How to Measure Health: A Stochastic Model for an Index of Health

Abstract

An index of health based on a continuous time Markov process is proposed for measuring the health of a defined population over a calendar year. The health of an individual is regarded as a dynamic phenomenon, varying on a continuum from optimum well-being to extreme illness. This continuum is divided into an ordered set of states: States of health. Transition probabilities from one health state to another over a time interval are expressed in terms of intensity functions. A basic concept underlying this index is the expected duration of stay in each health state during the year for each individual. The proposed index, which is equivalent to the fraction of the year a person is healthy, gives a numerical measurement of the health of an entire population with values ranging from zero to one. © 1973 Oxford University Press.