2025
Analysis of Cohort Stepped Wedge Cluster‐Randomized Trials With Nonignorable Dropout via Joint Modeling
Gasparini A, Crowther M, Hoogendijk E, Li F, Harhay M. Analysis of Cohort Stepped Wedge Cluster‐Randomized Trials With Nonignorable Dropout via Joint Modeling. Statistics In Medicine 2025, 44: e10347. PMID: 39963907, PMCID: PMC11833761, DOI: 10.1002/sim.10347.Peer-Reviewed Original ResearchConceptsStepped wedge cluster randomized trialDropout processNonignorable missing outcomesParallel-arm cluster-randomized trialsCluster randomized trialNonignorable dropoutsJoint longitudinal-survival modelLongitudinal submodelData-generating scenariosMissingness patternsJoint modeling methodologyCorrelation structureMonte Carlo simulationsLongitudinal outcomesJoint modelEffective parametrizationPrimary care practicesGeriatric care modelsCarlo simulationsFrail older adultsAssociation structureSubmodelsCare modelUsual careCare practicesWeighting methods for truncation by death in cluster-randomized trials
Isenberg D, Harhay M, Mitra N, Li F. Weighting methods for truncation by death in cluster-randomized trials. Statistical Methods In Medical Research 2025, 34: 473-489. PMID: 39885759, DOI: 10.1177/09622802241309348.Peer-Reviewed Original ResearchMeSH KeywordsCluster AnalysisHumansModels, StatisticalQuality of LifeRandomized Controlled Trials as TopicConceptsSurvivor average causal effectAverage causal effectCluster randomized trialAsymptotic variance estimatorsSubgroup treatment effectsCausal effectsPrincipal stratification frameworkFinite-sampleVariance estimationDistributional assumptionsIdentification assumptionsStratification frameworkPatient-centered outcomesNon-mortality outcomesOutcome modelQuality of lifeRandomized trialsIll patient populationMeasurement time pointsTruncationEstimationLength of hospital stayAssumptionsSurvivorsPatient populationAddressing selection bias in cluster randomized experiments via weighting
Papadogeorgou G, Liu B, Li F, Li F. Addressing selection bias in cluster randomized experiments via weighting. Biometrics 2025, 81: ujaf013. PMID: 40052595, DOI: 10.1093/biomtc/ujaf013.Peer-Reviewed Original ResearchMeSH KeywordsCluster AnalysisComputer SimulationHumansModels, StatisticalPatient SelectionRandomized Controlled Trials as TopicSelection BiasConceptsCluster-randomized experimentCluster randomized trialAverage treatment effectSelection biasInverse probability weightingOverall populationTreatment effectsCo-paymentControl armRecruited populationProbability weightingRandomized experimentRandomized trialsPopulationEstimation strategyTreatment assignmentIndividualsRecruitment assumptionR packageOverallAnalysis approachInterventionRecruitment
2024
How to achieve model-robust inference in stepped wedge trials with model-based methods?
Wang B, Wang X, Li F. How to achieve model-robust inference in stepped wedge trials with model-based methods? Biometrics 2024, 80: ujae123. PMID: 39499239, PMCID: PMC11536888, DOI: 10.1093/biomtc/ujae123.Peer-Reviewed Original ResearchConceptsTreatment effect estimandsWorking correlation structureSandwich variance estimatorExchangeable working correlation structureFunction of calendar timeEffect estimandsVariance estimationLink functionStepped wedge trialEstimandsTheoretical resultsCorrelation structureWedge trialsEstimating equationsCluster randomized trialG-computationLinear mixed modelsInferencePotential outcomesMisspecificationEstimationEffective structureModel-based methodsGeneralized estimating equationsMixed modelsBayesian pathway analysis over brain network mediators for survival data
Tian X, Li F, Shen L, Esserman D, Zhao Y. Bayesian pathway analysis over brain network mediators for survival data. Biometrics 2024, 80: ujae132. PMID: 39530270, PMCID: PMC11555425, DOI: 10.1093/biomtc/ujae132.Peer-Reviewed Original ResearchMeSH KeywordsAlzheimer DiseaseBayes TheoremBrainComputer SimulationHumansModels, StatisticalNerve NetNeuroimagingSurvival AnalysisConceptsAccelerated failure time modelFailure time modelBrain connectivityAlzheimer's Disease Neuroimaging Initiative studyMaximum information extractionResponse regressionBayesian approachInformation extractionTime modelSurvival dataNoisy componentsUnique edgeWhite matter fiber tractsNetwork configurationBrain networksInterconnection networksNetworkNetwork mediatorsBrainOptimal designs using generalized estimating equations in cluster randomized crossover and stepped wedge trials
