Impact of missing data on type 1 error rates in non-inferiority trials
Abstract
In this paper, a simulation study is conducted to systematically investigate the impact of different types of missing data on six different statistical analyses: four different likelihood-based linear mixed effects models and analysis of covariance (ANCOVA) using two different data sets, in non-inferiority trial settings for the analysis of longitudinal continuous data. ANCOVA is valid when the missing data are completely at random. Likelihood-based linear mixed effects model approaches are valid when the missing data are at random. Pattern-mixture model (PMM) was developed to incorporate non-random missing mechanism. Our simulations suggest that two linear mixed effects models using unstructured covariance matrix for within-subject correlation with no random effects or first-order autoregressive covariance matrix for within-subject correlation with random coefficient effects provide well control of type 1 error (T1E) rate when the missing data are completely at random or at random. ANCOVA using last observation carried forward imputed data set is the worst method in terms of bias and T1E rate. PMM does not show much improvement on controlling T1E rate compared with other linear mixed effects models when the missing data are not at random but is markedly inferior when the missing data are at random.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Linear mixed modelling is a frequently recommended method for handling missing data and correlated observations in repeated-measures data.
“Linear mixed modelling is a frequently recommended method for its abilities to handle missing data and correlated observations in repeated-measures data (Gueorguieva and Krystal, 2004; Yoo, 2010).”
From Internet-based treatment for vulvodynia (EMBLA) – Study protocol for a randomised controlled study (2021) · cites this paper as Yoo, 2010
Cites
RUBIN, D. B. (1976) · Biometrika · doi:10.1093/biomet/63.3.581 · Cited by 9,925 · see citing works
Random-Effects Models for Longitudinal Data
Laird, N. M., & Ware, J. H. (1982) · Biometrics · doi:10.2307/2529876 · Cited by 8,923 · see citing works
Statistical Analysis with Missing Data
Little, R. J. A., & Rubin, D. B. (2002) · Wiley Series in Probability and Statistics · doi:10.1002/9781119013563 · Cited by 8,473 · see citing works
Linear Mixed Models for Longitudinal Data
Molenberghs, G., & Verbeke, G. N. (2000) · Springer series in statistics · doi:10.1007/978-1-4419-0300-6 · Cited by 2,681 · see citing works
Unbalanced repeated-measures models with structured covariance matrices.
Jennrich, R., & Schluchter, M. (1986) · Biometrics · doi:10.2307/2530695 · Cited by 1,382 · see citing works
Application of random-effects pattern-mixture models for missing data in longitudinal studies.
Hedeker, D., & Gibbons, R. D. (1997) · Psychological Methods · doi:10.1037/1082-989x.2.1.64 · Cited by 907 · see citing works
Missing data in longitudinal studies
Laird, N. M. (1988) · Statistics in Medicine · doi:10.1002/sim.4780070131 · Cited by 713 · see citing works
Intention to treat – who should use ITT?
Lewis, J., & Machin, D. (1993) · British Journal of Cancer · doi:10.1038/bjc.1993.402 · Cited by 236 · see citing works
Mallinckrodt, C. H., Clark, W. S., & David, S. R. (2001) · Drug Information Journal · doi:10.1177/009286150103500418 · Cited by 160 · see citing works
Choosing the analysis population in non-inferiority studies: per protocol or intent-to-treat
Sanchez, M. M., & Chen, X. (2006) · Statistics in Medicine · doi:10.1002/sim.2244 · Cited by 147 · see citing works
Barnes, S. A., Mallinckrodt, C. H., Lindborg, S. R., & Carter, M. K. (2007) · Pharmaceutical Statistics · doi:10.1002/pst.310 · Cited by 44 · see citing works
Reporting delay and corrected incidence of multiple sclerosis
Esbjerg, S., Keiding, N., & Koch‐Henriksen, N. (1999) · Statistics in Medicine · doi:10.1002/(sici)1097-0258(19990715)18:13 · Cited by 38 · see citing works
