Lessons Learned From the Front Line: Outcomes of Noninvasive Ventilation for Coronavirus Disease 2019 Pneumonia in China.

The word “unprecedented” has been thrust into the vernacular during the current pandemic. Although historically, there have been numerous other, even more deadly pandemics, in this modern data-hungry era of rapid information dissemination, our common global experience in this moment is unique. As clinicians, we are accustomed to instantly accessing high-quality information to support our rational, evidence-based management decisions. We are comfortable that our decisions derive from sound medical reasoning and stand on a solid base of evidence, and are not pure conjecture.

Coronavirus disease 2019 (COVID-19) has yanked the comfort of having a base of supporting evidence out from under our feet. We no longer can operate within the relative certainty of guidelines or “best practice.” Our initial strategies were extrapolated from past experiences with other deadly coronavirus, but we have found that COVID-19 behaves quite differently (1). COVID-19 studies that are available are most often smaller, single institution, observational studies that limit their generalizability and applicability. When we can see a snapshot of critically ill COVID-19 patients with respiratory failure from a large, nationwide database in China, it provides some perspective and guidance on alternative management strategies. Globalization, while contributing to spread of the pandemic, can also provide some solutions for the rest of us, as we learn from our colleagues who have had to address these challenges first.

The study by Want et al (2), published in this issue of Critical Care Medicine, derived from a nationwide cohort of critically ill COVID-19 patients, describes 141 patients requiring ventilatory support. One-hundred twenty-two patients initially received noninvasive ventilation (NIV), of which 31 progressed to invasive mechanical ventilation (IMV) and 19 others were first supported on IMV. By segmenting a less sick subpopulation of the critically ill, it also shows the high mortality of patients who required IMV (50% mortality by the completion of the study by Want et al [2]). The study highlights three main areas of clinical interest: the preference for NIV in contradistinction to early intubation, the laboratory correlates among those requiring IMV, and the case demographics.

In the United States, early endotracheal intubation was the preferred initial strategy for critical care management because it secures source control in the patient and limits airborne exposure to the virus. Given the protracted course of COVID-related respiratory failure, it has generally been felt that IMV may lead to better outcomes though evidence for this is limited. NIV is more controversial as aerosolization of viral particles via high-flow oxygen delivery may place healthcare workers (HCWs) at higher risk and increase the need for personal protective equipment (PPE) at a time of widespread PPE shortages.

In China, they adopted a different approach. In the Diagnosis and Treatment Protocol for COVID-19 (Trial Version 7) released by National Health Commission and State Administration of Traditional Chinese Medicine in China (3), NIV is introduced first to COVID-19 patients with respiratory distress. IMV is recommended only if NIV failed to improve the respiratory distress or hypoxemia. Historically, European Respiratory Society guidelines have not recommended NIV for improving hypoxemia in patients with a pandemic viral illness due to insufficient evidence of benefit and the risk of nosocomial transmission (4). However, NIV has been gaining attention as an alternative means of respiratory support in clinical practice. Early application of prone position coupled with NIV in patients with viral pneumonia and moderate-to-severe acute respiratory distress syndrome (Pao2/Fio2 50–200 mm Hg) decreased the need of intubation (5). The markedly higher and dismal 50% mortality rate in the current study by Want et al (2) in patients requiring IMV compared with those managed with NIV alone (hazard ratio 2.95) strongly supports the goal of avoiding IMV to improve outcomes. In addition, use of NIV was associated with decreased secondary infections compared with the IMV cohort (16.5% vs 28%) although the study was not powered for this outcome. Longer term outcome data in the study by Want et al (2) are limited by the fact that many patients still remained hospitalized when data review was completed, and thus the true hospital mortality rate may be underreported.

Risk of COVID-19 infection in HCW remains an ongoing concern, particularly in the absence of source control. During the initial phase of the COVID-19 outbreak in China, 3,387 HCW were reported to have been infected. The authors note though that, with appropriate availability and use of PPE, HCW infections could be minimized even with the application of NIV. This is evidenced by the few HCW infections during the later period of the China outbreak despite wide use of NIV (6–8). In a study of infection in HCW in Wuhan, China, the HCW infection rate was only 1.1% and was lower on the COVID units than those in the low contagion units, also suggesting that with appropriate PPE, risk of nosocomial transmission is low (9).

In their analysis, the authors suggests that markers such as elevated admission D-dimer levels may predict patients requiring IMV and the note that elevation in C-reactive protein was associated with progression from NIV to IMV. Other markers recently reported to be associated with poorer outcomes such as interleukin-6 and ferritin were not available in this cohort. Targeting lower risk profile individuals for whom noninvasive strategies may be successful can facilitate resource allocation and appropriate ventilator usage and may improve patient outcomes. Other publications from the nationwide China cohort that this subset of invasively ventilated patients was abstracted from have demonstrated that clinical and laboratory profiles at admission are associated with risk of more critical illness (10).

The reported at-risk populations in this early COVID-19 population (male, older, smokers, hypertension) have now been verified in other populations (11). An additional, commonly described risk factor not reported in this cohort was obesity, which may reflect differences in prevalence of obesity in the Chinese population. In this Chinese cohort, chronic obstructive pulmonary disease and chronic renal disease were associated with higher risk for death for those requiring ventilation. These patients may just represent a sicker presenting cohort, but poor outcomes could potentially be related to the pathophysiology of the COVID 19 disease process, particularly our still-evolving understanding of the role of angiotensin-converting enzyme 2 receptors (12, 13).

At the time, this article was written, the database for effective COVID-19 directed therapeutics was nonexistent. We now have the benefit of studies suggesting several promising treatment modalities in our arsenal, including direct antiviral agents such as remdesivir as well as agents targeting the severe cytokine mediated inflammatory response important in the pathogenesis of COVID-19–associated respiratory failure (14, 15). Further studies investigating earlier initiation of these COVID targeted therapies in conjunction with a less invasive respiratory management strategies may demonstrate even more benefit in preventing need for mechanical intubation and the associated poor prognosis.

While the study by Want et al (2) does not allow for more subtle data interpretation, it does provide strong justification for the path of less invasive ventilator management and may help guide complex ventilation decisions in a time where information is sparse. Not by choice, our Chinese colleagues were the first to have to manage critically ill patients during this evolving pandemic and had to learn as they went. As the COVID-19 pandemic continues to across the world, the lessons learned from those first responder nations will be crucial to integrate into our still evolving global managem ent paradigm. These paradigms need to focus on optimal patient outcomes while simultaneously integrating our more complete understanding of transmission risks in the critically ill population. Stratifying populations of patients who can safely be started on NIV saves ventilators for those who need them may buy critical additional time for improving antiviral and antiinflammatory therapeutics to take effect and may not place healthcare providers in harm’s way if appropriate use of PPE is maintained.