Background

The use of the Internet as a research tool has dramatically increased in the past several years. Yet, the current literature favors the response rate achieved from paper-based studies. Knowledge of successful methods to increase participation in Internet-based research is scarce. The objective of this study is to examine the effects of different reminders to encourage study participation on overall response patterns in an Internet-based intervention study.

Methods

In 2008, 3,876 employees at four companies in the railway sector in Sweden were randomly e-mailed an Internet-based lifestyle questionnaire consisting of: A) questions, B) questions + interactive personalized automated feedback, or C) questions + interactive personalized automated feedback + telephone counseling. The questionnaire assessed health aspects including diet, physical activity, sleep, stress, alcohol/tobacco consumption, and motivation to change health. Interactive feedback was provided for all health sections; telephone counseling was offered for diet, physical activity, alcohol and smoking habits. Nine months later, a follow-up questionnaire (C) was e-mailed to examine health improvements. 4-5 and 11 e-mail reminders were sent at baseline and follow-up, respectively. Additional reminders (flyers, texts in internal media/bulletin board, information talks, SMS, visit by the research group etc) were also administered at the four companies, to various extents. The number of additional reminders was summarized and analyses were based on the total number of received additional reminders (low, moderate or high). Response patterns were examined in relation to basic characteristics, company, work type (office/field worker), received e-mail reminders, and total number of received additional reminders. As a result of the study, the companies received recommendations for future health implementations.

Results

38% and 36% completed the baseline and follow-up questionnaire, respectively. The majority of the participants was male, non-smokers, employed as field workers, and had a BMI ≥25. The 4-5 e-mail reminders increased the total response rate by 15%; the 11 e-mail reminders by 21%. Additional reminders had a marginal effect on total response rate, yet generated a positive effective on the response rate among office workers (71%). Since the planning process of the study, the company involved had the highest overall response rate (61%: P<0.001), despite receiving a moderate number of additional reminders. The employees at this company were almost 1.80 (CI: 1.55-2.08) times more likely to participate in the baseline questionnaire, compared to the company which entered the study just prior to the start and had the lowest overall response rate. Participant characteristics including sex, age, BMI, smoking, motivation to change health habits, and version of the completed questionnaire (A, B or C) were not associated with time of response. The highest participation at follow-up, however, was found for those who completed baseline questionnaire A, consisting solely of questions.

Conclusions

A well-established collaboration with the participants prior to study start and to send out e-mail reminders on a continuous basis are two effective strategies to increase the response rate in Internet-based studies. Additional reminders conducted in the work setting may only be effective among office workers participating in Internet-based studies.