Liu J, Li F. Optimal designs using generalized estimating equations in cluster randomized crossover and stepped wedge trials. Statistical Methods In Medical Research 2024, 33: 1299-1330. PMID: 38813761, DOI: 10.1177/09622802241247717.Peer-Reviewed Original ResearchMeSH KeywordsAlgorithmsCluster AnalysisCross-Over StudiesHumansModels, StatisticalRandomized Controlled Trials as TopicResearch DesignConceptsMaximin optimal designsStepped wedge cluster randomized trialLocally optimal designsCluster-period sizesClosed-form formulaCluster-randomized crossover trialCross-sectional sampling schemeInteger estimationOptimal design algorithmDesign algorithmLongitudinal cluster randomized trialsWorking correlation structureCluster randomized trialMethod of generalized estimating equationsTreatment effect estimatesSAS macroVariance expressionsExact valueCorrelation structureMaximinSampling schemeBetween-clusterOptimal designOptimization design researchEstimating equationsMultiply robust estimation of principal causal effects with noncompliance and survival outcomes
Cheng C, Guo Y, Liu B, Wruck L, Li F, Li F. Multiply robust estimation of principal causal effects with noncompliance and survival outcomes. Clinical Trials 2024, 21: 553-561. PMID: 38813813, DOI: 10.1177/17407745241251773.Peer-Reviewed Original ResearchMeSH KeywordsAspirinCardiovascular DiseasesCausalityHospitalizationHumansModels, StatisticalPragmatic Clinical Trials as TopicResearch DesignSurvival AnalysisConceptsPrincipal strataRight-censored survival outcomesPrincipal causal effectsCausal effectsSensitivity analysis strategyPrincipal ignorabilityRobust estimationIdentification assumptionsCensoringPragmatic clinical trialsTreatment assignmentTreatment noncomplianceMonotonicityEstimationAssess treatment effectsCardiovascular diseaseClinical trialsMultipliersTreatment effectsAssumptionsNoncomplianceAssessing treatment effect heterogeneity in the presence of missing effect modifier data in cluster-randomized trials
Blette B, Halpern S, Li F, Harhay M. Assessing treatment effect heterogeneity in the presence of missing effect modifier data in cluster-randomized trials. Statistical Methods In Medical Research 2024, 33: 909-927. PMID: 38567439, PMCID: PMC11041086, DOI: 10.1177/09622802241242323.Peer-Reviewed Original ResearchConceptsMultilevel multiple imputationHeterogeneous treatment effectsCluster randomized trialPotential effect modifiersMultiple imputationAssess treatment effect heterogeneityEffect modifiersTreatment effect heterogeneityComplete-case analysisMissingness mechanismIntracluster correlationSimulation studyUnder-coverageRandomized trialsEffect heterogeneityHealth StudyTreatment effectsContinuous outcomesClinical practiceImputationModel specificationMissingnessData methodsModified dataTrialsDoubly robust estimation and sensitivity analysis for marginal structural quantile models
Cheng C, Hu L, Li F. Doubly robust estimation and sensitivity analysis for marginal structural quantile models. Biometrics 2024, 80: ujae045. PMID: 38884127, DOI: 10.1093/biomtc/ujae045.Peer-Reviewed Original ResearchMeSH KeywordsAntihypertensive AgentsBiometryComputer SimulationElectronic Health RecordsHumansHypertensionModels, StatisticalConceptsQuantile modelDistribution of potential outcomesEfficient influence functionPotential outcome distributionsDoubly robust estimatorsTime-varying treatmentsSequential ignorability assumptionSemiparametric frameworkIgnorability assumptionVariance estimationOutcome distributionInfluence functionRobust estimationPotential outcomesEfficient computationFunction approachTime-varying confoundersElectronic health record dataEstimationTreatment assignmentHealth record dataEffect of antihypertensive medicationEquationsRecord dataAntihypertensive medications
2023
Designing individually randomized group treatment trials with repeated outcome measurements using generalized estimating equations
Wang X, Turner E, Li F. Designing individually randomized group treatment trials with repeated outcome measurements using generalized estimating equations. Statistics In Medicine 2023, 43: 358-378. PMID: 38009329, PMCID: PMC10939061, DOI: 10.1002/sim.9966.Peer-Reviewed Original ResearchMeSH KeywordsBiasCluster AnalysisComputer SimulationHumansModels, StatisticalResearch DesignSample SizeConceptsSample size proceduresConstant treatment effectCorrelation structureSize proceduresMarginal mean modelClosed-form sample size formulaCorrelation parametersSandwich variance estimatorGroup treatment trialsEquation approachExchangeable correlation structureSample size formulaBinary outcomesVariance estimatorEmpirical powerLinear timeMean modelCorrelation matrixDifferent correlation parametersEstimating EquationsSize formulaEquationsSample size calculationDifferent assumptionsProper sample size calculationA mixed model approach to estimate the survivor average causal effect in cluster‐randomized trials
