To the Editor:—We appreciated the recent article by Subramanian and colleagues exploring the impact of open access (OA) scheduling on diabetes care,1 but were left with two critical unanswered questions.

How do the authors conceptualize “open access”? The authors do not describe what changes intervention group clinics undertook to implement open access. Their statement that “The index date for the OA clinics was the date of OA implementation” suggests OA was implemented on a single day. We therefore speculate they use the term “open access” to describe only a scheduling strategy that emphasizes same-day appointments and does not allow future appointments to be scheduled. We understand OA to connote a much broader set of principles. As conceptualized by Mark Murray and others,2,3 OA is designed to eliminate unwanted delays for appointments and delays in the office through intensive, clinic-wide redesign of care delivery processes (e.g., balancing the supply of and demand for appointments, reducing the pre-existing appointment backlog, addressing bottlenecks, and several other redesign strategies). Further, the authors state that, with OA, patients “…are required to remember and schedule their appointments at a suggested interval,” which is not a property of OA systems as we understand them. As Murray has stated, “…when a patient needs a pre-scheduled or follow-up appointment, that appointment is given…Pre-scheduling patients with certain clinical needs ensures you will not lose them to follow-up.”4 The purpose of OA is to improve the accessibility of health care, not limit it by banning future appointments.

Did intervention practices successfully implement open access, as described in the literature? Though the authors hypothesized that “OA implementation would improve access to care,” they do not report any clinic-level data (e.g., time to the third available preventive care appointment) to suggest whether this hypothesis was supported and acknowledge they did not have a measure of the degree of OA implementation. This is particularly relevant given the study’s non-randomized design. At baseline, intervention practices were those selected for “less than optimal provider productivity and high missed appointment rates.” Whether or not OA was fully and successfully implemented and overcame these pre-intervention differences in clinic operations is essential information for interpreting the study results.

Waiting is ubiquitous in American medicine, is costly, and may be harmful to patient health. We agree with the authors wholeheartedly that OA is worthy of further study. Unfortunately, it appears this study did not actually evaluate OA. Instead, it seems the intervention was simply a scheduling strategy that limited access by not allowing patients to schedule appointments in the future.