Priorities in biomedical research. Indices of burden.

The relative burden of different diseases is a legitimate component in the consideration of priorities for biomedical research. Simple prevalence statistics do not directly reflect the different burdens on services imposed by diseases of comparable prevalence. This paper sets out for each 54 categories of disease five indices of the burden on services, based respectively on inpatient days, outpatient referrals, consultations in family practice, sickness benefit, and loss of expectation of life. There is considerable variation in the rank-order of categories of disease, in their contribution to the five burdens; but for each burden the number of categories accounting for 50% of the total burden is not large, ranging from 3 to 9 out of the possible 54.