Applying Last-birthday respondent selection method to an oral health telephone survey in an Iranian population

Document Type : Original Article


1 Dental student, School of Dentistry, Mashhad University of Medical Sciences, Mashhad, Iran.

2 PhD in Biostatistics, Research Fellow and Biostatistician, University of Sydney, Sydney, NSW, Australia.

3 Community Oral Health Dept., School of Dentistry, Mashhad University of Medical Sciences


Background: A representative sample in health surveys ensures the findings can be reliably generalized to the target population. Conducting oral health surveys through telephone interviews has become more common, and ensuring respondent randomness is necessary for any health survey. Several techniques have been suggested. This paper reports applying the last-birthday method as a within-household random selection method for the first time in an oral health telephone survey in Iran.
Methods: This study was part of a larger research in which adult citizens' self-perceived oral health was compared with an objective dental examination. The last-birthday method randomly selected a household member for each attempted landline number, asking the primary respondent to select an eligible family member with the most recent birthday. The selected respondent was then either contacted or replaced with another respondent from the same household based on the research criteria.
Results: Of the 6745 called numbers, 1771 were invalid, 3129 did not respond, 364 were not households, and 771 declined to be interviewed. Finally, 710 respondents entered the random selection procedure, of which 53 had no eligible family member to select. The sample selection method's difficulty caused 36 refusals. Of the 621 selected final respondents, 30 could not be contacted or refused upon introduction. The total percentage of "selection" and "post-selection” dropouts that could be attributed to the sample selection method was 7.41%. Based on the American Association for Public Opinion Research (AAPOR Response Rate 1) guidelines, the minimum response rate for the interview was 13%, and the AAPOR Response Rate 3 was 39.6%. In all characteristics except for employment status (P = 0.488), the final participant's demographic characteristics were significantly different from those of the city population (P < 0.0001).
Conclusion: Oral health science can make use of the last-birthday selection strategy. This technique seems to obtain a reasonably representative sample through a respondent-friendly selection process in telephone surveys.


Main Subjects