Methodological aspects of a computer-assisted telephone interview survey of oral health

Despite the reported benefits of computer-assisted telephone interview (CATI) methods, experiences from their use in Australian oral health surveys have not been described. This report aimed to present methodological aspects of a CATI survey conducted in the five mainland states. A response rate of 66 per cent was obtained, yielding 4050 completed interviews. Analysis revealed generally small levels of non-response bias: persons who avoided or delayed dental treatment because of cost and non-health card holders were harder to contact, while non-English speakers and persons aged 20-29 years were less likely to participate. A total of 1770 person hours of interview time was spent on the survey: 64.5 per cent of that time was spent on the telephone with an average of 10 minutes 17 seconds per call (13 minutes 37 seconds per completed call). Only seven questions had missing data for more than 1 per cent of respondents. Comprehension of questions and cooperation with the interview was rated by interviewers as 'good' or 'very good' for more than 90 per cent of respondents. The CATI method was highly efficient and yielded good quality data for the survey.