Random digit dialing (ROD) has been a valuable tool for recruiting controls for population-based case-control studies in the United States for more than twenty years. Traditional ROD methods screen only households with landline telephones for study eligibility. The overall proportion of households with cellular but no landline telephone service has increased rapidly from 2% in 2003 to 10% in 2005. There is evidence that this proportion will continue to increase in the future. Since individuals with cellular-only telephone service are younger, less affluent and less likely to be married than the general population, the sampling frame for traditional landline ROD may be particularly unrepresentative of these groups. However, there are only limited available data regarding differences in lifestyle characteristics between individuals with cell phone- only service versus those with landline service which are relevant to many epidemiologic studies. It is unclear if it is feasible to include cellular telephones in ROD sampling frames and if characteristics of individuals with only cell phones are different enough from those with landline telephones to warrant the additional effort and expense. The proposed study will develop, field test and evaluate a dual-frame ROD sample that will include both landline and cellular telephones. Five hundred fifty randomly selected men and women 20-44 years of age with cellular telephones, but no landline telephone service, will be identified and interviewed by telephone. Selected characteristics of these men and women will be compared to 800 respondents of the same age and gender with landline telephones recruited using traditional ROD methods for two case-control studies conducted during similar time periods. The results of this study will have important implications for the future use of ROD for epidemiologic studies. Lav description: A common way of studying the causes of cancer is to compare the characteristics (such as diet, medical history, and family history) of people diagnosed with cancer to a control group of people without cancer. One of the most useful methods of identifying a good control group is a process called random digit dialing (ROD), which involves recruiting controls by dialing randomly generated phone numbers. Traditional ROD has excluded people who only have a cell phone, and so this study will evaluate the extent to which excluding these people could bias studies that investigate the causes of cancer, since the characteristics of people who only have a cell phone may be different from those who have a landline phone. ? ? ? ? ? ? ? ? ? ? ?
Voigt, Lynda F; Schwartz, Stephen M; Doody, David R et al. (2011) Feasibility of including cellular telephone numbers in random digit dialing for epidemiologic case-control studies. Am J Epidemiol 173:118-26 |