Modes of Survey and Reported Biases: Study of Latino Health

Studies have shown mixed and biased results from modes of interview that seek personal, sensitive or stigmatizing information. This study compares the results of face-to-face (298) and telephone (290) interviews of Latinos with arthritis on levels of pain, disability, coping skills, and willingness to participate in self-management programs. Latino adults were screened using the CDC's arthritis questions. A 50-question survey was administered, and participants received $20 gift certificates.

There were no statistical differences between the two groups on most demographic variables, insurance coverage, degree of pain and use of over-the-counter drugs. Reporting on degree of pain does not appear to be a function of mode of interview. Our findings confirm previous studies in which Latinos reported high levels of pain. Those interviewed face-to-face differed from telephone groups in their greater interest in specific self-management classes. Higher interest among face-to-face interviewees may be due to lower health status, earlier onset of arthritis, more activity limitations, more co-morbidities, greater likelihood of reporting early retirement, and job changes.

Significant differences between the two groups on several key variables do not appear to stem from interview modes; rather, from factors such as type of work, number of years lived in this country, and beliefs and attitudes about arthritis. Although the two groups are similar on a variety of measures, sole reliance on telephone interviews may lead to over-estimation of prescription use and clinic visits and lower estimation of co-morbidities and activity limitations. Ethnicity appears to have no influence in modes of interview.