Background: Evidence-based COVID-19 management has evolved with unprecedented speed; however, the rate and fidelity with which institutions have adopted new evidence into practice is unknown.

Methods: We surveyed members of the Hospital Medicine Reengineering Network (HOMERuN) from 12/17/20-2/10/21 and compared their institutional COVID-19 management recommendations to available evidence from pivotal randomized controlled trials (RCTs) and treatment guidelines from the National Institutes of Health, Infectious Diseases Society of America, and American Society of Hematology. This study was deemed non-human subjects research.

Results: Of 83 sites, 52 (63%) responded, 51 (98%) of whom issued internal COVID-19 management guidance. Multidisciplinary committees generated recommendations. Membership included infectious diseases (98%), hospital medicine (88%), infection control (86%), and critical care (80%). Recommendations were disseminated through email (84%), institutional websites (82%), and integration into the electronic health record as COVID-19 specific ordersets (73%) and note templates (65%). The percentage of institutions recommending each studied intervention for specific patient populations is shown in Figure 1, alongside simplified guidelines and RCT data.

Conclusions: Three themes emerged. First, translation from evidence to practice was remarkably rapid for interventions supported by aligned national guidelines and high-quality studies. A striking example is the 94-100% adoption of dexamethasone for patients requiring >4L of oxygen – only 6-8 months after the RECOVERY trial demonstrated a survival benefit. The lone exception to this trend was baricitinib; however, evidence and guidelines were released the week prior to our study’s release, potentially marking the translational speed limit for academic medical centers (AMCs). This rapidity trend was also observed when evidence and guidelines converged against interventions, although leaving improvement opportunities as only 80% of institutions recommended against use of dexamethasone for patients with SpO2 >94%. Second, institutions favored treatment over not treatment. 69% of sites recommended remdesivir for mechanically-ventilated patients and 81% recommended dexamethasone for patients requiring 1-2L of oxygen. We suspect this reflects biases to “just do something” when uncertain, likely exacerbated by inconsistent study and guideline disease severity definitions. Third, AMCs innovated – 67% of sites implemented awake proning, 35% limited remdesivir use to “early or viral phase of illness” for specific subgroups, and 17% recommended therapeutic anticoagulation for specific subgroups. This demonstrates AMC willingness to bridge knowledge gaps with expert opinion. AMCs are capable of responding nimbly to emerging data and shifting guidelines, though we found both overtreatment and innovation in response to knowledge gaps. As COVID-19 continues to rapidly spread, with new data in its wake, we hope that AMCs sustain their flexible and rapid approaches to clinical care standardization. While factors unique to the pandemic likely shaped this successful response, we hope some strategies – such as use of focused multidisciplinary teams and novel information sharing tools – can be harnessed to accelerate the translation of evidence to bedside for COVID-19 and beyond.

IMAGE 1: Figure 1: Results Summary Table