When participants fail to respond to health surveys this may bias the results. Findings suggest that non-response is increasingly due to difficulties contacting participants, rather than other reasons such as death or participants refusing to take part, but little is known about how this type of non-response might influence results.

The study compared people who responded and those who could not be contacted on health factors (e.g. smoking) and the relationship between health factors (e.g. between smoking and self-rated health). It found that those who could not be contacted were more likely to bcomputer-313840_1280e less educated, stressed about money, separated, divorced or widowed, to have experienced violence, and to be a smoker. For relationships between health factors, the effect of monetary stress on self-rated health was higher for those who responded, but all other relationships were similar. These findings suggest that despite the loss of participants the relationships between health factors were found to be accurate.

Citation: Powers J, Tavener M, Graves A, Loxton D, ‘Loss to follow-up was used to estimate bias in a longitudinal study: A new approach’, Journal of Clinical Epidemiology, (2015)

Contact: Jennifer Powers jenny.powers@newcastle.edu.au

Leave a Reply

Your email address will not be published. Required fields are marked *