Selection of elderly controls using random digit dialing.

The experience of identifying and soliciting elderly male controls, using a nonclustered random digit dialing procedure in a case-control study, is presented. For elderly controls (ages 65-84) 3.5 times more residential telephone numbers were required than for controls encompassing a much broader age range (ages 40-84). This is a function of the proportion of elderly in the population and their lower response to telephone identification. Elderly controls, age 70 and older, also had lower participation rates.