Problems in statistical analysis of attrition in randomized controlled clinical trials of antidepressant for geriatric depression

Moonseong Heo, Andrew C. Leon, Barnett S. Meyers, George S. Alexopoulos

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Attrition from clinical trials is unavoidable in geriatric psychiatry and beyond. It results in incomplete data and consequently imposes three fundamental challenges: greater bias, reduced power, and less generalizability. In an effort to assess the extent of attrition and the relevance of statistical methods applied to analyze incomplete data in geriatric psychiatry, we reviewed 69 published antidepressant randomized clinical trials conducted since 1975. The median attrition rate estimated from these trials was 26.6%; nevertheless, we found that many trials lack data analytic strategies to address the problem of attrition. Most of the applied statistical analyses involved chi-square tests, t-tests, and analysis of variance (ANOVA), each of which assume that data are missing completely at random. Even when imputation for missing data due to attrition was attempted, only the last observation carried forward (LOCF) method was implemented. The LOCF imputation can actually increase bias of the results in the analysis of repeatedly measured outcomes. In addition, despite the longitudinal nature of repeatedly measured outcomes, the statistical methods used are for analysis of cross-sectional data. Thus, the data analytic strategies did not adequately meet the challenges arising from attrition. We encourage the use of mixed-effects models to reduce the impact of attrition on bias, power and generalizability in antidepressant RCTs for geriatric depression. For imputation, we recommend use of multiple imputation methods instead of LOCF.

Original languageEnglish (US)
Pages (from-to)178-185
Number of pages8
JournalCurrent Psychiatry Reviews
Volume3
Issue number3
DOIs
StatePublished - Aug 2007
Externally publishedYes

Fingerprint

Geriatrics
Antidepressive Agents
Randomized Controlled Trials
Depression
Geriatric Psychiatry
Observation
Chi-Square Distribution
Analysis of Variance
Cross-Sectional Studies
Clinical Trials
Power (Psychology)

Keywords

  • ANOVA
  • Hamilton Rating Scale of Depression
  • Intent-to-treat sample
  • Last observation carried forward method
  • Montgomery-Asberg Depression Rating Scale

ASJC Scopus subject areas

  • Psychiatry and Mental health

Cite this

Problems in statistical analysis of attrition in randomized controlled clinical trials of antidepressant for geriatric depression. / Heo, Moonseong; Leon, Andrew C.; Meyers, Barnett S.; Alexopoulos, George S.

In: Current Psychiatry Reviews, Vol. 3, No. 3, 08.2007, p. 178-185.

Research output: Contribution to journalArticle

Heo, Moonseong ; Leon, Andrew C. ; Meyers, Barnett S. ; Alexopoulos, George S. / Problems in statistical analysis of attrition in randomized controlled clinical trials of antidepressant for geriatric depression. In: Current Psychiatry Reviews. 2007 ; Vol. 3, No. 3. pp. 178-185.
@article{b6f39acf923e4f289665605b113b56fd,
title = "Problems in statistical analysis of attrition in randomized controlled clinical trials of antidepressant for geriatric depression",
abstract = "Attrition from clinical trials is unavoidable in geriatric psychiatry and beyond. It results in incomplete data and consequently imposes three fundamental challenges: greater bias, reduced power, and less generalizability. In an effort to assess the extent of attrition and the relevance of statistical methods applied to analyze incomplete data in geriatric psychiatry, we reviewed 69 published antidepressant randomized clinical trials conducted since 1975. The median attrition rate estimated from these trials was 26.6{\%}; nevertheless, we found that many trials lack data analytic strategies to address the problem of attrition. Most of the applied statistical analyses involved chi-square tests, t-tests, and analysis of variance (ANOVA), each of which assume that data are missing completely at random. Even when imputation for missing data due to attrition was attempted, only the last observation carried forward (LOCF) method was implemented. The LOCF imputation can actually increase bias of the results in the analysis of repeatedly measured outcomes. In addition, despite the longitudinal nature of repeatedly measured outcomes, the statistical methods used are for analysis of cross-sectional data. Thus, the data analytic strategies did not adequately meet the challenges arising from attrition. We encourage the use of mixed-effects models to reduce the impact of attrition on bias, power and generalizability in antidepressant RCTs for geriatric depression. For imputation, we recommend use of multiple imputation methods instead of LOCF.",
keywords = "ANOVA, Hamilton Rating Scale of Depression, Intent-to-treat sample, Last observation carried forward method, Montgomery-Asberg Depression Rating Scale",
author = "Moonseong Heo and Leon, {Andrew C.} and Meyers, {Barnett S.} and Alexopoulos, {George S.}",
year = "2007",
month = "8",
doi = "10.2174/157340007781369694",
language = "English (US)",
volume = "3",
pages = "178--185",
journal = "Current Psychiatry Reviews",
issn = "1573-4005",
publisher = "Bentham Science Publishers B.V.",
number = "3",