Type I error rates from likelihood‐based repeated measures analyses of incomplete longitudinal data
Mallinckrodt, C. H., Kaiser, C. J., Watkin, J. G., Detke, M. J., Molenberghs, G., & Carroll, R. J. (2004) · Pharmaceutical Statistics · doi:10.1002/pst.131 · Cited by 36 · see citing works
Park, T., Park, J.-K., & Davis, C. S. (2001) · Statistics in Medicine · doi:10.1002/sim.859 · Cited by 19 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Andersson, G., Hesser, H., Veilord, A., Svedling, L., Andersson, F., Sleman, O., Mauritzson, L., Sarkohi, A., Claesson, E., Zetterqvist, V., et al. (2013) · Journal of Affective Disorders · doi:10.1016/j.jad.2013.08.022 · Cited by 195 · see citing works
Ly, K. H., Topooco, N., Cederlund, H., Wallin, A., Bergström, J., Molander, O., Carlbring, P., & Andersson, G. (2015) · PLOS ONE · doi:10.1371/journal.pone.0126559 · Cited by 171 · see citing works
Blom, K., Tillgren, H. T., Wiklund, T., Danlycke, E., Forssén, M., Söderström, A., Johansson, R., Hesser, H., Jernelöv, S., Lindefors, N., et al. (2015) · Behaviour Research and Therapy · doi:10.1016/j.brat.2015.05.002 · Cited by 172 · see citing works
Garamszegi, L. Z., & Møller, A. P. (2011) · Systematic Biology · doi:10.1093/sysbio/syr060 · Cited by 89 · see citing works
Missing Data in Noninferiority Trials
Wiens, B. L., & Rosenkranz, G. K. (2013) · Statistics in Biopharmaceutical Research · doi:10.1080/19466315.2013.847383 · Cited by 11 · see citing works
Beukes, E. W., Baguley, D., Allen, P. M., Manchaiah, V. K. C., & Andersson, G. (2017) · Trials · doi:10.1186/s13063-017-1931-6 · Cited by 9 · see citing works
Internet-based treatment for vulvodynia (EMBLA) – Study protocol for a randomised controlled study
Hess Engström, A., Kullinger, M., Jawad, I., Hesselman, S., Buhrman, M., Högberg, U., & Skalkidou, A. (2021) · Internet Interventions · doi:10.1016/j.invent.2021.100396 · Cited by 7 · see citing works
“handle missing data and correlated observations in repeated-measures data (Gueorguieva and Krystal, 2004; Yoo, 2010).”
MethodeCBT Versus Standard Individual CBT for Paediatric Obsessive–Compulsive Disorder
Babiano-Espinosa, L., Skarphéðinsson, G., Weidle, B., Wolters, L. H., Compton, S. N., Ivarsson, T., & Skokauskas, N. (2022) · Child Psychiatry & Human Development · doi:10.1007/s10578-022-01350-7 · Cited by 7 · see citing works
Sidi, Y., & Harel, O. (2020) · Contemporary Clinical Trials Communications · doi:10.1016/j.conctc.2020.100567 · Cited by 4 · see citing works
Rabe, B. A., & Bell, M. L. (2019) · Statistics in Biopharmaceutical Research · doi:10.1080/19466315.2019.1677493 · Cited by 2 · see citing works
Controversial (Unresolved) Issues in Noninferiority Trials
Wiens, B. L. (2014) · Chapman & Hall/CRC eBooks · doi:10.1201/b17716-7
- Authors
- Yoo, B.
- Affiliations
- Bristol-Myers Squibb (United States) (US)
- Corresponding author
- Yoo, Bongin
- Venue
- Pharmaceutical Statistics
- Publisher
- Wiley
- Published
- 2010-04
- Volume
- 9
- Issue
- 2
- Pages
- 87-99
- ISSN
- 1539-1604 (ISSN-L)
- 1539-1612
- OpenAlex
- W2073148597
- Cited by
- 12
- References
- 15
- Citations by year
- 2022: 1 · 2021: 1 · 2020: 1 · 2019: 1 · 2017: 1 · 2015: 2 · 2014: 1 · 2013: 2 · 2012: 1
- FWCI
- 0.55
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Pharmaceutical Statistics (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Clinical Trials
- Health Sciences › Medicine › Pediatrics, Perinatology and Child Health › Pharmaceutical studies and practices
- Keywords
- missing data, type I error rate, linear mixed models, covariance analysis, pattern-mixture models, missing at random, last observation carried forward, within-subject correlation
- MeSH terms
- Clinical Trials as Topic, Computer Simulation, Data Interpretation, Statistical, Lactones, Longitudinal Studies, Therapeutic Equivalency, Models, Statistical, Randomized Controlled Trials as Topic
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
