Is there a difference in health estimates between people with listed and unlisted telephone numbers?

OBJECTIVE: To examine the characteristics of people with a silent telephone by demographic and health-related issues and to assess the potential bias in telephone surveys that use the White Pages as the sampling frame.
METHOD: A representative annual population health survey of South Australians aged 15 years and over conducted in 1994, 1998, 1999 and 2002. Self-reported information on telephone connection and listings in the White Pages was provided by participants. Questions were included in the 1998 survey on reasons why telephone number was not listed.
RESULTS: The rate of unlisted telephone numbers significantly increased (17.3% to 20.2%) in South Australia between 1994 and 2002. People with an unlisted number are more likely to be living in the metropolitan area, single adult households, current smokers, and in the younger age groups. In the multivariate analyses the only health and risk factor variable that was consistently significant over the years was smoking status. The main reason for having an unlisted number was to avoid market research or telemarketing calls (33%).
CONCLUSION: Telephone surveys using White Pages samples are suitable to collect information for prevalence on most commonly collected self-reported health conditions and health-related issues. The exception is smoking status. IMPLICATIONS: The use of Whites Pages telephone samples is a suitable method for obtaining population health information because it is efficient when there is limited time and funds.