Response to comment on O’Hara et al. Understanding Patient Preferences Regarding Limb Salvage for Diabetic Foot Ulcers: A Discrete Choice Experiment. Diabetes Care 2025;48:1517–1523

We thank Dr. Reed for her feedback (1) on our recent publication in Diabetes Care (2). Our design, like any study, was guided by a priori assumptions and heuristic decisions. We consider these choices to be justifiable within the context of this study and unlikely to qualitatively alter the interpretation of the results.

We acknowledge that the calculation used for our sample size was introduced decades ago. However, old is not necessarily antiquated. The calculation endures as a reasonable estimate without requiring prior data to inform distributional assumptions, unlike most recent approaches. This study was the pilot work to inform larger studies in this population. In addition, we recognize that the subgroup analyses were underpowered, which was explicitly stated in the first paragraph of the Research Design and Methods section.

We used three bins of surgical risk for several reasons. Simplifying probabilities into ordered categories reduces cognitive burden. Pairing verbal labels, such as “low,” “medium,” and “high,” with numeric anchors improves comprehension and reliability relative to numeric probabilities alone (3,4). Furthermore, individuals do not perceive probabilities linearly and tend to overweight moderate risks (5). Categorizing risk into qualitative ranges is supported by decision psychology (5). We acknowledge that the phrase “marginal rates of substitution” implies an infinitesimal change, and “trade-off” would have been more appropriate in this context. Finally, our use of the term “linear” in this context refers to sequential deltas of similar proportion.

We agree that the relative importance estimates should always be interpreted within the context of attributes and levels included in the design. In our study, we selected the attributes and ranges based on patient interviews and expert consensus to reflect the decisional dilemma. While relative importance is inherently conditional on the design, our objective was not to claim universal weightings of preferences but rather to characterize the trade-offs patients face in this clinical scenario.

While these design decisions can be debated, our paper provides a proof of concept that veterans with diabetes can separate the attributes of postrecovery mobility and amputation level in a discrete choice experiment. The findings are foundational for developing a decision aid for people with a diabetic foot ulcer who are candidates for surgery.

Funding Statement

This work was supported in part by VA Merit award CX-0016-01 from the U.S. Department of Veterans Affairs Clinical Sciences Research and Development Service and National Institutes of Health University of Maryland Pepper Center (award P30 AG028747).