}

TY - JOUR

T1 - Problems in statistical analysis of attrition in randomized controlled clinical trials of antidepressant for geriatric depression

AU - Heo, Moonseong

AU - Leon, Andrew C.

AU - Meyers, Barnett S.

AU - Alexopoulos, George S.

PY - 2007/8

Y1 - 2007/8

N2 - Attrition from clinical trials is unavoidable in geriatric psychiatry and beyond. It results in incomplete data and consequently imposes three fundamental challenges: greater bias, reduced power, and less generalizability. In an effort to assess the extent of attrition and the relevance of statistical methods applied to analyze incomplete data in geriatric psychiatry, we reviewed 69 published antidepressant randomized clinical trials conducted since 1975. The median attrition rate estimated from these trials was 26.6%; nevertheless, we found that many trials lack data analytic strategies to address the problem of attrition. Most of the applied statistical analyses involved chi-square tests, t-tests, and analysis of variance (ANOVA), each of which assume that data are missing completely at random. Even when imputation for missing data due to attrition was attempted, only the last observation carried forward (LOCF) method was implemented. The LOCF imputation can actually increase bias of the results in the analysis of repeatedly measured outcomes. In addition, despite the longitudinal nature of repeatedly measured outcomes, the statistical methods used are for analysis of cross-sectional data. Thus, the data analytic strategies did not adequately meet the challenges arising from attrition. We encourage the use of mixed-effects models to reduce the impact of attrition on bias, power and generalizability in antidepressant RCTs for geriatric depression. For imputation, we recommend use of multiple imputation methods instead of LOCF.

AB - Attrition from clinical trials is unavoidable in geriatric psychiatry and beyond. It results in incomplete data and consequently imposes three fundamental challenges: greater bias, reduced power, and less generalizability. In an effort to assess the extent of attrition and the relevance of statistical methods applied to analyze incomplete data in geriatric psychiatry, we reviewed 69 published antidepressant randomized clinical trials conducted since 1975. The median attrition rate estimated from these trials was 26.6%; nevertheless, we found that many trials lack data analytic strategies to address the problem of attrition. Most of the applied statistical analyses involved chi-square tests, t-tests, and analysis of variance (ANOVA), each of which assume that data are missing completely at random. Even when imputation for missing data due to attrition was attempted, only the last observation carried forward (LOCF) method was implemented. The LOCF imputation can actually increase bias of the results in the analysis of repeatedly measured outcomes. In addition, despite the longitudinal nature of repeatedly measured outcomes, the statistical methods used are for analysis of cross-sectional data. Thus, the data analytic strategies did not adequately meet the challenges arising from attrition. We encourage the use of mixed-effects models to reduce the impact of attrition on bias, power and generalizability in antidepressant RCTs for geriatric depression. For imputation, we recommend use of multiple imputation methods instead of LOCF.

KW - ANOVA

KW - Hamilton Rating Scale of Depression

KW - Intent-to-treat sample

KW - Last observation carried forward method

KW - Montgomery-Asberg Depression Rating Scale

UR - http://www.scopus.com/inward/record.url?scp=34548336388&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548336388&partnerID=8YFLogxK

U2 - 10.2174/157340007781369694

DO - 10.2174/157340007781369694

M3 - Article

AN - SCOPUS:34548336388

VL - 3

SP - 178

EP - 185

JO - Current Psychiatry Reviews

JF - Current Psychiatry Reviews

SN - 1573-4005

IS - 3

ER -