Factors impacting questionnaire response in a Dutch retrospective cohort study

To study questionnaire length, type of consent, approach to recruitment, and subject characteristics on participation in epidemiologic studies.

As part of a health survey among Dutch subjects treated for ear, nose, and throat disorders in childhood, we conducted a pilot study of 200 individuals who were randomly assigned to one of four categories, defined by length of questionnaire (long vs. short) and type of consent form (basic vs. multi-option). In addition, among 8402 subjects eligible to be in the main study (average age 41 years in 1997), we examined the effect of approach to recruitment and subject characteristics on participation rates.

The pilot study showed a non-significant 10% increase in participation rate using the shorter questionnaire, but no differences by type of consent form. In the full survey, the participation rate was 49% after the first mailing. Response increased by 15% after a written reminder and by 10% after a telephone survey. The total participation rate was 74%. Attained age, sex, exposure status, age at exposure, and response to an earlier survey were determinants of participation rates. Among male non-participants, outright refusal was less frequent than non-response. The refusal rate, unlike the non-response rate, was positively associated with older age at time of survey.

Health survey participation is influenced by questionnaire length, frequency of contact, and subject characteristics.