2019 HSR&D/QUERI National Conference
4085 — Development of a Novel Measure of Timely Access to Primary Care Services in VHA
Lead/Presenter: Peter Kaboli, COIN - Iowa City
All Authors: Kaboli PJ (Iowa City VA), Augustine, MR (Bronx VA), Batten, A (VA Puget Sound) Nelson, KM (VA Puget Sound)
Timeliness, or the ability to provide care when requested, is a critical component of access. Current measures of timely access to primary care (PC) may not account for patient urgency or modern scheduling practices. Our objective was to develop a novel measure for identifying clinic performance in fulfilling next-day and walk-in requests after adjusting for patient demographics and risk.
We performed a retrospective, longitudinal analysis identifying clinic-level success in fulfilling timely care requests within 48 hours from 6/2013 to 6/2017. Using VA administrative data from 174 PC clinics in 8 VISNs, we defined a timely care request as any walk-in or first next available with successful fulfillment defined as a: completed encounter within 48 hours in the same clinic with any PC provider; different clinic within the same facility; non-VA visit paid by VA; hospital admission; secure message; or telehealth visit. Failure to provide timely care was determined if the patient was not seen within 48 hours, appointment cancelled by the clinic, or appointment fulfilled in the emergency department for conditions considered PC treatable. Bayesian hierarchical regression modeled annual timely care success rate within each clinic, controlling for age, sex, race, rurality, and patient risk; two- and three-way interactions between rurality, patient risk, and time were also tested.
From 2013 to 2017, 512,113 patients made 1,271,007 timely care requests with the number increasing 19% from 267,276 to 318,634. Of all timely care requests, 85-90% were successfully fulfilled: 98% by VA with 62% in PC, 32% in other VA clinics, and 3% in VA mental health clinics. Non-VA care and secure messaging fulfilled 2% and 1%, respectively. Of 174 clinics, mean overall success rate was 87-88% with an upper bound of 95-97%. The Bayesian credible interval of the mean success rate widened over time with the lower bound decreasing from 72% to 64%. Over four years of observation, 19 clinics (11%) were identified as outliers at providing timely care; two and three-way interactions were not significant.
We developed a novel metric of timely care, which included patient perceived urgency, modern scheduling practices, and multiple care delivery modalities. Robust models identified a high-level of timely care success across VA and differentiated struggling clinics.
While further validation is needed, this measure offers a novel procedure to assess access while providing real-time, actionable intelligence to clinical leadership.