Random digit dialing for control selection. A review and a caution on its use in studies of childhood cancer.

Random digit dialing is an effective method for creating a probability sample of households for telephone interviews. Epidemiologists have successfully used the technique to select controls for population-based case-control studies. However, a modified version of random digit dialing used in some institution-based studies of childhood cancer etiology involves a selection process which is biased against children from families with closely spaced births. The author discusses alternative approaches to reducing this bias in control group selection.