Mixed contact methods to improve response to a postal questionnaire

Postal questionnaires remain an important method of collecting data in trials. However, a high non-response rate can lead to biases, which may undermine the validity of the study.

To assess a simple method of trying to improve response rates in an occupational health trial evaluating an intervention to prevent hand dermatitis in nurses.

The trial employed questionnaires at t = 0, t = 1 month and t = 12 months. The t = 1 month questionnaire was posted to study participants (student and intensive care nurses) together with a free postage reply envelope. After 2 weeks, an e-mail was sent to non-responders reinforcing the need for completed questionnaires to be returned. Two weeks later, non-responders were sent another hard copy of the questionnaire, along with an accompanying letter. Six weeks after posting the initial questionnaires, non-responders were sent an SMS text message or were telephoned to remind them to return the questionnaire.

The response rates for the 744 student nurses were 8% (no reminder), 27% (after first reminder), 22% (after second reminder) and 27% (after the third reminder), resulting in a response rate of 63%. The response rates for the 959 intensive care nurses were 9% (no reminder), 24% (after first reminder), 24% (after second reminder) and 31% (after third reminder), resulting in a final response rate of 63%.

We found that a series of regular reminders including a third, personalized reminder by SMS text or telephone had a positive impact on non-responders.