Aggregating and Analyzing Daily Drinking Data in Clinical Trials: A Comparison of Type I Errors, Power, and Bias.

<bold>Objective: </bold>Statistical analyses in alcohol clinical trials often use longitudinal daily drinking data (e.g., percentage of drinking days) to test treatment efficacy. Such data can be aggregated and analyzed in many ways. To assess how statistical analytic decisions may influence substan...

Full description

Bibliographic Details
Published in:Journal of Studies on Alcohol & Drugs Vol. 77; no. 6; pp. 986 - 992
Main Authors: Hallgren, Kevin A., Atkins, David C., Witkiewitz, Katie
Format: journal article
Published: Alcohol Research Documentation Nov2016
Subjects:
Online Access:View this record in EBSCOhost
fields @attributes:
  recordID: 1
pdfLink:
plink: https://search.ebscohost.com/login.aspx?direct=true&db=ssf&AN=119250639&site=ehost-live
header:
  @attributes:
    shortDbName: ssf
    uiTerm: 119250639
    longDbName: Social Sciences Full Text (H.W. Wilson)
    uiTag: AN
  controlInfo:
    bkinfo:
    jinfo:
      jid:
        19371888
        30V1
      jtl: Journal of Studies on Alcohol & Drugs
      issn: 19371888
      maglogo: N
    pubinfo:
      dt: Nov2016
      vid: 77
      iid: 6
      pid: 343
      pub: Alcohol Research Documentation
    artinfo:
      ui:
        119250639
        10.15288/jsad.2016.77.986
      ppf: 986
      ppct: 6
      formats:
      tig:
        atl: Aggregating and Analyzing Daily Drinking Data in Clinical Trials: A Comparison of Type I Errors, Power, and Bias.
      aug:
        au:
          Hallgren, Kevin A.
          Atkins, David C.
          Witkiewitz, Katie
        affil:
          Department of Psychiatry and Behavioral Sciences, University of Washington, Seattle, Washington
          Department of Psychology, Center on Alcoholism, Substance Abuse, and Addictions, University of New Mexico, Albuquerque, New Mexico
      su:
        Clinical trials
        Alcohol drinking
        Comparative studies
        Longitudinal method
        Research methodology
        Medical cooperation
        Regression analysis
        Research
        Research funding
        Statistics
        Data analysis
        Evaluation research
        Research bias
      sug:
        subj:
          Clinical trials
          Alcohol drinking
          Research and Development in the Physical, Engineering, and Life Sciences (except Biotechnology)
          Drinking Places (Alcoholic Beverages)
          Comparative studies
          Longitudinal method
          Research methodology
          Medical cooperation
          Regression analysis
          Research
          Research funding
          Statistics
          Data analysis
          Evaluation research
          Research bias
      ab: <bold>Objective: </bold>Statistical analyses in alcohol clinical trials often use longitudinal daily drinking data (e.g., percentage of drinking days) to test treatment efficacy. Such data can be aggregated and analyzed in many ways. To assess how statistical analytic decisions may influence substantive results, the current report compares different aggregation methods (e.g., computing percentages of drinking days vs. using daily binary indicators of drinking) and statistical methods (i.e., least squares regression, linear mixed-effects models [LMM], generalized linear mixed models [GLMM], and generalized estimating equations [GEE]) for testing the effects of treatment on drinking outcomes in clinical trials.<bold>Method: </bold>A simulation study repeatedly resampled daily drinking data from the treatment period of the Combined Pharmacotherapies and Behavioral Interventions for Alcohol Dependence (COMBINE) Study at different sample sizes. Treatment effects in each data set were modeled using different aggregation and statistical methods.<bold>Results: </bold>Type I error rates were near the expected rate for most models but on occasion were mildly elevated when disaggregated daily drinking data were analyzed using GLMM or GEE with an exchangeable correlation structure. Most methods yielded similar statistical power, although power decreased when modeling disaggregated daily drinking with GLMM and had mixed increases and decreases when the longitudinal nature of data was ignored by using fully aggregated data with independent samples t tests.<bold>Conclusions: </bold>When testing treatment main effects, relatively simpler statistical methods with fewer repeated measures may perform equally well or better than more complicated methods. Patterns of significance and treatment effect size estimates are likely comparable across most studies that use different aggregation and statistical methods, but differences between these methods may occasionally have an important impact on conclusions in clinical trials.
      pubtype: Academic Journal
      doctype: journal article
      src: R
    language: English
    refInfo:
    copyright:
      @attributes:
        flag: N
    holdings:
      @attributes:
        islocal: N