS程序，把SAS的分析结果输出到EXCEL中，并可以控制：EXCEL的文件打开、关闭、保存、设置字体和框线等操作，使得统计分析报告可以全自动地产生。以下就用2个实例说明如何利用DDE产生统计表和统计图。     

集装箱检验检疫风险分析系统的设计

目的采用国际通用的SAS统计分析平台编制计算机程序,实现进出境集装箱检验检疫风险分析与评估。方法采用SAS编程平台适用性评估、集装箱业务流程分析、软件应用分析等环节确立应用SAS统计分析平台编制集装箱检验检疫风险分析系统的可行性,并举例说明实现部分代表性模块的程序代码。结果 SAS编程平台适合进出境集装箱检验检疫风险分析系统的研制与开发,该设计的风险分析系统涵盖数据管理层、核心应用层的6个子模块,实现15项功能。结论基于SAS的计算机辅助集装箱风险分析系统可以有效加强出入境检验检疫机构对集装箱的检验检疫与监督管理,为防控外来疫情及有毒有害物质传入传出提供智能决策支持。     

Marks, scores and grades: scaling and aggregating student assessment outcomes

The term marks conflates the concepts of scores (raw test performance) and grades (level of performance). Neither scores nor grades represent interval scales, and therefore properly speaking arithmetic means should not be calculated during aggregation. The distributions of scores from a variety of kinds of assessment are considered, and ways of converting scores to grades are discussed. Methods of aggregation are also considered, and several strategies for implementing these via spreadsheets are made available. It is recommended that: 1 Scores should always be converted to grades before aggregation. The process of converting scores to grades requires both subject-specific skills, and familiarity with educational principles. 2 Whatever grade scale is used, it should be readily distinguishable from scores. 3 The median should be calculated as the measure of overall performance, not the arithmetic mean. 4 The interquartile range should be calculated as the measure of dispersion. 5 Students should be informed of both their score and grade for each assessment. 6 Where possible, assessment should report performance by individual assessment, not by a single aggregated mark. 7 When aggregation takes place, it should be possible to aggregate student performance by type of assessment as well as by academic subject. 8 Students who perform inconsistently should receive particular scrutiny during assessment.     

分类树中QUEST算法与多水平logistic模型的联合应用与比较

目的探讨分类树中的QUEST算法和多水平logistic模型联合应用于分析层次资料的可行性.方法首先介绍模型的原理,据此提出联合应用的具体思路,然后采用分析实例进行深入探讨,并用预测准确率和ROC曲线对模型拟合效果加以比较.结果QUEST算法能非常准确地搜索出主要影响因素,并能更加深入地刻画因变量、自变量间的复杂联系,在此基础上再进行多水平logistic模型会更加高效、准确.结论QUEST算法和多水平logistic模型各有优势,将它们联合用于层次资料的分析将更能保证分析的正确性和完善性.     

The Safe Dates program: 1-year follow-up results

OBJECTIVES: An earlier report described desirable 1-month follow-up effects of the Safe Dates program on psychological, physical, and sexual dating violence. Mediators of the program-behavior relationship also were identified. The present report describes the 1-year follow-up effects of the Safe Dates program. METHODS: Fourteen schools were in the randomized experiment. Data were gathered by questionnaires in schools before program activities and 1 year after the program ended. RESULTS: The short-term behavioral effects had disappeared at 1 year, but effects on mediating variables such as dating violence norms, conflict management skills, and awareness of community services for dating violence were maintained. CONCLUSIONS: The findings are considered in the context of why program effects might have decayed and the possible role of boosters for effect maintenance.     

基因表达数据的随机森林逐步判别分析方法

目的给出一种新的随机森林算法,它能在建模过程中自动对变量进行筛选,建立“最优”判断模型。方法采用变量重要性评分和逐步迭代算法选择有作用的变量;通过实际基因表达数据考核其应用效果,并使用R语言编程做模拟试验验证其有效性。结果三种疾病基因表达数据的判别模型,在包含很少量的基因情况下便获得了理想的分类效果;模拟试验则显示在类间区分度较大的情况下,随机森林逐步判别分析的效果明显,能有效地将有作用的变量保留在模型中,提高模型的判别效果;在类间区分度不够大的情况下分类效果提高不明显。结论随机森林逐步判别分析可以有效地应用于基因表达数据的基因筛选和分类研究,但要特别注意由随机波动对分析结果造成的影响。     

