[The validity of repeated mail surveys concerning smoking habits for Japanese physicians]

A cross-sectional study of smoking prevalence among medical doctors was performed using questionnaires. Mailing four times proved essential in order to obtain a reasonable number of responses from the subjects. The objective of the present study was to analyze the smoking characteristics of the subjects who returned the questionnaires after the second to fourth mailings.

A set of an anonymous questionnaire and a letter from the President of the Japan Medical Association (JMA) requesting cooperation was mailed with a return envelope with the subject's name and address written to 3,000 male and 1,500 female doctors randomly selected from the roster membership of the JMA. The survey was conducted between February and June 2000.

The collection rate after the first mailing was 66%, while the subtotal collection rate for the second through fourth mailings was 21%, bringing the grand-total to 87%. The total prevalence of smoking among the subjects who had sent back the questionnaire on the second to fourth mailings was approximately 1.5 times higher than for those who had sent back the questionnaire after the initial mailing for both male and female subjects. As to other characteristics of the subjects who only responded after the second to fourth mailings were: working in hospitals (odds ratios; male: 1.39, female: 1.47), not giving smoking cessation guidance (odds ratio; male: 0.58), and not recognizing the idea that doctors should not smoke (odds ratio; female: 0.67).

The results suggested that for future surveys on smoking to be conducted in academic institutions or work-places, efforts to obtain responses from those who do not answer the first attempt should take into consideration these points.