Markov Chains with Measurement Error: Estimating the ‘True’ Course of a Marker of the Progression of Human Immunodeficiency Virus Disease

SUMMARY A Markov chain is a useful way of describing cohort data. Longitudinal observations of a marker of the progression of the human immunodeficiency virus (HIV), such as CD4 cell count, measured on members of a cohort study, can be analysed as a continuous time Markov chain by categorizing the CD4 cell counts into stages. Unfortunately, CD4 cell counts are subject to substantial measurement error and short timescale variability. Thus, fitting a Markov chain to raw CD4 cell count measurements does not determine the transition probabilities for the true or underlying CD4 cell counts; the measurement error results in a process that is too rough. Assuming independent measurement errors, we propose a likelihood-based method for estimating the 'true'or underlying transition probabilities. The Markov structure allows efficient calculation of the likelihood by using hidden Markov model methodology. As an example, we consider CD4 cell count data from 430 HIV-infected participants in the San Francisco Men's Health Study by categofizing the marker data into seven stages; up to 17 observations are available for each individual.We find that including measurement error both produces a significantly better fit and provides a model for CD4 progression that is more biologically reasonable.

[1]  P. Bacchetti.  Estimating the Incubation Period of AIDS by Comparing Population Infection and Diagnosis Patterns , 1990 .

[2]  John G. Kemeny,et al.  Finite Markov chains , 1960 .

[3]  B. Ripley,et al.  Statistical models for prevalent cohort data. , 1995, Biometrics.

[4]  Bradley P. Carlin,et al.  Hierarchical Bayes Models for the Progression of HIV Infection Using Longitudinal CD4+ Counts. , 1992 .

[5]  Lisa Sattenspiel,et al.  Modeling and analyzing HIV transmission: the effect of contact patterns , 1988 .

[6]  Jeremy M. G. Taylor,et al.  A Stochastic Model for Analysis of Longitudinal AIDS Data , 1994 .

[7]  I M Longini,et al.  A simulation model of AIDS in San Francisco: I. Model formulation and parameter estimation. , 1991, Mathematical biosciences.

[8]  Jens Perch Nielsen,et al.  A framework for consistent prediction rules based on markers , 1993 .

[9]  I. Longini,et al.  Estimation of incidence of HIV infection using cross-sectional marker surveys. , 1994, Biometrics.

[10]  N P Jewell,et al.  Estimating patterns of CD4 lymphocyte decline using data from a prevalent cohort of HIV infected individuals. , 1994, Statistics in medicine.

[11]  L. Baum,et al.  A Maximization Technique Occurring in the Statistical Analysis of Probabilistic Functions of Markov Chains , 1970 .

[12]  J. Phair,et al.  Effect of CD4+ Cell Count Measurement Variability on Staging HIV‐1 Infection , 1992, Journal of acquired immune deficiency syndromes.

[13]  Rolf Langeheine,et al.  Latent variables Markov models. , 1994 .

[14]  P. Solomon.  Models for Infectious Human Diseases: AIDS: modelling and predicting , 1996 .

[15]  R. Holder,et al.  Long-Term Trends in Biochemical Data Obtained from Two Population Surveys , 1986, Annals of clinical biochemistry.

[16]  T R Holford,et al.  Change points in the series of T4 counts prior to AIDS. , 1995, Biometrics.

[17]  M. Wulfsohn,et al.  Modeling the Relationship of Survival to Longitudinal Data Measured with Error. Applications to Survival and CD4 Counts in Patients with AIDS , 1995 .

[18]  Halina Frydman,et al.  A Nonparametric Estimation Procedure for a Periodically Observed Three‐State Markov Process, with Application to Aids , 1992 .

[19]  P Pezzotti,et al.  Estimation of the acquired immunodeficiency syndrome incubation period in intravenous drug users: a comparison with male homosexuals. , 1992, American journal of epidemiology.

[20]  R. Capocaccia,et al.  An age and period reconstruction of the HIV epidemic in Italy. , 1994, International journal of epidemiology.

[21]  Lawrence R. Rabiner,et al.  A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.

[22]  A. Gelfand,et al.  Hierarchical Bayes Models for the Progression of HIV Infection Using Longitudinal CD4 T-Cell Numbers , 1992 .

[23]  Paul F. Lazarsfeld,et al.  Latent Structure Analysis. , 1969 .

[24]  F. H. Garrison,et al.  HISTORY OF MEDICINE , 1914 .

[25]  Carlo Berzuini,et al.  Bayesian graphical models of the natural history of HIV infection , 1995 .

[26]  William H. Press,et al.  Numerical recipes in C. The art of scientific computing , 1987 .

[27]  D R Fredkin,et al.  Bayesian restoration of single-channel patch clamp recordings. , 1992, Biometrics.

[28]  J. Rice,et al.  Maximum likelihood estimation and identification directly from single-channel recordings , 1992, Proceedings of the Royal Society of London. Series B: Biological Sciences.

[29]  N D Le,et al.  Exact likelihood evaluation in a Markov mixture model for time series of seizure counts. , 1992, Biometrics.

[30]  A. Dawid,et al.  On Testing the Validity of Sequential Probability Forecasts , 1993 .

[31]  R Brookmeyer,et al.  Reconstruction and future trends of the AIDS epidemic in the United States. , 1991, Science.

[32]  U. Dafni,et al.  Modeling the Progression of HIV Infection , 1991 .

[33]  M. Newton,et al.  Stochastic Modeling of Early Hematopoiesis , 1995 .

[34]  L. Collins,et al.  Latent Class Models for Stage-Sequential Dynamic Latent Variables , 1992 .

[35]  A. Verdecchia,et al.  A back-calculation method to estimate the age and period HIV infection intensity, considering the susceptible population. , 1995, Statistics in medicine.

[36]  I. Longini,et al.  Estimating the stage-specific numbers of HIV infection using a Markov model and back-calculation. , 1992, Statistics in medicine.

[37]  D. Thomas,et al.  Simultaneously modelling censored survival data and repeatedly measured covariates: a Gibbs sampling approach. , 1996, Statistics in medicine.

[38]  Yudi Pawitan,et al.  Modeling Disease Marker Processes in AIDS , 1993 .

[39]  Ira M. Longini,et al.  The stages of HIV infection: waiting times and infection transmission probabilities , 1989 .

[40]  J. Levy,et al.  Sexual practices and risk of infection by the human immunodeficiency virus. The San Francisco Men's Health Study. , 1987, JAMA.

[41]  Cristiana Larizza,et al.  A Unified Approach for Modeling Longitudinal and Failure Time Data, With Application in Medical Monitoring , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[42]  Victor DeGruttola,et al.  Modeling The Relationship Between Progression Of CD4-Lymphocyte Count And Survival Time , 1992 .

[43]  A. Krämer.  Markers as time-dependent covariates in relative risk regression. MAP Workshop (Multi-cohort Analysis Project). , 1993, Statistics in medicine.

[44]  I M Longini,et al.  The dynamics of CD4+ T-lymphocyte decline in HIV-infected individuals: a Markov modeling approach. , 1991, Journal of acquired immune deficiency syndromes.