This study examines the potential coverage bias in telephone surveys. Data were analyzed from the first phase of the third National Health and Nutrition and Examination Survey conducted from 1988 to 1991. In that survey, 10,120 persons 17 years and older were interviewed and 9034 were examined. About 2.7% of respondents reported not having a telephone. Differences in demographic and lifestyle variables, but not physiological or anthropometric variables, existed between persons with a telephone and those without one. Respondents without a telephone were more Likely to report that an impairment or health problems limited their work or activities. Compared with respondents with a telephone, those without one were more likely to be current smokers, to be less physically active, to never have had their brood pressure checked or have had it checked more than 5 years ago, and to never have had their cholesterol checked. Based on data from a 24-hour dietary recall, persons without a telephone consumed less vitamin A, Vitamin C, vitamin E, and carotene than did respondents with a telephone. However, prevalence estimates of health characteristics obtained from telephone surveys in populations with high telephone coverage are unlikely to be seriously affected by coverage bias nor are conclusions of comparisons involving populations with low telephone coverage. Published by Elsevier Science Inc.
