Response bias associated with asking former study participants to assist in new research

Avoidance and inhibition do not predict nonrespondent bias among patients with inflammatory bowel disease

Background

It has been suggested that participant withdrawal from studies can bias estimates. However, this is only possible when withdrawers and nonwithdrawers differ in an important way. We tested the hypothesis that withdrawers are more likely than nonwithdrawers to be avoidant and negatively affected.

Methods

A total of 1160 participants with inflammatory bowel disease were recruited at different sites in Switzerland. Their levels of avoidance coping and negative affectivity were rated by means of 2 short baseline questionnaires. One year later, they were sent a longer follow-up questionnaire. The primary outcome was return versus non-return of the follow-up questionnaire within 3 months. After controlling for potential confounders identified in an extensive literature search, we estimated the odds of returning the follow-up questionnaire for 1 standard deviation of avoidance coping and negative affectivity.

Results

The odds ratio for 1 standard deviation was 1.03 (95% confidence interval: 0.89–1.18) for avoidance coping and 1.02 (0.89–1.17) for negative affectivity.

Conclusions

The odds of returning the questionnaires did not depend on avoidance coping or negative affectivity.