Comparison of techniques for handling missing covariate data within prognostic modelling studies: A simulation study

BMC Medical Research Methodology

Volumn 10, Issue , 2010, Pages

  Marshall, Andrea ^a,b   Altman, Douglas G ^a   Royston, Patrick ^c   Holder, Roger L ^d  

^a UNIVERSITY OF OXFORD   (United Kingdom)

^b UNIVERSITY OF WARWICK   (United Kingdom)

^c UNIVERSITY COLLEGE LONDON   (United Kingdom)

^d UNIVERSITY OF BIRMINGHAM   (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

ARTICLE; BREAST TUMOR; COMPARATIVE STUDY; COMPUTER SIMULATION; EPIDEMIOLOGY; FEMALE; HUMAN; METHODOLOGY; MULTIVARIATE ANALYSIS; PROPORTIONAL HAZARDS MODEL; STATISTICAL ANALYSIS;

BIAS (EPIDEMIOLOGY); BREAST NEOPLASMS; COMPUTER SIMULATION; DATA INTERPRETATION, STATISTICAL; FEMALE; HUMANS; MULTIVARIATE ANALYSIS; PROPORTIONAL HAZARDS MODELS; RESEARCH DESIGN;

EID: 77249126457     PISSN: None     EISSN: 14712288     Source Type: Journal    
DOI: 10.1186/1471-2288-10-7     Document Type: Article

References (44)

