Background In many clinical trials, data are collected longitudinally over time. In such studies, missingness, in particular dropout, is an often encountered phenomenon.
Methods We discuss commonly used but often problematic methods such as complete case analysis and last observation carried forward and contrast them with broadly valid and easy to implement direct-likelihood methods. We comment on alternatives such as multiple imputation and the expectation-maximization algorithm.
Results We apply these methods in particular to data from a study with continuous outcomes. The outcomes are modelled using a general linear mixed-effects model. The bias with CC and LOCF is established in the case study and the advantages of the direct-likelihood approach shown.
Conclusions We have established formal but easy to understand arguments for a shift towards a direct-likelihood paradigm when analysing incomplete data from longitudinal clinical trials, necessitating neither imputation nor deletion.
Get full access to this article
View all access options for this article.
References
1.
1. Hill AB. Principles of medical statistics, seventh edition. London: The Lancet, 1961.
2.
2. Schwartz D, Lellouch J. Explanatory and pragmatic attitudes in therapeutic trials. Journal of Chronic Diseases1967; 20: 637–648.
3.
3. Armitage P. Attitudes in clinical trials. Statistics in Medicine1998; 17: 2675–2683.
4.
4. McMahon AD. Study control, violators, inclusion criteria and defining explanatory and pragmatic trials. Statistics in Medicine2002; 21: 1365–1376.
5.
5. Fischer K, Goetghebeur E. Structural mean effects of noncompliance: estimating interaction with baseline prognosis and selection effects. Journal of the American Statistical Association2004; 99: 918–928.
6.
6. Pocock SJ. Clinical trials: a practical approach. Chichester: John Wiley, 1983.
7.
7. Pocock SJ. Advances in Biometry. Clinical trials: a statistician's perspective. In Armitage P, David HA eds. New York: Wiley, 1996.
8.
8. Heyting A, Tolboom J, Essers J. Statistical handling of dropouts in longitudinal clinical trials. Statistics in Medicine1992; 11: 2043–2061.
9.
9. Mallinckrodt CH, Clairk WS, Carroll RJ, Molenberghs G. Assessing response profiles from incomplete longitudinal clinical trial data under regulatory considerations. Journal of Biopharmaceutical Statistics2003a; 13: 179–190.
10.
10. Mallinckrodt CH, Sanger TM, Dube S et al. Assessing and interpreting treatment effects in longitudinal clinical trials with missing data. Biological Psychiatry2003b; 53: 754–760.
11.
11. Lavori PW, Dawson R, Shera D. A multiple imputation strategy for clinical trials with truncation of patient data. Statistics in Medicine1995; 14: 1913–1925.
12.
12. Siddiqui O, Ali MW. A comparison of the random-effects pattern mixture model with last observation carried forward (LOCF) analysis in longitudinal clinical trials with dropouts. Journal of Biopharmaceutical Statistics1998; 8: 545–563.
13.
13. Verbeke G, Molenberghs G. Linear mixed models for longitudinal data. New York: Springer, 2000.
14.
14. Molenberghs G, Verbeke G. Models for discrete longitudinal data. New York: Springer, 2005.
15.
15. Ware JH. Interpreting incomplete data in studies of diet and weight loss. New England Journal of Medicine2003; 348: 2136–2137.
16.
16. Molenberghs G, Thijs H, Jansen I et al. Analyzing incomplete longitudinal clinical trial data. Biostatistics2004; 5: 445–464.
17.
17. Rubin DB. Inference and missing data. Biometrika1976; 63: 581–592.
18.
18. Little RJA, Rubin DB. Statistical analysis with missing data. New York: John Wiley & Sons, 2002.
19.
19. Gibbons RD, Hedeker D, Elkin I et al. Some conceptual and statistical issues in analysis of longitudinal psychiatric data. Archives of General Psychiatry1993; 50: 739–750.
20.
20. Mallinckrodt CH, Clark WS, Stacy RD. Type I error rates from mixed-effects model repeated measures versus fixed effects analysis of variance with missing values imputed via last observation carried forward. Drug Information Journal2001a; 4: 1215–1225.
21.
21. Mallinckrodt CH, Clark WS, Stacy RD. Accounting for dropout bias using mixed-effects models. Journal of Biopharmaceutical Statistics2001b; 1 & 2: 9–21.
22.
22. Mallinckrodt CH, Watkin JG, Molenberghs G, Carroll RJ. Choice of the primary analysis in longitudinal clinical trials. Pharmaceutical Statistics2004; 3: 161–169.
23.
23. Little RJA, Yau L. Intent-to-treat analysis in longitudinal studies with drop-outs. Biometrics1996; 52: 1324–1333.
24.
24. Liu G, Gould AL. Comparison of alternative strategies for analysis of longitudinal trials with dropouts. Journal of Biopharmaceutical Statistics2002; 12: 207–226.
25.
25. Laird NM, Ware JH. Random effects models for longitudinal data. Biometrics1982; 38: 963–974.
26.
26. Rubin DB. Multiple imputation for nonresponse in surveys. New York: John Wiley & Sons, 1987.
27.
27. Dempster AP, Laird NM, Rubin DB. Maximum likelihood from incomplete data via the EM algorithm (with discussion). Journal of the Royal Statistical Society, Series B1977; 39: 1–38.
28.
28. Potthoff RF, Roy SN. A generalized multivariate analysis of variance model useful especially for growth curve problems. Biometrika1964; 51: 313–326.
29.
29. Little RJA, Rubin, DB. Statistical analysis with missing data. New York: John Wiley, 1987.
