Safety Assessment of a Noninvasive Respiratory Protocol for Adults With COVID-19

Abstract

As evidence emerged supporting noninvasive strategies for coronavirus disease 2019 (COVID-19)-related respiratory distress, we implemented a noninvasive COVID-19 respiratory protocol (NCRP) that encouraged high-flow nasal cannula (HFNC) and self-proning across our healthcare system. To assess safety, we conducted a retrospective chart review evaluating mortality and other patient safety outcomes after implementation of the NCRP protocol (April 3, 2020, to April 15, 2020) for adult patients hospitalized with COVID-19, compared with preimplementation outcomes (March 15, 2020, to April 2, 2020). During the study, there were 469 COVID-19 admissions. Fewer patients underwent intubation after implementation (10.7% [23 of 215]), compared with before implementation (25.2% [64 of 254]) (P < .01). Overall, 26.2% of patients died (24% before implementation vs 28.8% after implementation; P = .14). In patients without a do not resuscitate/do not intubate order prior to admission, mortality was 21.8% before implementation vs 21.9% after implementation. Overall, we found no significant increase in mortality following implementation of a noninvasive respiratory protocol that decreased intubations in patients with COVID-19.

FIG

Interrupted Time Series Analysis of Intubation Rates by Date of Arrival. Circle size represents relative number of patients seen each day. The slope of the preimplementation intubation rate increased about 2% per day (incidence rate ratio (IRR), 1.015; P = .54) whereas after the implementation of the noninvasive COVID-19 respiratory protocol, the slope decreased about 2% per day (IRR, 0.977; P = .56). Although there was a significant change in level after implementation, (IRR, 0.44; 95% CI, 0.23-0.83; P = .012), the slopes prior to and after the protocol implementation were not statistically different (P = .41).