The impact of missing data on analyses of a time-dependent exposure in a longitudinal cohort: A simulation study

Emerging Themes in Epidemiology

Volumn 10, Issue 1, 2013, Pages

  Karahalios, Amalia ^a,b   Baglietto, Laura ^a,b   Lee, Katherine J ^c,d   English, Dallas R ^a,b   Carlin, John B ^a,c   Simpson, Julie A ^a,b  

^a UNIVERSITY OF MELBOURNE   (Australia)

^b CANCER EPIDEMIOLOGY CENTRE   (Australia)

^c MURDOCH CHILDREN'S RESEARCH INSTITUTE   (Australia)

^d UNIVERSITY OF MELBOURNE   (Australia)

Author keywords

Complete case analysis; Missing exposure; Multiple imputation; Repeated exposure measurement; Simulation study

Indexed keywords

ACCURACY; ADULT; AGED; ARTICLE; CANCER EPIDEMIOLOGY; CANCER MORTALITY; CANCER RISK; CASE STUDY; COHORT ANALYSIS; COLORECTAL CANCER; DATA ANALYSIS; FEMALE; HUMAN; LONGITUDINAL STUDY; MAJOR CLINICAL STUDY; MALE; MONTE CARLO METHOD; OUTCOME ASSESSMENT; REFERENCE DATABASE; SIMULATION; TIME; WAIST CIRCUMFERENCE;

EID: 84882954625     PISSN: None     EISSN: 17427622     Source Type: Journal    
DOI: 10.1186/1742-7622-10-6     Document Type: Article

References (49)

