Quantifying Patient Preferences and Expectations About Diabetic Retinopathy Monitoring.

Abstract

Importance: Diabetic retinopathy (DR) is the leading cause of blindness among adults in the US. The US Centers for Disease Control and Prevention recommends annual DR monitoring for all individuals with diabetes, yet monitoring rates remain below 70%.

Objective: To evaluate how patient preferences and expectations about DR monitoring are associated with expected monitoring adherence behaviors.

Design, setting, and participants: In this survey study, a web-enabled survey instrument was developed and implemented with a discrete-choice experiment to characterize patient preferences for outcomes of DR monitoring and graded-pair questions to quantify patients' expectations about the impact of DR monitoring on blindness risk. The survey was conducted through ResearchMatch, a US National Institutes of Health-developed online platform, among adults with self-reported, physician-diagnosed diabetes. Recruitment occurred between September 15, 2023, and October, 17, 2023, and data analysis occurred between October 2023 and December 2023. Results from the 2 tasks were combined to derive patients' expected monitoring behavior following a recently proposed treatment adherence framework. The survey instrument was pretested in cognitive interviews and validated for the purposes of this study.

Exposure: Survey-based discrete-choice experiment and graded-pair questions.

Main outcomes and measures: Participants' relative preferences for DR-related blindness risk reductions, monitoring time, and out-of-pocket monitoring costs were quantified, as well as the degree to which participants expected adherence to monitoring to affect the risk of blindness. By combining how much participants valued specific reductions in blindness risk (relative to monitoring costs) and their expected risk reduction through monitoring, the rate at which patients would maximize the benefit of monitoring appointments was assessed.

Results: The survey was completed satisfactorily by 304 respondents of 542 individuals invited to participate. Mean (SD) respondent age was 40.5 (11.2) years, and 169 respondents (56.1%) were female. Reductions in blindness risk were valuable to participants. Participants required a 3.87 (95% CI, 1.91-5.88) percentage-point reduction in 5-year blindness risk to be fully adherent to an annual 53-minute monitoring visit with a $26 co-payment, but respondents expected DR monitoring to reduce the 5-year blindness risk by 0.71 (95% CI, 0.21-1.28) percentage points.

Conclusions and relevance: In this online survey study among adults with diabetes, measurement of patient preferences and expectations about DR monitoring with properly validated instruments offered an opportunity to assess patient health behaviors. The association between preferences and monitoring expectations was generally consistent with monitoring nonadherence among adults with diabetes and offers insights that may help address inconsistent DR monitoring.