I was intrigued by the editorial [1] in the 1 June issue of Annals. It seems that the benefit of screening does not accrue to the "average person" in a population but to the persons who are found to have a particular disease. In the example given in the editorial, in a disease with a prevalence of 1:1000, screening is said to "provide a maximum benefit ... of only 2 to 18 days." If only one person with the disease were found, that person might receive a benefit of 49.3 years of life.

Little doubt exists that any one of us would find 49 years of life a considerable benefit. The question is whether the other 999 persons would wish to pay to have that benefit accrue to the single individual. Health care is provided to individuals. Clearly, methods are needed to focus on appropriate individuals to screen for less prevalent conditions. Makers of public policy, however, must never lose sight of the fact that the doctor-patient relationship exists between two individuals. Screening for a low-prevalence disease needs to be the result of a negotiation in which the patient is informed of the low likelihood and the potentially great nature of the benefit.