1. A review of reporting missing data in cohort studies with repeated assessment of exposure measures. Karahalios A, Baglietto L, English D, Simpson J, BMC Med Res Methodol 2012 12 96 10.1186/1471-2288-12-96 22784200
2. Missing data: a systematic review of how they are reported and handled. Eekhout I, de Boer RM, Twisk JWR, de Vet HCW, Heymans MW, Epidemiology 2012 23 5 729 732 10.1097/EDE.0b013e3182576cdb 22584299
3. Comparison of techniques for handling missing covariate data within prognostic modelling studies: a simulation study. Marshall A, Altman DG, Royston P, Holder RL, BMC Med Res Methodol 2010 10 7 10.1186/1471-2288-10-7 20085642
4. Bias and efficiency of multiple imputation compared with complete-case analysis for missing covariate values. White IR, Carlin JB, Stat Med 2010 29 28 2920 31 10.1002/sim.3944 20842622
5. Imputation of missing values is superior to complete case analysis and the missing-indicator method in multivariable diagnostic research: a clinical example. van der Heijden GJMG, Donders ART, Stijnen T, Moons KGM, J Clin Epidemiol 2006 59 10 1102 1109 10.1016/j.jclinepi.2006.01.015 16980151
6. Biased estimation of the odds ratio in case-control studies due to the use of ad hoc methods of correcting for missing values for confounding variables. Vach W, Blettner M, Am J Epidemiol 1991 134 8 895 907 1670320
7. SAS Insitute Inc, SAS OnlineDoc, Version 8 Cary, NC: SAS Institute, Inc. 2000
8. StataCorp, Stata statistical software: Release 11 College Station, TX: StataCorp LP 2009
9. Little RJA, Rubin DB, Statistical analysis with missing data (2nd edition) New York: J Wiley & Sons 2002
10. Bias due to missing exposure data using complete-case analysis in the proportional hazards regression model. Demissie S, LaValley MP, Horton NJ, Glynn RJ, Cupples LA, Stat Med 2003 22 4 545 557 10.1002/sim.1340 12590413
11. Unpredictable bias when using the missing indicator method or complete case analysis for missing confounder values: an empirical example. Knol MJ, Janssen KJM, Donders ART, Egberts ACG, Heerdink ER, Grobbee DE, Moons KGM, Geerlings MI, J Clin Epidemiol 2010 63 7 728 736 10.1016/j.jclinepi.2009.08.028 20346625
12. Using the outcome for imputation of missing predictor values was preferred. Moons KGM, Donders RART, Stijnen T, Harrell FEJr, J Clin Epidemiol 2006 59 10 1092 1101 10.1016/j.jclinepi.2006.01.009 16980150
13. Missing data methods for dealing with missing items in quality of life questionnaires. A comparison by simulation of personal mean score, full information maximum likelihood, multiple imputation, and hot deck techniques applied to the SF-36 in the French 2003 decennial health survey. Peyre H, Leplège A, Coste J, Qual Life Res 2011 20 2 287 300 10.1007/s11136-010- 9740-3 20882358
14. Impact of missing data due to drop-outs on estimators for rates of change in longitudinal studies: a simulation study. Touloumi G, Babiker AG, Pocock SJ, Darbyshire JH, Stat Med 2001 20 24 3715 3728 10.1002/sim.1114 11782028
15. Missing covariate data in medical research: to impute is better than to ignore. Janssen KJM, Donders ART, Harrell FE Jr, Vergouwe Y, Chen Q, Grobbee DE, Moons KGM, J Clin Epidemiol 2010 63 7 721 727 10.1016/j.jclinepi.2009.12.008 20338724
16. A comparison of imputation techniques for handling missing predictor values in a risk model with a binary outcome. Ambler G, Omar RZ, Royston P, Stat Methods Med Res 2007 16 3 277 298 10.1177/0962280206074466 17621472
17. Joint modeling of missing data due to non-participation and death in longitudinal aging studies. Rajan KB, Leurgans SE, Stat Med 2010 29 21 2260 2268 10.1002/sim.4010 20645281
18. Weighted estimating equations for longitudinal studies with death and non-monotone missing time-dependent covariates and outcomes. Shardell M, Miller RR, Stat Med 2008 27 7 1008 1025 10.1002/sim.2964 17579923
19. The Melbourne Collaborative Cohort Study. Giles GG, English DR, IARC Sci Publ 2002 156 69 70 12484128
20. Influence of obesity on the risk of developing colon cancer. Frezza EE, Wachtel MS, Chiriva-Internati M, Gut 2006 55 2 285 291 10.1136/gut.2005.073163 16239255
21. Body size and composition and colon cancer risk in men. MacInnis R, English D, Hopper J, Haydon A, Gertig D, Giles G, Cancer Epidemiol Biomarkers Prev 2004 13 4 553 15066919
22. Body size and composition and colon cancer risk in women. MacInnis R, English D, Hopper J, Gertig D, Haydon A, Giles G, Int J Cancer 2006 118 6 1496 1500 10.1002/ijc.21508 16187280
23. Body size and composition and risk of rectal cancer (Australia). MacInnis R, English D, Haydon A, Hopper J, Gertig D, Giles G, Cancer Causes Control 2006 17 10 1291 1297 10.1007/s10552-006-0074-y 17111261
