Ultra-rapid synthesis of secondary sources for hospital COVID-19 guidance

Article type
Authors
Mitchell M1, Flores E1, Connolly J2, Goel R2, Gottschalk A2, Kavanagh N2, Philipson B2, Wang S2, Wood C2, Mull N1
1University of Pennsylvania Health System
2University of Pennsylvania School of Medicine
Abstract
Background: The COVID-19 pandemic has brought about an urgent need for rapid evidence-based guidance to inform local decision making on infection control, procedures for safely discharging and following up on patients who have been hospitalized, and diverse other topics.

Objectives: 1--Rapidly identify existing guidance for topics such as use, extended use, and re-use of medical masks and respirators; and present that information in a way that facilitates rapid decision-making and action. 2--Provide updated review and synthesis to hospital stakeholders as new guidance is made available and new clinical questions emerge.

Methods: We adapted an existing rapid HTA synthesis product to summarize clinical guidelines, policies from other hospitals, and key society position statements where available. The new prototype product is called a Rapid Guidance Summary. Seven medical students volunteered to assist with gathering evidence and writing reports. Updates to guidance sources were monitored using a combination of technology support (VisualPing, Vancouver, BC, Canada) and daily manual review.
We used simple evidence tables to describe the guidance, and concordance tables to report agreement and variations among different sources. Each report included a summary table on the first page reporting key recommendations and the quantity and source of of support for those recommendations. Reports were disseminated directly to chief medical officers and infection control staff at our affiliated hospitals and outpatient/home care entities and to all system providers via the hospital intranet site.

Results: In the first 10 days of the program, we completed 6 reports, one of which was an update and elaboration on the first report. Students served as lead analysts on two of those six reports. The mean time needed to complete a report was 3.5 days (range 1-5 days). Reports averaged 5 pages in length (range 2-9 pages), citing 4 or 5 major guidelines and 3 to 6 hospital policies.
Very little guidance was found for some topics, such as criteria for deeming a COVID-19 patient ready for discharge. In other areas such as personal protective equipment (PPE) requirements, there was considerably more guidance, and our challenge was to organize and present that information as succinctly as possible.
We identified several areas where guidelines and/or hospital policies were in disagreement. In other areas, hospital policies were frequently based on the same CDC and WHO guidance, so their agreement represented less of an expert consensus than it would if the policies were developed independently.

Conclusions: Evidence-based practice centers can provide timely guidance to clinicians and hospital administrators in a crisis, if centers are willing to adopt new methods as necessary and utilize non-traditional sources such as policy documents from peer hospitals.