Applied Medical Statistics Using SAS

Written with medical statisticians and medical researchers in mind, this intermediate-level reference explores the use of SAS for analyzing medical data. Applied Medical Statistics Using SAS covers the whole range of modern statistical methods used in the analysis of medical data, including regression, analysis of variance and covariance, longitudinal and survival data analysis, missing data, generalized additive models (GAMs), and Bayesian methods. The book focuses on performing these analyses using SAS, the software package of choice for those analysing medical data. Features Covers the planning stage of medical studies in detail; several chapters contain details of sample size estimation Illustrates methods of randomisation that might be employed for clinical trials Covers topics that have become of great importance in the 21st century, including Bayesian methods and multiple imputation Its breadth and depth, coupled with the inclusion of all the SAS code, make this book ideal for practitioners as well as for a graduate class in biostatistics or public health. Complete data sets, all the SAS code, and complete outputs can be found on an associated website: http://support.sas.com/amsus

An Introduction to SASIntroductionThe User InterfaceSAS ProgramsReading Data—The Data StepModifying SAS DataThe Proc StepGlobal StatementsSAS GraphicsODS—The Output Delivery SystemSaving Output in SAS Data Sets—ods outputEnhancing OutputSAS MacrosSome Tips for Preventing and Correcting ErrorsStatistics and Measurement in MedicineIntroductionA Brief History of Medical StatisticsMeasurement in MedicineAssessing Bias and Reliability of MeasurementsDiagnostic TestsSummaryClinical TrialsIntroductionClinical TrialsHow Many Participants Do I Need in My Trial?The Analysis of Data from Clinical TrialsSummaryEpidemiologyIntroductionTypes of Epidemiological StudyRelative Risk and Odds RatiosSample Size Estimation for Epidemiologic StudiesSimple Analyses for Data from Observational StudiesSummaryMeta-analysisIntroductionStudy SelectionPublication BiasThe Statistics of Meta-analysisAn Example of the Application of Meta-analysisMeta-analysis on Sparse DataMetaregressionSummaryAnalysis of Variance and CovarianceIntroductionA Simple Example of One-Way Analysis of VarianceMultiple Comparison ProceduresA Factorial ExperimentUnbalanced DesignsNonparametric Analysis of VarianceAnalysis of CovarianceSummaryScatter Plots, Correlation, Simple Regression, and SmoothingIntroductionThe Scatter Plot and Correlation CoefficientSimple Linear Regression and Locally Weighted RegressionLocally Weighted RegressionThe Aspect Ratio of a Scatter PlotEstimating Bivariate DensitiesScatter Plot MatricesSummaryMultiple Linear RegressionIntroductionThe Multiple Linear Regression ModelSome Examples of the Application of the Multiple Linear Regression ModelIdentifying a Parsimonious ModelChecking Model Assumptions: Residuals and OtherRegression DiagnosticsThe General Linear ModelSummaryLogistic RegressionIntroductionLogistic RegressionTwo Examples of the Application of Logistic RegressionDiagnosing a Logistic Regression ModelLogistic Regression for 1:1 Matched StudiesPropensity ScoresSummaryThe Generalised Linear ModelIntroductionGeneralised Linear ModelsApplying the Generalised Linear ModelResiduals for GLMsOverdispersionSummaryGeneralised Additive ModelsIntroductionScatter Plot SmoothersAdditive and Generalised Additive ModelsExamples of the Application of GAMsSummaryThe Analysis of Longitudinal Data IIntroductionGraphical Displays of Longitudinal DataSummary Measure Analysis of Longitudinal DataSummary Measure Approach for Binary ResponsesSummaryThe Analysis of Longitudinal Data II: Linear Mixed-Effects Models for Normal Response VariablesIntroductionLinear Mixed-Effects Models for Repeated Measures DataDropouts in Longitudinal DataSummaryThe Analysis of Longitudinal Data III: Non-Normal ResponsesIntroductionMarginal Models and Conditional ModelsAnalysis of the Respiratory DataAnalysis of Epilepsy DataSummarySurvival AnalysisIntroductionThe Survivor Function and the Hazard FunctionComparing Groups of Survival TimesSample Size EstimationSummaryCox’s Proportional Hazards Models for Survival DataIntroductionModelling the Hazard Function: Cox’s RegressionTime-Varying CovariatesRandom-Effects Models for Survival DataSummaryBayesian MethodsIntroductionBayesian EstimationMarkov Chain Monte CarloPrior DistributionsModel Selection When Using a Bayesian ApproachSome Examples of the Application of Bayesian StatisticsSummaryMissing ValuesIntroductionPatterns of Missing DataMissing Data MechanismsExploring MissingnessDealing with Missing ValuesImputing Missing ValuesAnalysing Multiply Imputed DataSome Examples of the Application of Multiple ImputationSummaryReferences

Geoff Der, Brian S. Everitt