A controlled trial of envelope colour for increasing response rates in older women

Abstract

Background and aims: Postal questionnaires are widely used in health research to provide measurable outcomes in areas such as quality of life. Participants who fail to return postal questionnaires can introduce non-response bias. Previous studies within populations over the age of 65 years have shown that response rates amongst older people can be 60% or less. The current study sought to investigate whether envelope colour affected response rates in a study about the effectiveness of screening older women for osteoporosis. Methods: A total of 2803 eligible female participants aged between 70 and 85 were sent an invitation pack from their GP practice. The invitation was either in a brown or white envelope and contained a matching pre-paid reply envelope. A study questionnaire was also sent out in brown or white envelopes 1 week after consenting to participate in the trial. Results: The overall response rate was 78%. There was little evidence of an effect of envelope colour on response to the invitation to participate in the trial (OR 1.04, 95% CI 0.87–1.24). Similarly, there was no influence of envelope colour on the number of participants returning their questionnaires (OR 0.99, 95% CI 0.60–1.63). There was weak evidence of an effect of envelope colour on the response rates of the consent process (OR 0.86, 95% CI 0.74–1.00). When we updated a recent meta-analysis with the results of this study, there was a non-statistically- significant trend for greater response rates with brown envelopes compared with white envelopes (OR 1.19, 95% CI 0.86–1.64, I²=92%). However, the results where influenced by one study and when this study was excluded the pooled estimate was 0.98 (95% CI 0.89–1.08, I²=0%). Conclusion: This study found no evidence to suggest envelope colour has an effect on response to participate in a trial or questionnaire returns. There is weak evidence to suggest envelope colour may affect consent into a trial.