Wang W, Tong G, Hirani S, Newman S, Halpern S, Small D, Li F, Harhay M. A mixed model approach to estimate the survivor average causal effect in cluster‐randomized trials. Statistics In Medicine 2023, 43: 16-33. PMID: 37985966, DOI: 10.1002/sim.9939.Peer-Reviewed Original ResearchORTH.Ord: An R package for analyzing correlated ordinal outcomes using alternating logistic regressions with orthogonalized residuals
Meng C, Ryan M, Rathouz P, Turner E, Preisser J, Li F. ORTH.Ord: An R package for analyzing correlated ordinal outcomes using alternating logistic regressions with orthogonalized residuals. Computer Methods And Programs In Biomedicine 2023, 237: 107567. PMID: 37207384, DOI: 10.1016/j.cmpb.2023.107567.Peer-Reviewed Original ResearchMeSH KeywordsBiasCluster AnalysisComputer SimulationHumansLogistic ModelsLongitudinal StudiesModels, StatisticalConceptsOrdinal outcomesSandwich estimatorR packageSimulation studyCorrelated ordinal dataFinite sample biasesNumber of clustersCovariance estimationMarginal modelsEquationsParameter estimatesOrdinal responsesAssociation parametersCluster associationsBias correctionOrdinal dataEstimatorEstimating EquationsNominal levelMarginal meansResidualsEstimationPairwise odds ratiosAssociation modelGEE modelAccounting for expected attrition in the planning of cluster randomized trials for assessing treatment effect heterogeneity
Tong J, Li F, Harhay M, Tong G. Accounting for expected attrition in the planning of cluster randomized trials for assessing treatment effect heterogeneity. BMC Medical Research Methodology 2023, 23: 85. PMID: 37024809, PMCID: PMC10077680, DOI: 10.1186/s12874-023-01887-8.Peer-Reviewed Original ResearchConceptsSample size methodsSample size proceduresSize proceduresTreatment effect heterogeneityHeterogeneous treatment effectsSize methodMissingness ratesSample size formulaSample size estimationMissingness indicatorsEffect heterogeneityReal-world examplesSimulation studyIntracluster correlation coefficientInflation methodSize formulaAverage treatment effectResultsSimulation resultsSample size estimatesSize estimationMissingnessSample sizeClustersEstimationFormulaA Bayesian Approach for Estimating the Survivor Average Causal Effect When Outcomes Are Truncated by Death in Cluster-Randomized Trials
Tong G, Li F, Chen X, Hirani S, Newman S, Wang W, Harhay M. A Bayesian Approach for Estimating the Survivor Average Causal Effect When Outcomes Are Truncated by Death in Cluster-Randomized Trials. American Journal Of Epidemiology 2023, 192: 1006-1015. PMID: 36799630, PMCID: PMC10236525, DOI: 10.1093/aje/kwad038.Peer-Reviewed Original ResearchGEEMAEE: A SAS macro for the analysis of correlated outcomes based on GEE and finite-sample adjustments with application to cluster randomized trials
Zhang Y, Preisser J, Li F, Turner E, Toles M, Rathouz P. GEEMAEE: A SAS macro for the analysis of correlated outcomes based on GEE and finite-sample adjustments with application to cluster randomized trials. Computer Methods And Programs In Biomedicine 2023, 230: 107362. PMID: 36709555, PMCID: PMC10037297, DOI: 10.1016/j.cmpb.2023.107362.Peer-Reviewed Original ResearchMeSH KeywordsCluster AnalysisComputer SimulationLongitudinal StudiesModels, StatisticalRandomized Controlled Trials as TopicConceptsNumber of clustersBias-corrected estimationCorrelation structurePopulation-averaged interpretationMarginal regression modelsDeletion diagnosticsEstimating EquationsFinite-sample adjustmentInfluence of observationsLarge valuesStandard errorEquationsSandwich estimatorVariance estimatorCook's distanceSAS macroDesign of clusterCount outcomesLongitudinal responseCorrelation parametersValid inferencesCorrelated outcomesFlexible specificationBiased estimatesEstimator
2022
Simulating time-to-event data subject to competing risks and clustering: A review and synthesis
Meng C, Esserman D, Li F, Zhao Y, Blaha O, Lu W, Wang Y, Peduzzi P, Greene E. Simulating time-to-event data subject to competing risks and clustering: A review and synthesis. Statistical Methods In Medical Research 2022, 32: 305-333. PMID: 36412111, PMCID: PMC11654122, DOI: 10.1177/09622802221136067.Peer-Reviewed Original ResearchIs the Product Method More Efficient Than the Difference Method for Assessing Mediation?
