SELECTION OF NEIGHBORHOOD CONTROLS FOR A STUDY OF FACTORS RELATED TO THE DIAGNOSIS OF CERVICAL-CANCER

The level of effort required to generate neighborhood controls for a statewide matched case-control study of cervical cancer was investigated, with the aim of identifying hard-to-reach demographic subgroups. Cross reference telephone directories were used to identify households on the same street as the case. Letters were then sent to the households, followed by 'phone calls. A total of 2920 households were contacted to obtain 147 controls. Overall, 63.6% of age-eligible contacts participated in the study. In 49.3% of all households the major reason for not obtaining a control was ''no age-eligible woman''. Level of effort required to obtain a matched control was greater for black women than for white women-on average 24 letters and 40 'phone calls for black women us 12 letters and 20 calls for white women. Fewer eligible younger women refused to be interviewed than older. No marked differences were noted when the data were stratified by urban-rural area of residence.