A Model Relating Quality of Life to Latent Health Status and Survival

  • Mei-Ling Ting Lee
  • G. Alex Whitmore


QOL assessments measure a subject’s state of health and well-being in a global sense or with reference to particular domains of function, symptoms and living. In this chapter, we define a QOL process for a subject as a continuous-time stochastic process that is periodically assessed by means of a survey instrument. The QOL process is assumed to have three components: a survival component that is correlated with the survival time of the subject, a palliative component that reflects a subject’s comfort, freedom from pain and other aspects of well-being that are not correlated with survival, and a noise component that represents a combination of measurement errors and extraneous effects. We present a statistical model that can be used to distinguish between the survival component and the combined palliative and noise components. Our model also provides a way of incorporating markers of health status in the analysis and evaluating their importance. The model assumes that health status and related health markers follow a joint stochastic process. The markers are assumed to be observable whereas the health status process is assumed to be latent or unobservable. The primary endpoint, which we take to be death, is assumed to be triggered when this latent process first crosses a failure threshold level. Inferences for the model are based on censored survival data and marker measurements. Covariates, such as treatment variables, risk factors and other baseline variables, are related to the model parameters through generalized linear regression functions. We interpret the model as a joint process for QOL, latent health status and, possibly, markers of health status. The portion of the QOL process that is correlated with the latent health status process forms the survival component of QOL. The proposed model provides a rich conceptual framework for the study of QOL issues and offers a flexible and tractable methodology for associated statistical inferences. The model and methods are illustrated by a small case demonstration.


Survival Component Failure Threshold Censor Survival Data Stationary Independent Increment Initial Health Status 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Cox, D. R. (1999). Some remarks on failure-times, surrogate markers, degradation, wear and the quality of life. Lifetime Data Analysis 5, 307–314.PubMedCrossRefGoogle Scholar
  2. Glasziou, P.P., Cole, B.F., Gelber, R.D., Hilden, J., Simes, R.J. (1998). Quality adjusted survival analysis with repeated quality of life measurements. Statistics in Medicine 17, 1215–1229.PubMedCrossRefGoogle Scholar
  3. Lee, M.-L.T., V. DeGruttola and D. Schoenfeld (2000). A model for markers and latent health status. J. Royal. Statist Soc., Series B 62, 747–762.CrossRefGoogle Scholar
  4. Lee, M.-L.T. and G. A. Whitmore (2000a). Assumptions of a Latent Survival Model. In: Proceeding of the Conference on Goodness of Fit, 1999, Paris, France, forthcoming.Google Scholar
  5. Lee, M.-L.T. and G. A. Whitmore (2000b). Distribution-free inference methods for latent disease progression and survival. Technical paper, Channing Laboratory, Harvard Medical School.Google Scholar
  6. Whitmore, G. A., Crowder, M.J. and Lawless, J.F. (1998). Failure inference from a marker process based on a bivariate Wiener model. Lifetime Data Analysis 4, 229–251.PubMedCrossRefGoogle Scholar
  7. Zhao, H. and A. A. Tsiatis (1997). A consistent estimator for the distribution of quality-adjusted survival time. Biometrika 84, 339–348.CrossRefGoogle Scholar
  8. Zhao, H. and A. A. Tsiatis (1999). Efficient estimation of the distribution of quality-adjusted survival time. Biometrics 55, 231–236.CrossRefGoogle Scholar
  9. Zhao, H. and A. A. Tsiatis (2000). Estimating mean quality-adjusted lifetime with censored data. Sankyha, Series B 62, 175–188.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2002

Authors and Affiliations

  • Mei-Ling Ting Lee
    • 1
    • 2
  • G. Alex Whitmore
    • 3
  1. 1.Channing LaboratoryBrigham & Women’s HospitalUSA
  2. 2.Harvard UniversityUSA
  3. 3.McGill UniversityCanada

Personalised recommendations