Cheng C, Spiegelman D, Li F. Is the Product Method More Efficient Than the Difference Method for Assessing Mediation? American Journal Of Epidemiology 2022, 192: 84-92. PMID: 35921210, PMCID: PMC10144745, DOI: 10.1093/aje/kwac144.Peer-Reviewed Original ResearchMeSH KeywordsBiomedical ResearchEpidemiologic StudiesEswatiniHumansMediation AnalysisModels, StatisticalTwo weights make a wrong: Cluster randomized trials with variable cluster sizes and heterogeneous treatment effects
Wang X, Turner EL, Li F, Wang R, Moyer J, Cook AJ, Murray DM, Heagerty PJ. Two weights make a wrong: Cluster randomized trials with variable cluster sizes and heterogeneous treatment effects. Contemporary Clinical Trials 2022, 114: 106702. PMID: 35123029, PMCID: PMC8936048, DOI: 10.1016/j.cct.2022.106702.Peer-Reviewed Original ResearchMeSH KeywordsCluster AnalysisComputer SimulationEarly Detection of CancerHumansModels, StatisticalRandomized Controlled Trials as TopicSample SizeConceptsInverse cluster sizeVariable cluster sizesCluster sizeCorrelation matrixTreatment effect estimatesCluster correlationEquation frameworkEstimation characteristicsTheoretical derivationSimulation studyAverage treatment effectHeterogeneous treatment effectsDistinct weightsEstimandsCluster levelHierarchical nestingMatrixHypothetical populationEstimatesValid resultsDerivationClustersConceptual populationEstimationEffect estimates
2021
Estimating the natural indirect effect and the mediation proportion via the product method
Cheng C, Spiegelman D, Li F. Estimating the natural indirect effect and the mediation proportion via the product method. BMC Medical Research Methodology 2021, 21: 253. PMID: 34800985, PMCID: PMC8606099, DOI: 10.1186/s12874-021-01425-4.Peer-Reviewed Original ResearchConceptsInterval estimatorsApproximate estimatorExact estimatorMultivariate delta methodFinite sample performanceProduct methodNon-negligible biasBinary outcomesRare outcome assumptionExact expressionDelta methodVariance estimationEmpirical performanceEstimatorCommon data typesBootstrap approachBinary mediatorNatural indirect effectSample sizeImpact of complex, partially nested clustering in a three-arm individually randomized group treatment trial: A case study with the wHOPE trial
Tong G, Seal KH, Becker WC, Li F, Dziura JD, Peduzzi PN, Esserman DA. Impact of complex, partially nested clustering in a three-arm individually randomized group treatment trial: A case study with the wHOPE trial. Clinical Trials 2021, 19: 3-13. PMID: 34693748, PMCID: PMC8847260, DOI: 10.1177/17407745211051288.Peer-Reviewed Original ResearchConceptsGroup treatment trialsIntraclass correlation coefficientTreatment trialsTreatment sessionsHealth optionsEducation trialThree-armWhole health teamFuture trial designNumber of cliniciansGroup treatment designTrue intraclass correlation coefficientsGroup treatment sessionsTreatment armsClinical trialsClinician levelMultiple cliniciansBACKGROUND/Health teamsOutcome dataTreatment groupsTrial designGroup educationClinical scenariosDifferent clinicians
This site is protected by hCaptcha and its Privacy Policy and Terms of Service apply