1. Missing covariate data within cancer prognostic studies: a review of current reporting and proposed guidelines. A Burton DG Altman, British Journal of Cancer 2004 91 1 4 8 10.1038/sj.bjc.6601907 15188004
2. Missing data in epidemiologic studies. W Vach M Blettner P Armitage T Colton, Encyclopedia of Biostatistics New York: John Wiley & Sons 1998 2641 2654
3. Bias due to missing exposure data using complete-case analysis in the proportional hazards regression model. S Demissie MP LaValley NJ Horton RJ Glynn LA Cupples, Statistics in Medicine 2003 22 4 545 557 10.1002/sim.1340 12590413
4. Using the EM-algorithm for survival data with incomplete categorical covariates. SR Lipsitz JG Ibrahim, Lifetime Data Analysis 1996 2 1 5 14 10.1007/BF00128467 9384645
5. Estimating equations with incomplete categorical covariates in the Cox model. SR Lipsitz JG Ibrahim, Biometrics 1998 54 3 1002 1013 10.2307/2533852 9750248
6. Maximum likelihood estimation for linear regression models with right censored outcomes and missing predictors. XL Meng N Schenker, Computational Statistics & Data Analysis 1999 29 4 471 483 10.1016/S0167-9473(98)00074-7
7. Multiple Imputation for Nonresponse in Surveys. DB Rubin, New York: John Wiley and Sons 2004
8. Statistical Analysis with Missing Data, Second edition. RJA Little DB Rubin, New York: John Wiley and Sons 2002
9. Multiple imputation of missing blood pressure covariates in survival analysis. S van Buuren HC Boshuizen DL Knook, Statistics in Medicine 1999 18 6 681 694 10.1002/(SICI)1097-0258(19990330)18:6<681::AID-SIM71>3.0.CO;2-R 10204197
10. Multiple-imputation inferences with uncongenial sources of input. XL Meng, Statistical Science 1994 9 4 538 558
11. Evaluation of Proc Impute and Schafer's imputation software. M Hu S Salvucci S Weng MP Cohen, Proceedings of the survey research methods section of the American Statistical Association. Chicago, Illinois 1996 287 292
12. Missing data: our view of the state of the art. JL Schafer JW Graham, Psychological Methods 2002 7 2 147 177 10.1037/1082-989X.7.2.147 12090408
13. The NHANES III multiple imputation project. J Schafer T Ezzati-Rice W Johnson M Khare R Little D Rubin, Proceedings of the Survey Research Methods Section of the American Statistical Association. Chicago, Illnois 1996 28 37
14. Partially parametric techniques for multiple imputation. N Schenker JMG Taylor, Computational Statistics & Data Analysis 1996 22 4 425 446 10.1016/0167-9473(95)00057-7
15. Multiple imputation versus data enhancement for dealing with missing data in observational health care outcome analyses. PD Faris WA Ghali R Brant CM Norris PD Galbraith ML Knudtson, Journal of Clinical Epidemiology 2002 55 2 184 191 10.1016/S0895-4356(01)00433-4 11809357
16. A critical look at methods for handling missing covariates in epidemiologic regression analyses. S Greenland WD Finkle, American Journal of Epidemiology 1995 142 12 1255 1264 7503045
17. Double-semiparametric method for missing covariates in Cox regression models. HY Chen, Journal of the American Statistical Association 2002 97 458 565 576 10.1198/016214502760047096
18. Non-ignorable missing covariate data in survival analysis: a case-study of an International Breast Cancer Study Group trial. AH Herring JG Ibrahim SR Lipsitz, Journal of the Royal Statistical Society Series C-Applied Statistics 2004 53 2 293 310 10.1046/j.1467-9876.2003.05168.x (Pubitemid 38597505)
19. Diagnostic research on routine care data prospects and problems. R Oostenbrink KGM Moons SE Bleeker HA Moll DE Grobbee, Journal of Clinical Epidemiology 2003 56 6 501 506 10.1016/S0895-4356(03)00080-5 12873643
20. Regression Modeling Strategies with Applications to Linear Models, Logistic Regression, and Survival Analysis. FE Harrell, New York: Springer-Verlag 2001
21. Imputations of missing values in practice: Results from imputations of serum cholesterol in 28 cohort studies. F Barzi M Woodward, American Journal of Epidemiology 2004 160 1 34 45 10.1093/aje/kwh175 15229115
22. Dealing with missing data. J Scheffer, Research Letters in the Information and Mathematical Sciences 2002 3 153 160
23. R: A language and environment for statistical computing. R Development Core Team, Vienna, Austria: R Foundation for Statistical Computing 2004
24. Modelling the effects of standard prognostic factors in node-positive breast cancer. German Breast Cancer Study Group (GBSG). W Sauerbrei P Royston H Bojar C Schmoor M Schumacher, British Journal of Cancer 1999 79 11-12 1752 1760 10.1038/sj.bjc.6690279 10206288
25. The design of simulation studies in medical statistics. A Burton DG Altman P Royston RL Holder, Statistics in Medicine 2006 25 24 4279 4292 10.1002/sim.2673 16947139
26. Generating survival times to simulate Cox proportional hazards models. R Bender T Augustin M Blettner, Statistics in Medicine 2005 24 11 1713 1723 10.1002/sim.2059 15724232
27. A prognostic model for ovarian cancer. TG Clark ME Stewart DG Altman H Gabra JF Smyth, British Journal of Cancer 2001 85 7 944 952 10.1054/bjoc.2001. 2030 11592763
28. A comparison of inclusive and restrictive strategies in modern missing data procedures. LM Collins JL Schafer CM Kam, Psychological Methods 2001 6 4 330 351 11778676
29. A new measure of prognostic separation in survival data. P Royston W Sauerbrei, Statistics in Medicine 2004 23 5 723 748 10.1002/sim.1621 14981672
30. Adjusting regression attenuation in the Cox proportional hazards model. FH Kong, Journal of Statistical Planning and Inference 1999 79 1 31 44 10.1016/S0378-3758(98)00178-5
31. Analysis of Incomplete Multivariate Data. JL Schafer, New York: Chapman and Hall 1997
32. Combining estimates of interest in prognostic modelling studies after multiple imputation: current practice and guidelines. A Marshall D Altman R Holder P Royston, BMC Medical Research Methodology 2009 9 1 57 10.1186/1471-2288-9-57 19638200
33. Significance levels from repeated p-values with multiply-imputed data. KH Li XL Meng TE Raghunathan DB Rubin, Statistica Sinica 1991 1 1 65 92
34. How should variable selection be performed with multiply imputed data? AM Wood IR White P Royston, Statistics in Medicine 2008 27 17 3227 3246 10.1002/sim.3177 18203127
35. Multiple imputation in health-care databases: an overview and some applications. DB Rubin N Schenker, Statistics in Medicine 1991 10 4 585 598 10.1002/sim.4780100410 2057657
36. A comparison of imputation methods in a longitudinal randomized clinical trial. LQ Tang JW Song TR Belin J Unutzer, Statistics in Medicine 2005 24 14 2111 2128 10.1002/sim.2099 15889392
37. Multiple imputation after 18+ years. DB Rubin, Journal of the American Statistical Association 1996 91 434 473 489 10.2307/2291635
38. Modelling and imputation of semicontinuous survey variables. JL Schafer MK Olsen, The Methodology Center, Penn State University, USA 2000
39. A potential for bias when rounding in multiple imputation. NJ Horton SR Lipsitz M Parzen, American Statistician 2003 57 4 229 232 10.1198/0003130032314
40. Imputing missing covariate values for the Cox model. I White P Royston, Statistics in Medicine 2009 28 15 1982 1998 10.1002/sim.3618 19452569
41. norm: Analysis of multivariate normal datasets with missing values. JL Schafer AA Novo, R package version 1.0.9 2002
42. mix: Estimation/multiple Imputation for Mixed Categorical and Continuous Data. JL Schafer, R package version 1.0.4 2003
43. mice: Multivariate Imputation by Chained Equations library. S van Buuren CGM Oudshoorn, R package version 1.13.1 2005
44. Hmisc: Harrell Miscellaneous library for R statistical software. FE Harrell, R package 2.2-3 2004