肝功能检验建议性报告软件临床分析管理的优势

介绍一种方便于临床分析、鉴别诊断和分析管理的临床检验专家系统。利用人工神经网络（Artificial Neural.Networks,ANN）挖掘患者基本信息和实验室数据,进行综合统计分析,对已确诊的患者（肝炎、肝癌、肝硬化和胆囊结石）的生化检验项目进行梳理,得出相应疾病生化指标的临床阳性预测值,然后以各项生化指标为多因素变量、诊断为输出变量建立ANN预测模型。另外抽取肝功异常并已明确诊断的患者60例进行生化指标测定,利用ANN系统,综合分析预测临床符合率,同时构建对初诊病人的树状筛查程序和直观判读报告软件的开发。利用专家诊断系统对肝脏疾病预测的准确率分别为：肝炎80.0%、肝硬化86.7%、肝癌66.7%和胆囊结石73.3%,优化后的报告格式直观反映患者的病情变化。借助于建议性报告软件可快速简易地作出对肝胆系统疾病鉴别诊断,达到医疗资源优化利用的目的。     

肿瘤危险因素的筛选与多重共线性的诊断和处理

目的 结合肺癌危险因素研究中变量的筛选过程,探讨在涉及较多自变量的大型多回归分析中,变量间多重共线性的诊断和处理法,方法 首先将经单因素分析筛选的变量进行相关分析,得出相关系数矩阵Ｒ的特征值,用主成分分析法判定自变量间是否存在多重共线性以及存在几个多重共线性关系。然后将这些自变量进行正交旋转,取得旋转后公因子所对应的自变量及其多重共线性关系,结合专业知识和以往研究的经验加以去除。结果 将去除重共线     

A profile of the Medicare Current Beneficiary Survey

This article presents the logic, methods, and capabilities of a major new source of data on the Medicare population, the Medicare Current Beneficiary Survey (MCBS). The survey originated from the need to provide valid estimates of various kinds of health care spending, such as long-term care spending or expenditures by different age groups, to describe the effects of the Medicare program on its beneficiaries, and to model the effects of proposed program changes. Presented here is an account of the MCBS sampling and data collection design and the analytic strengths of the resulting data. Of special interest are the use of Computer-Assisted Personal Interviewing (CAPI); sampling from Medicare enrollment files; design for both cross-sectional and longitudinal analysis; surveying both community and facility residents; and merging survey and administrative data.     

An alternative approach to analyzing occupational mortality data

It is widely recognized that analyzing occupational mortality by calculating standardized mortality ratios based on death rates from the general population is subject to a number of limitations. An alternative approach described in this report takes advantage of the fact that comparisons of mortality by subgroups and assessments of trends in mortality are often of equal or greater interest than overall assessments and that such comparisons do not require an external control. A computer program is available for performing the needed calculations for several diseases. This program provides tests for differences and trends among subcategories defined by variables such as length of employment, job category, or exposure measurements and also provides control for age, calendar year, and several other potentially confounding variables.     

A multivariate analysis of the attitudinal and perceptual determinants of completion of a weight-reduction program

Reducing attrition in programs, especially in weight-reduction programs, is a major challenge for nutritionists. The purpose of this study was to investigate the determinants of completion of a weight-reduction program and to provide suggestions for retaining participants. A model that explains attrition behavior was tested on a group of 309 participants in a comprehensive 10-week weight-reduction program. The dependent variable was program completion. The independent variables were program, personal, attitudinal, environmental, demographic, and intention. A principal-components factor analysis was used to determine the structure of the variables related to program completion. Six major factors were identified: self-assurance, social ease, social support, perceived importance, program satisfaction and demographic variables. Stepwise multiple-regression analysis determined the factors that most effectively predicted program completion. Path analysis explored the direct and indirect effects on program completion. The results indicated that self-assurance was the most significant factor influencing program completion. Program satisfaction did not have a direct effect on program completion when considered in the presence of the other independent variables. It is suggested that to retain participants in weight-control programs, weight-control educators should develop participants' self-assurance, social support and social ease, and help participants perceive the program as important to their health and well-being.     

Predicting dissemination of a disaster mental health "Train-the-Trainer" program

Objective: Disaster mental health (DMH) is vital to comprehensive disaster preparedness for communities. A train-the-trainer (TTT) model is frequently used in public health to disseminate knowledge and skills to communities, although few studies have examined its success. We report on the development and implementation of a DMH TTT program and examine variables that predict dissemination. Methods: This secondary analysis examines 140 community-based mental health providers' participation in a TTT DMH program in 2005-2006. Instructors' dissemination of the training was followed for 12 months. Bivariate and multivariate analyses were conducted to predict dissemination of the training program. Results: Sixty percent of the trainees in the DMH TTT program conducted training programs in the 12-month period following being trained. The likelihood of conducting training programs was predicted by a self-report measure of perceptions of transfer of training. The number of individuals subsequently trained (559) was predicted by prior DMH training and sex. No other variables predicted dissemination of DMH training. Conclusions: The TTT model was moderately successful in disseminating DMH training. Intervention at?the organizational and individual level, as well as training modifications, may increase cost-effective dissemination of DMH training.