Preferences of people with diabetes for diabetic retinopathy screening: a discrete choice experiment

The ever-increasing prevalence of diabetes places pressure on the provision of diabetic retinopathy screening services. As the first study of its kind, we aimed to determine preferences for diabetic retinopathy screening in people with diabetes and to examine the trade-offs between frequency of screening and other service attributes.

A questionnaire including a discrete choice experiment was administered to people (n = 198) attending diabetic retinopathy screening at eight clinics across Wales, United Kingdom. The discrete choice experiment contained eight pairwise choices in which screening provision was described by five attributes: frequency of screening; travel time; results time; ability of screening to detect other changes; and explanation of results. Data were analysed using logistic regression techniques.

We gained a response rate of 86.4% from the 198 questionnaires administered at clinics; 160 complete responses were analysed. Respondents valued four out of the five attributes [ability of screening to detect other changes (P = 0.000), explanation of results (P = 0.024), frequency of screening (P = 0.000) and travel time (P = 0.007)]. Results time was insignificant (P = 0.122). Respondents were willing to wait an additional 12, 2 and 1 month between screening tests to have a test that was able to detect additional changes, to have their results explained in person rather than by letter and to have a 15-min reduction in travel time, respectively.

Respondents were willing to accept a longer screening interval, as long as preferences for other attributes of service provision (ability of screening to detect other changes, explanation of results and travel time) were made available.