COVID-19 management has evolved rapidly, creating challenges for implementation. This study was conducted to assess the fidelity with which academic medical centers (AMCs) adopted evidence into practice.
This survey study was deemed exempt from review and informed consent by the University of California, San Francisco, because it does not involve human participants. Response rates were computed according to the American Association for Public Opinion Research (AAPOR) reporting guideline.
We surveyed members of the Hospital Medicine Reengineering Network (Figure 1)1 from December 17, 2020, to February 10, 2021, and compared their institutional recommendations for COVID-19 management with available evidence at that time from pivotal randomized clinical trials (RCTs) (eAppendix in the Supplement) and guidelines from the National Institutes of Health,2 Infectious Diseases Society of America,3 and American Society of Hematology.4 Data were analyzed from February 10 to March 4, 2021.
Of 83 hospitals contacted, 52 (63%) responded. Hospitalist leaders involved in the direct care of patients with COVID-19 provided responses. A total of 49 sites (94%) self-identified as AMCs; the remaining 3 sites (6%) identified as AMC-affiliated teaching hospitals. Fifty-one sites (98%) issued internal COVID-19 management guidance. Guidance at 48 sites (94%) was generated by multidisciplinary committees, including infectious disease (47 sites [98%]), infection control (43 sites [90%]), hospital medicine (42 sites [88%]), and critical care (40 sites [83%]). Of 51 sites with internal COVID-19 management guidance, recommendations were disseminated most commonly through email (43 sites [84%]), institutional websites (42 sites [82%]), and integration into the electronic health record as COVID-19-specific order sets (37 sites [73%]) and note templates (33 sites [65%]). The percentage of institutions recommending each studied intervention is shown in Figure 2, alongside simplified RCT findings and guidelines. Notable results include 94% to 100% of sites recommending dexamethasone for patients requiring at least 4 L of oxygen, 69% recommending remdesivir for patients receiving mechanical ventilation, 81% recommending dexamethasone for patients requiring 1 to 2 L of oxygen, 67% implementing awake proning, 35% limiting remdesivir use to the “early or viral phase of illness,” and 17% recommending D-dimer–based therapeutic anticoagulation. The proportion of sites recommending each intervention varied from a low of 10% for convalescent plasma for patients without antibodies to 100% for dexamethasone in patients requiring mechanical ventilation.
In this survey study, 3 themes emerged from our analysis. First, translation from evidence to practice guidelines was remarkably complete for interventions supported by aligned national guidelines and high-quality studies. A striking example is the near universal adoption of dexamethasone among patients requiring at least 4 L of supplementary oxygen only 6 to 8 months after the RECOVERY trial demonstrated a survival benefit. The lone exception to this trend was baricitinib; however, new evidence and guidelines were released within 1 week of survey distribution. Practice convergence was also observed when evidence and guidelines aligned against interventions, as seen in the infrequent recommendation of dexamethasone for patients with oxygen saturation greater than 94%, as measured by pulse oximetry. However, clear opportunity for improvement still exists.
Second, institutions favored treatment over not treatment, particularly when guidelines diverged from each other or from the underlying evidence, as exemplified by 69% to 81% of sites recommending remdesivir or dexamethasone, respectively, when evidence or guidelines conflicted. We suspect this finding reflects systemic biases to do something rather than nothing when faced with uncertainty, likely exacerbated by inconsistent definitions of disease severity across studies and guidelines.
Finally, AMCs demonstrated a willingness to innovate across a range of interventions. Novel interventions, such as awake proning, phase-of-illness–restricted remdesivir, and D-dimer–based therapeutic anticoagulation, varied widely with respect to patient selection and procedural specifics, but collectively, and with 17% to 67% of sites recommending 1 or more of such novel interventions, they demonstrate that AMCs were sophisticated consumers of information, willing to bridge knowledge gaps with expert opinion. Additional research is needed to understand how AMCs monitor innovation outcomes and deimplement practices when negative evidence emerges.
Limitations of this study include a 37% nonresponse rate, reliance on self-reporting, lack of longitudinal follow-up, and lack of data on actual clinical practice and outcomes. Nonetheless, our findings demonstrate that AMCs were capable of responding nimbly to emerging data and shifting guidelines, although both overtreatment and experimentation were observed where significant uncertainty persisted. While factors unique to the early pandemic likely shaped this performance, 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.
Accepted for Publication: January 16, 2022.
Published: April 4, 2022. doi:10.1001/jamanetworkopen.2022.5657
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Berger AC et al. JAMA Network Open.
Corresponding Author: Alan A. Kubey, MD, Thomas Jefferson University Hospital, 833 Chestnut St, Ste 701, Philadelphia, PA 19107 (alan.kubey@jefferson.edu).
Author Contributions: Drs Berger and Kubey had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Berger and Simchoni contributed equally as co–first authors.
Concept and design: Berger, Simchoni, Auerbach, Brode, Raffel, Kubey.
Acquisition, analysis, or interpretation of data: Berger, Simchoni, Auerbach, Kuperman, Raffel, Kubey.
Drafting of the manuscript: Berger, Simchoni, Auerbach, Kubey.
Critical revision of the manuscript for important intellectual content: All authors.
Obtained funding: Auerbach.
Administrative, technical, or material support: Berger, Auerbach, Kubey.
Supervision: Auerbach, Raffel, Kubey.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by the Gordon and Betty Moore Foundation (grant No. 8856) and Agency for Healthcare Research and Quality (grant No. R01HS027369) for the Hospital Medicine Reengineering Network (HOMERuN) COVID-19 Collaborative.
Role of the Funder/Sponsor: The sponsors had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.