This Article Full Text (PDF) All Versions of this Article: 0272989X08330121v1 29/4/532    most recent References Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Scopus Google Scholar Articles by Chan, D. C. Articles by Weinstein, M. C. PubMed PubMed Citation Articles by Chan, D. C. Articles by Weinstein, M. C. Social Bookmarking                         What's this?

Quantitative Risk Stratification in Markov Chains with Limiting Conditional Distributions

Department of Economics, Massachusetts Institute of Technology, Cambridge, Massachusetts, dcchan{at}partners.org, Department of Medicine, Brigham and Women's Hospital, Boston, Massachusetts

Philip K. Pollett, PhD

Department of Mathematics, University of Queensland, Australia

Milton C. Weinstein, PhD

Department of Health Policy and Management, Harvard School of Public Health, Boston, Massachusetts

Background . Many clinical decisions require patient risk stratification. The authors introduce the concept of limiting conditional distributions, which describe the equilibrium proportion of surviving patients occupying each disease state in a Markov chain with death. Such distributions can quantitatively describe risk stratification. Methods. The authors first establish conditions for the existence of a positive limiting conditional distribution in a general Markov chain and describe a framework for risk stratification using the limiting conditional distribution. They then apply their framework to a clinical example of a treatment indicated for high-risk patients, first to infer the risk of patients selected for treatment in clinical trials and then to predict the outcomes of expanding treatment to other populations of risk. Results. For the general chain, a positive limiting conditional distribution exists only if patients in the earliest state have the lowest combined risk of progression or death. The authors show that in their general framework, outcomes and population risk are interchangeable. For the clinical example, they estimate that previous clinical trials have selected the upper quintile of patient risk for this treatment, but they also show that expanded treatment would weakly dominate this degree of targeted treatment, and universal treatment may be cost-effective. Conclusions . Limiting conditional distributions exist in most Markov models of progressive diseases and are well suited to represent risk stratification quantitatively. This framework can characterize patient risk in clinical trials and predict outcomes for other populations of risk.

Key Words: risk stratification • Markov models • mathematical models • decision analysis.

This version was published on July 1, 2009

Medical Decision Making, Vol. 29, No. 4, 532-540 (2009)
DOI: 10.1177/0272989X08330121


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?