1. Are hospitals providing clinical guidance on COVID-19 to their frontline workers? And if so, how is this guidance developed and disseminated?

• A resounding yes: Nearly all (98%) institutions provide some form of guidance on the evaluation and management of COVID-19.
• Stakeholders involved: The majority of institutions generate guidance through highly multidisciplinary teams including Infectious Disease (94%), Infection Control (84%), Hospital Medicine (84%), Critical Care (76%), Pulmonology (57%), and Emergency Medicine (45%).
• Update frequency: Guidance is most commonly updated as needed (59%) or multiple times per month (37%).
• Dissemination: The vast majority of sites provide guidance through a combination of methods, including email (84%), an institutional website (82%), a group wiki or document sharing site (57%), and training sessions for providers about to go on service (53%). In addition, the majority integrate clinical guidance with the electronic medical record (EMR) in the form of COVID-19-specific order sets (73%), note templates (65%), and best practice alerts (55%).
• Single source of truth: In addition to the multimodal approach above, the vast majority (84%) of sites have “a single source of truth” where frontline providers can efficiently access current recommendations. This was most commonly implemented as an institution-wide website, although a significant portion of Hospital Medicine groups have developed their own hospitalist-specific resource.
• Recommendations / takeaways: Clear and accessible guidelines are a vital tool to improve standardization of care and to help frontline providers synthesize large amounts of emerging data. Multidisciplinary input is likely necessary to improve buy-in. We were surprised that email was the most widely used modality for dissemination, and instead we recommend implementing a “single source of truth” and integrating guidance with the EMR in the form of order sets and note templates in order to improve ease-of-use for busy frontline providers, though the need for frequent updates may be difficult to manage at many sites.

Limitations of the survey / areas for future study: This study collected data from surveys rather than direct artifact review, given the diversity of tools utilized even at individual sites. In addition, we only obtained 1 survey response from each institution, typically from a hospital medicine leader. Therefore, we were unable to ascertain which modes of dissemination were most effective or preferred by frontline providers.