24. Weight change and cancer risk in a cohort of more than 65,000 adults in Austria. Rapp K, Klenk J, Ulmer H, Concin H, Diem G, Oberaigner W, Schroeder J, Ann Oncol 2008 19 4 641 648 18056917
25. Prospective weight change and colon cancer risk in male US health professionals. Thygesen LC, Grønbaek M, Johansen C, Fuchs CS, Willett WC, Giovannucci E, Int J Cancer 2008 123 5 1160 1165 10.1002/ijc.23612 18546286
26. Lohman T, Roche A, Martorell R (Eds), Anthropometric standardization reference manual Champaign IL: Kinetics Books 1988
27. The design of simulation studies in medical statistics. Burton A, Altman DG, Royston P, Holder RL, Stat Med 2006 25 24 4279 4292 10.1002/sim.2673 16947139
28. Simulation of correlated continuous and categorical variables using a single multivariate distribution. Tannenbaum S, Holford N, Lee H, Peck C, Mould D, J Pharmacokinet Pharmacodyn 2006 33 6 773 794 10.1007/s10928-006-9033-1 17053984
29. Generating survival times to simulate Cox proportional hazards models. Bender R, Augustin T, Blettner M, Stat Med 2005 24 11 1713 1723 10.1002/sim.2059 15724232
30. Modeling the drop-out mechanism in repeated-measures studies. Little RJ, J Am Stat Assoc 1995 90 431 1112 1121 10.1080/01621459.1995.10476615
31. Multiple imputation for multivariate missing-data problems: a data analyst's perspective. Schafer J, Olsen M, Multivariate Behav Res 1998 33 4 545 571 10.1207/s15327906mbr3304-5
32. Imputing missing covariate values for the Cox model. White I, Royston P, Stat Med 2009 28 15 1982 1998 10.1002/sim.3618 19452569
33. Rubin D, Multiple imputation for nonresponse in surveys New York: J Wiley & Sons 1987
34. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. Sterne J, White I, Carlin J, Spratt M, Royston P, Kenward M, Wood A, Carpenter J, BMJ 2009 338 2393 10.1136/bmj.b2393 19564179
35. Why missing data matter in the longitudinal study of adolescent development: using the 4-H Study to understand the uses of different missing data methods. Jelicić H, Phelps E, Lerner RM, J Youth Adolesc 2010 39 7 816 835 10.1007/s10964-010-9542-5 20446025
36. Reweighting estimators for Cox regression with missing covariate data: analysis of insulin resistance and risk of stroke in the Northern Manhattan Study. Xu Q, Paik MC, Rundek T, Elkind MSV, Sacco RL, Stat Med 2011 30 28 3328 3340 10.1002/sim.4380 21965165
37. Body size, weight change, and risk of colon cancer. Bassett JK, Severi G, English DR, Baglietto L, Krishnan K, Hopper JL, Giles GG, Cancer Epidemiol Biomarkers Prev 2010 19 11 2978 2986 10.1158/1055-9965.EPI-10-0543 20870733
38. A prospective study of body mass index, weight change, and risk of cancer in the proximal and distal colon. Laake I, Thune I, Selmer R, Tretli S, Slattery ML, Veierød MB, Cancer Epidemiol Biomarkers Prev 2010 19 6 1511 1522 10.1158/1055-9965.EPI-09-0813 20501754
39. Review: a gentle introduction to imputation of missing values. Donders ART, van der Heijden GJMG, Stijnen T, Moons KGM, J Clin Epidemiol 2006 59 10 1087 1091 10.1016/j.jclinepi.2006.01.014 16980149
40. Multiple imputation for missing data: fully conditional specification versus multivariate normal imputation. Lee KJ, Carlin JB, Am J Epidemiol 2010 171 5 624 632 10.1093/aje/kwp425 20106935
41. On the stationary distribution of iterative imputations. Liu J, Gelman A, Hill J, Su YS, 2010 arXiv preprint arXiv:1012.2902
42. Using modern missing data methods with auxiliary variables to mitigate the effects of attrition on statistical power. Graham J, Missing data: analysis and design New York: Springer 2012 253 275
43. Recovery of information from multiple imputation: a simulation study. Lee KJ, Carlin JB, Emerg Themes Epidemiol 2012 9 3 10.1186/1742-7622-9-3 22695083
44. R Development Core Team, R: A language and environment for statistical computing. Software Vienna, Austria: R Foundation for Statistical Computing 2004
45. IBM SPSS statistics for windows. IBM Corp, 2012 Version 21.0, Armonk, NY
46. The use and reporting of multiple imputation in medical research-a review. Mackinnon A, J Intern Med 2010 268 6 586 593 10.1111/j.1365-2796.2010. 02274.x 20831627
47. Missing data: our view of the state of the art. Schafer JL, Graham JW, Psychol Methods 2002 7 2 147 177 12090408
48. Assumptions. Schafer J, Analysis of incomplete multivariate data New York: Chapman and Hall 1997
49. A Bayesian proportional hazards regression model with non-ignorably missing time-varying covariates. Bradshaw PT, Ibrahim JG, Gammon MD, Stat Med 2010 29 29 3017 3029 10.1002/sim.4076 20960582