Biostatistics (2004), 5, 3, pp. 445-464
Biostatistics Vol. 5 No. 3 © Oxford University Press 2004; all rights reserved.
Analyzing incomplete longitudinal clinical trial data
Center for Statistics, Limburgs Universitair Centrum, Universitaire Campus, B-3590 Diepenbeek, Belgium
geert.molenberghs{at}luc.ac.be
London School of Hygiene and Tropical Medicine, London, UK
Eli Lilly and Company, Indianapolis, IN, USA
Department of Statistics, Texas A & M University, College Station, TX, USA
* To whom corespondence should be addressed.
Using standard missing data taxonomy, due to Rubin and co-workers, and simple algebraic derivations, it is argued that some simple but commonly used methods to handle incomplete longitudinal clinical trial data, such as complete case analyses and methods based on last observation carried forward, require restrictive assumptions and stand on a weaker theoretical foundation than likelihood-based methods developed under the missing at random (MAR) framework. Given the availability of flexible software for analyzing longitudinal sequences of unequal length, implementation of likelihood-based MAR analyses is not limited by computational considerations. While such analyses are valid under the comparatively weak assumption of MAR, the possibility of data missing not at random (MNAR) is difficult to rule out. It is argued, however, that MNAR analyses are, themselves, surrounded with problems and therefore, rather than ignoring MNAR analyses altogether or blindly shifting to them, their optimal place is within sensitivity analysis. The concepts developed here are illustrated using data from three clinical trials, where it is shown that the analysis method may have an impact on the conclusions of the study.
Keywords: Complete case analysis; Ignorability; Last observation carried forward; Missing at random; Missing completely at random; Missing not at random
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Van Assche, D. Van Caspel, J. Vanlauwe, J. Bellemans, D. B. Saris, F. P. Luyten, and F. Staes Physical Activity Levels After Characterized Chondrocyte Implantation Versus Microfracture in the Knee and the Relationship to Objective Functional Outcome With 2-year Follow-up Am. J. Sports Med., November 1, 2009; 37(1_suppl): 42S - 49S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Pausova, C. Syme, M. Abrahamowicz, Y. Xiao, G. T. Leonard, M. Perron, L. Richer, S. Veillette, G. D. Smith, O. Seda, et al. A Common Variant of the FTO Gene Is Associated With Not Only Increased Adiposity but Also Elevated Blood Pressure in French Canadians Circ Cardiovasc Genet, June 1, 2009; 2(3): 260 - 269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Rannou, J. Dimet, I. Boutron, G. Baron, F. Fayad, Y. Mace, J. Beaudreuil, P. Richette, P. Ravaud, M. Revel, et al. Splint for Base-of-Thumb Osteoarthritis: A Randomized Trial Ann Intern Med, May 19, 2009; 150(10): 661 - 669. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. N. Keene, B. Celli, J. A. Anderson, G. T. Ferguson, C. R. Jenkins, P. W. Jones, J. Vestbo, K. Knobil, J. C. Yates, and P. M. A. Calverley Lung function decline in COPD trials Eur. Respir. J., March 1, 2009; 33(3): 708 - 709. [Full Text] [PDF]
|
|
|
|
 |
|
 A.-W. Chan, A. Hrobjartsson, K. J Jorgensen, P. C Gotzsche, and D. G Altman Discrepancies in sample size calculations and data analyses reported in randomised trials: comparison of publications with protocols BMJ, December 4, 2008; 337(dec04_1): a2299 - a2299. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Thiebaut and S. Walker When it is better to estimate a slope with only one point QJM, October 1, 2008; 101(10): 821 - 824. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. P. McMahon, D. L. Kelly, D. L. Boggs, L. Li, Q. Hu, J. M. Davis, and W. T. Carpenter Jr. Feasibility of Reducing the Duration of Placebo-Controlled Trials in Schizophrenia Research Schizophr Bull, March 1, 2008; 34(2): 292 - 301. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Beunckens, G. Molenberghs, H. Thijs, and G. Verbeke Incomplete hierarchical data Statistical Methods in Medical Research, October 1, 2007; 16(5): 457 - 492. [PDF]
|
|
|
|
|
|
M. G Kenward and J. Carpenter Multiple imputation: current perspectives Statistical Methods in Medical Research, June 1, 2007; 16(3): 199 - 218. [Abstract] [PDF]
|
|
|
|
 |
|
 E. A. Lingard and D. L. Riddle Impact of Psychological Distress on Pain and Function Following Knee Arthroplasty J. Bone Joint Surg. Am., June 1, 2007; 89(6): 1161 - 1169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. R White, J. Carpenter, S. Evans, and S. Schroter Eliciting and using expert opinions about dropout bias in randomized controlled trials Clinical Trials, April 1, 2007; 4(2): 125 - 139. [Abstract] [PDF]
|
|
|
|
 |
|
 I. Jansen, A. Van den Troost, G. Molenberghs, A. A. Vermulst, and J. R. M. Gerris Modeling Partially Incomplete Marital Satisfaction Data Sociological Methods Research, August 1, 2006; 35(1): 113 - 136. [Abstract] [PDF]
|
|
|
|
 |
|
 C. K. Enders A Primer on the Use of Modern Missing-Data Methods in Psychosomatic Medicine Research Psychosom Med, May 1, 2006; 68(3): 427 - 436. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. G. Baker, G. M. Fitzmaurice, L. S. Freedman, and B. S. Kramer Simple adjustments for randomized trials with nonrandomly missing or censored outcomes arising from informative covariates Biostat., January 1, 2006; 7(1): 29 - 40. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Katz Methodological issues in clinical trials of opioids for chronic pain Neurology, December 29, 2005; 65(12_suppl_4): S32 - S49. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. L. Rollman, B. H. Belnap, S. Mazumdar, P. R. Houck, F. Zhu, W. Gardner, C. F. Reynolds III, H. C. Schulberg, and M. K. Shear A Randomized Trial to Improve the Quality of Treatment for Panic and Generalized Anxiety Disorders in Primary Care Arch Gen Psychiatry, December 1, 2005; 62(12): 1332 - 1341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. C. Johnson and G. Raghu Clinical trials in idiopathic pulmonary fibrosis: a word of caution concerning choice of outcome measures Eur. Respir. J., November 1, 2005; 26(5): 755 - 758. [Full Text] [PDF]
|
|
|
|
|
|
C. Beunckens, G. Molenberghs, and M. G Kenward Direct likelihood analysis versus simple forms of imputation for missing data in randomized clinical trials Clinical Trials, October 1, 2005; 2(5): 379 - 386. [Abstract] [PDF]
|
|
|
|
|
|
I. Sousa, A. G. Chetwynd, and P. J. Diggle Exploratory analysis of longitudinal trials with staggered intervention times Biostat., July 1, 2005; 6(3): 479 - 485. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. H Mallinckrodt, C. J Kaiser, J. G Watkin, G. Molenberghs, and R. J Carroll The effect of correlation structure on treatment contrasts estimated from incomplete clinical trial data with likelihood-based repeated measures compared with last observation carried forward ANOVA Clinical Trials, December 1, 2004; 1(6): 477 - 489. [Abstract] [PDF]
|
|