Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls

被引:5203
作者
Sterne, Jonathan A. C. [1 ]
White, Ian R. [2 ]
Carlin, John B. [3 ,4 ]
Spratt, Michael [1 ]
Royston, Patrick [5 ,6 ]
Kenward, Michael G. [7 ]
Wood, Angela M. [8 ]
Carpenter, James R. [7 ]
机构
[1] Univ Bristol, Dept Social Med, Bristol BS8 2PR, Avon, England
[2] MRC, Biostat Unit, Inst Publ Hlth, Cambridge CB2 0SR, England
[3] Murdoch Childrens Res Inst, Clin Epidemiol & Biostat Unit, Parkville, Vic 3052, Australia
[4] Univ Melbourne, Parkville, Vic 3052, Australia
[5] MRC, Clin Trials Unit, Canc Grp, London NW1 2DA, England
[6] MRC, Clin Trials Unit, Stat Methodol Grp, London NW1 2DA, England
[7] Univ London London Sch Hyg & Trop Med, Med Stat Unit, London WC1E 7HT, England
[8] Inst Publ Hlth, Dept Publ Hlth & Primary Care, Cambridge, England
来源
BMJ-BRITISH MEDICAL JOURNAL | 2009年 / 339卷
关键词
VALUES; MODELS;
D O I
10.1136/bmj.b2393
中图分类号
R5 [内科学];
学科分类号
100201 [内科学];
摘要
We are enthusiastic about the potential for multiple imputation and other methods 14 to improve the validity of medical research results and to reduce the waste of resources caused by missing data. The cost of multiple imputation analyses is small compared with the cost of collecting the data. It would be a pity if the avoidable pitfalls of multiple imputation slowed progress towards the wider use of these methods. It is no longer excusable for missing values and the reason they arose to be swept under the carpet, nor for potentially misleading and inefficient analyses of complete cases to be considered adequate. We hope that the pitfalls and guidelines discussed here will contribute to the appropriate use and reporting of methods to deal with missing data.
引用
收藏
页码:157 / 160
页数:9
相关论文
共 31 条
[1]
Multiple imputation for missing data - A cautionary tale [J].
Allison, PD .
SOCIOLOGICAL METHODS & RESEARCH, 2000, 28 (03) :301-309
[2]
[Anonymous], 1997, Analysis of Incomplete Multivariate Data, DOI [DOI 10.1201/9780367803025, DOI 10.1201/9781439821862]
[3]
Robustness of a multivariate normal approximation for imputation of incomplete binary data [J].
Bernaards, Coen A. ;
Belin, Thomas R. ;
Schafer, Joseph L. .
STATISTICS IN MEDICINE, 2007, 26 (06) :1368-1382
[4]
Cost-effectiveness in clinical trials: using multiple imputation to deal with incomplete cost data [J].
Burton, Andrea ;
Billingham, Lucinda Jane ;
Bryan, Stirling .
CLINICAL TRIALS, 2007, 4 (02) :154-161
[5]
CARPENTER J, COMMENTS COMPUTATION
[6]
Carpenter J.R., 2008, MISSING DATA RANDOMI
[7]
Sensitivity analysis after multiple imputation under missing at random: a weighting approach [J].
Carpenter, James R. ;
Kenward, Michael G. ;
White, Ian R. .
STATISTICAL METHODS IN MEDICAL RESEARCH, 2007, 16 (03) :259-275
[8]
A comparison of multiple imputation and doubly robust estimation for analyses with missing data [J].
Carpenter, James R. ;
Kenward, Michael G. ;
Vansteelandt, Stijn .
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES A-STATISTICS IN SOCIETY, 2006, 169 :571-584
[9]
A comparison of inclusive and restrictive strategies in modern missing data procedures [J].
Collins, LM ;
Schafer, JL ;
Kam, CM .
PSYCHOLOGICAL METHODS, 2001, 6 (04) :330-351
[10]
On the performance of random-coefficient pattern-mixture models for non-ignorable drop-out [J].
Demirtas, H ;
Schafer, JL .
STATISTICS IN MEDICINE, 2003, 22 (16) :2553-2575