Analyzing Longitudinal Health-Related Quality of Life Data: Missing Data and Imputation Methods
Health-related quality of life (HRQL) outcomes are frequently incorporated into clinical trials of new medical treatments. Problems associated with missing data, multiplicity of outcomes, and longitudinal data structure complicate the statistical analysis of HRQL data. Various simple and complex imputation techniques and statistical methods have been evaluated to deal with missing at random or missing not at random HRQL data. I used simulations to examine how well four relatively simple imputation methods reproduce known population statistics. The findings suggest that all the imputation methods provide acceptable estimates of the missing HRQL data when less than 10% of the data is missing. The empirical Bayes method was best at reproducing population characteristics, even when missing data rates exceeded 25%. The remaining imputation approaches began to introduce bias into the imputed estimates when missing data rates exceeded 20% across treatment groups. Future research should compare empirical Bayes or multiple imputation with statistical analysis approaches to handling missing HRQL outcome data.
KeywordsImputation Method Last Observation Carry Forward Root Mean Square Residual HRQL Score HRQL Outcome
Unable to display preview. Download preview PDF.
- 5.Diehr, P., Patrick, D.L., Hedrick, S., Rothman, M., Grembowski, D., Raghunathan, T.E. and Beresford, S. (1995). Including deaths when measuring health status over time. Medical Care 33 (suppl), AS 164–172.Google Scholar
- 8.Fairclough, D. Multiple imputation for non-random missing data in longitudinal studies of health-related quality of life. International Workshop on Statistical design, measurements and Analysis of Health Related Quality of Life, University of South Brittany, Arradon, France, October 16–17, 2000.Google Scholar
- 9.Fayers, P.M. and Machin, D. (2000). Quality of Life: Assessment, Analysis and Interpretation. Chichester: John Wiley & Sons.Google Scholar
- 10.Molenberghs, G. Sensitivity analysis of longitudinal quality of life data. International Workshop on Statistical design, measurements and Analysis of Health Related Quality of Life, University of South Brittany, Arradon, France, October 16–17, 2000.Google Scholar
- 11.Fairclough, D.L. (1998). Methods of analysis of longitudinal studies of health-related quality of life. In: Staquet, M.J., Hays, R.D. and Fayers, P.M. (eds), Quality of Life Assessment in Clinical Trials: Methods and Practice. New York: Oxford University Press.Google Scholar
- 13.Besarab, A., Bilton, W.K., Browne, J.K., Egrie, J.C., Nissenson, A.R., Okamoto, D.M., Schwab, S.J. and Goodkin, D.A. (1998). The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoetin. New England Journal of Medicine 339, 584–590.PubMedCrossRefGoogle Scholar
- 14.Ware, J.E., Snow, K.K., Kosinski, M. and Gandek, B. (1993) SF-36 Health Survey: Manual and Interpretation Guide. Boston: The Health Institute, New England Medical Center.Google Scholar
- 16.Bollen, K.A. (1989) Structural Equations with Latent Variables. New York: John Wiley & Sons.Google Scholar
- 17.SAS Institute (1996). SAS/STAT User’s Guide, Release 6.03. Cary, NC: SAS Institute.Google Scholar
- 18.Zar, J.H. (1996). Biostatistical Analysis. 3rd Edition. Upper Saddle River, NJ: Prentice Hall.Google Scholar