The COVID-19 pandemic has caused a wide range of ambulatory and elective outpatient services to be deferred or delayed. This will create massive future backlogs during the recovery phase. This project examined strategies used in VA to systematically track delayed care, and evaluated the potential impact of countermeasures to improve access in the COVID-19 recovery phase, using gastrointestinal endoscopy as an exemplar.
Aim 1: To identify a standardized, flexible, and low-burden process to identify and track deferred services.
Aim 2a: To develop a framework for making triage/prioritization decisions for healthcare needs in the system recovery phase that will be applicable to a wide array of ambulatory and elective medical services.
Aim 2b: Use simulation modeling to evaluate the impact of hypothetical countermeasures (e.g., selective rescheduling strategies) on future resource utilization and access.
Aim 3: To evaluate Veterans' perceptions of the acceptability of the alternative prioritization strategies identified in Aim 2.
In collaboration with VISN 10 leadership, we surveyed Radiology and Gastroenterology Service/Section Chiefs to examine how deferred services were tracked during the COVID-19 pandemic, and how patients were being prioritized during the ramp up. As part of the survey, we requested that any guidance documents they were using be shared with us. We then employed a rapid evidence synthesis process to identify published clinical prioritization frameworks.
Next, we used discrete event simulation to better understand whether the implementation of several evidence-based practice (EBP) strategies (i.e., "Exchange" - exchanging colorectal cancer screening by colonoscopy with screening by FIT, a stool-based home test; "Extend" - extending the surveillance interval for patients with low-risk polyps from 5 years to 7 years, in accordance with a recent change in guidelines) could reduce demand for routine screening and surveillance colonoscopy and increase access for necessary diagnostic and surveillance colonoscopy services. The model, with a length of about 3 years, examined a hypothetical endoscopy unit that had a pre-COVID capacity of 110 procedures weekly; the unit was part of an integrated healthcare system without external referrals to community care. We made the following assumptions: 1) during the early pandemic, capacity was reduced to 5 percent of pre-COVID levels for 10 weeks; 2) capacity incrementally increased back to 100 percent by 30 weeks; 3) the number of referrals each week was 113; and 4) patients with the highest priority were always seen first, as recommended by VA GI National Program guidance, with all others joining a waiting line (i.e., queue). Using published VA data, referrals included (in descending order of priority) 23% for diagnostic colonoscopy, 28% for upper endoscopy, 10% for high-risk polyp surveillance, 17% for low-risk polyp surveillance, and 22% for screening colonoscopy. (The Extend strategy would be appropriate for low-risk polyp surveillance and the Exchange strategy would be appropriate for screening colonoscopy.)
Finally, we met with local Veterans Research Engagement Committee (VREC) members to help us better understand: 1) how Veterans feel about the proposed EBP strategies to help decrease wait times, and 2) how we can best communicate these types of recommendations to patients.
We received survey responses from 12 (60%) of Service/Section Chiefs in VISN 10. We found little evidence of novel approaches to tracking (i.e., most sites were using the consult toolbox or a simple excel sheet). Additionally, leads were adopting the guidance distributed by VACO for prioritization of services. Early in our rapid evidence synthesis, we discovered a pre-print of a systematic review that examined tools that support patient prioritization and outlined the prioritization criteria most commonly included. Thus, a complete synthesis by our team was not necessary, and we instead created a structured set of questions that could assist clinicians in systematically evaluating whether a service could be appropriately postponed, changed to an alternative that may be safer during a pandemic, or foregone in the best interest of the patient. We identified common clinical scenarios that illustrated each of the situations [For example: Question - Is there another testing or treatment strategy with equivalent evidence of effectiveness, but which is safer or uses less health care resources (e.g., FIT testing vs colonoscopy, ophthalmologic exam vs retinal photography)? If yes, choose the alternative strategy.].
Our modeling exercise illustrated that with usual care, the average wait time across all procedural indications was nearly 6 months, driven predominantly by wait times for screening colonoscopy, which exceeded one year; more than 3,000 patients remained in the queue at the end of our study period. The Exchange strategy reduced average wait times by 56% and the Extend strategy by 37%. Notably, with Exchange the wait time for screening colonoscopy was reduced to zero since all CRC screening was done using FIT, and with Extend the wait times for screening still exceeded 8 months. Seven VREC members met with us to discuss the alternative strategies. Members indicated that they would be open to the EBP alternatives if they were shared verbally by their PCP or someone else with whom they have a well-established relationship, so they could ask follow-up questions as needed, and included information on the effectiveness of the strategy.
The COVID-19 pandemic has created an opportunity and urgency to maximize the value of care provided by VA, and in the process increase Veteran access. Our work shows how simulation modeling can help VA clinical leaders evaluate and select evidence-based strategies to maximize access to critical services when it is limited by external or internal events. We are now partnering with a national program office to examine how the use of alternative evidence based practice strategies can improve Veteran access to needed diagnostic and surveillance endoscopy services.
None at this time.
None at this time.
None at this time.
None at this time.