Incidence and prevalence of impairments

Estimates of YLD are based on whatever descriptive epidemiological data are available for each disease. The key measures of interest are the incidence of the disease; the prevalence, duration and remission rate of the impairments it causes; and excess mortality from the disease. Ideally, this information should be available on an age-specific basis and estimates would exist for each region of interest.

Evidently, these parameters are interrelated. Incident cases either recover or die and the time that it takes them to do so determines the prevalence of the condition. Moreover, between them, incidence by age and the survival times of cases that die determine the number of deaths by age from the disease. Few sources of data, however measure all aspects of this jigsaw and different data sources tend to use different case definitions, age intervals, and so on.

IHME uses a software tool known as DisMod to model each disease and its sequelae based on whatever data are available on its natural history. The aim is to produce estimates of the epidemiological parameters of interest that are both internally consistent and consistent with the estimates made of deaths from the disease.

In addition, if independent estimates exist of the prevalence of an impairment that results from multiple diseases, the disease-specific estimates of the impairment are constrained to sum to its overall prevalence. Examples of such impairments include vision and hearing loss, anaemia, infertility and intellectual disability.