Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing data
Abstract Background The primary analysis in a longitudinal randomized controlled trial is sometimes a comparison of arms at a single time point. While a two-sample t-test is often used, missing data are common in longitudinal studies and decreases power by reducing sample size. Mixed models for repe...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2016-04-01
|
| Series: | BMC Medical Research Methodology |
| Subjects: | |
| Online Access: | http://link.springer.com/article/10.1186/s12874-016-0144-0 |
| _version_ | 1819040987079507968 |
|---|---|
| author | Erin L. Ashbeck Melanie L. Bell |
| author_facet | Erin L. Ashbeck Melanie L. Bell |
| author_sort | Erin L. Ashbeck |
| collection | DOAJ |
| description | Abstract Background The primary analysis in a longitudinal randomized controlled trial is sometimes a comparison of arms at a single time point. While a two-sample t-test is often used, missing data are common in longitudinal studies and decreases power by reducing sample size. Mixed models for repeated measures (MMRM) can test treatment effects at specific time points, have been shown to give unbiased estimates in certain missing data contexts, and may be more powerful than a two sample t-test. Methods We conducted a simulation study to compare the performance of a complete-case t-test to a MMRM in terms of power and bias under different missing data mechanisms. Impact of within- and between-person variance, dropout mechanism, and variance-covariance structure were all considered. Results While both complete-case t-test and MMRM provided unbiased estimation of treatment differences when data were missing completely at random, MMRM yielded an absolute power gain of up to 12 %. The MMRM provided up to 25 % absolute increased power over the t-test when data were missing at random, as well as unbiased estimation. Conclusions Investigators interested in single time point comparisons should use a MMRM with a contrast to gain power and unbiased estimation of treatment effects instead of a complete-case two sample t-test. |
| first_indexed | 2024-12-21T09:17:49Z |
| format | Article |
| id | doaj.art-b654149f6d404c1bade1d08bb07636db |
| institution | Directory Open Access Journal |
| issn | 1471-2288 |
| language | English |
| last_indexed | 2024-12-21T09:17:49Z |
| publishDate | 2016-04-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Research Methodology |
| spelling | doaj.art-b654149f6d404c1bade1d08bb07636db2022-12-21T19:09:06ZengBMCBMC Medical Research Methodology1471-22882016-04-011611810.1186/s12874-016-0144-0Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing dataErin L. Ashbeck0Melanie L. Bell1Department of Epidemiology and Biostatistics, University of ArizonaDepartment of Epidemiology and Biostatistics, University of ArizonaAbstract Background The primary analysis in a longitudinal randomized controlled trial is sometimes a comparison of arms at a single time point. While a two-sample t-test is often used, missing data are common in longitudinal studies and decreases power by reducing sample size. Mixed models for repeated measures (MMRM) can test treatment effects at specific time points, have been shown to give unbiased estimates in certain missing data contexts, and may be more powerful than a two sample t-test. Methods We conducted a simulation study to compare the performance of a complete-case t-test to a MMRM in terms of power and bias under different missing data mechanisms. Impact of within- and between-person variance, dropout mechanism, and variance-covariance structure were all considered. Results While both complete-case t-test and MMRM provided unbiased estimation of treatment differences when data were missing completely at random, MMRM yielded an absolute power gain of up to 12 %. The MMRM provided up to 25 % absolute increased power over the t-test when data were missing at random, as well as unbiased estimation. Conclusions Investigators interested in single time point comparisons should use a MMRM with a contrast to gain power and unbiased estimation of treatment effects instead of a complete-case two sample t-test.http://link.springer.com/article/10.1186/s12874-016-0144-0Complete-caseLongitudinalMean response profileMissing dataMixed modelPower |
| spellingShingle | Erin L. Ashbeck Melanie L. Bell Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing data BMC Medical Research Methodology Complete-case Longitudinal Mean response profile Missing data Mixed model Power |
| title | Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing data |
| title_full | Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing data |
| title_fullStr | Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing data |
| title_full_unstemmed | Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing data |
| title_short | Single time point comparisons in longitudinal randomized controlled trials: power and bias in the presence of missing data |
| title_sort | single time point comparisons in longitudinal randomized controlled trials power and bias in the presence of missing data |
| topic | Complete-case Longitudinal Mean response profile Missing data Mixed model Power |
| url | http://link.springer.com/article/10.1186/s12874-016-0144-0 |
| work_keys_str_mv | AT erinlashbeck singletimepointcomparisonsinlongitudinalrandomizedcontrolledtrialspowerandbiasinthepresenceofmissingdata AT melanielbell singletimepointcomparisonsinlongitudinalrandomizedcontrolledtrialspowerandbiasinthepresenceofmissingdata |