The revenge of the ‘late COVID adopters’


The COVID-19 pandemic has stressed all aspects of the world’s health care systems. The sheer volume of pandemic-related research produced over the past year has been challenging to process. This is as it should be, given its unprecedented spread and related morbidity and mortality. However, such rapid production and application leaves little time for proper vetting. Large numbers of providers adopted suggested, but largely unproven, practices that deviated from pre–COVID-19 guidelines. These “early adopters” theorized that COVID-19–related disease processes were different, necessitating a modification to existing practices.

Dr. Aaron Holley

Dr. Aaron Holley

While many unproven approaches were suggested and implemented, I’ll focus on two approaches. First, throughout the pandemic, many have argued that COVID-19 causes a novel acute respiratory distress syndrome (ARDS) phenotype. Early on, a group of prominent Italian ARDS researchers made a compelling case for physiological differences, concluding that early intubation was required to avoid large transpulmonary pressure swings. The logic was that COVID-19 causes significant gas-exchange abnormality without the typical effect on elastance. The resulting increase in respiratory drive would generate vigorous inspiratory effort, overstretch a relatively compliant lung, and lead to further injury.

Other equally prominent researchers countered this argument. Martin Tobin drew on physiology, while Arthur Slutsky and Niall Ferguson used emerging data to make their case. Tobin and colleagues cautioned against early intubation for anyone who could be maintained using noninvasive support. In August 2020 (well into the pandemic and after more data were available), Slutsky and colleagues argued that ARDS caused by COVID-19 wasn’t much different from lung injury due to other causes.

Two more recent studies published online recently are relevant to the debate over COVID-19 ARDS. One was a prospective study and the other a retrospective study; both had comparison groups, and both came to the same conclusions. Overall, COVID-19 ARDS isn’t much different from ARDS due to other causes. These studies were comprehensive in their comparisons and measures of outcomes, but they were both rather small and included patients from one and two hospitals, respectively. The discussions of both provide a nice review of the existing literature on COVID-19 ARDS.

A second controversial, but unproven, COVID-19 practice is aggressive anticoagulation. Early reports of a high prevalence of venous thromboembolism (VTE) in patients with COVID-19 pushed many to recommend empirically increasing prophylaxis. Most of the data guiding this approach were from retrospective, observational studies that suffered from selection bias. Early on, many of the studies were from China, where baseline VTE prophylaxis rates were low. Despite these limitations, many physicians acted on the basis of these data. An arbitrarily defined “intermediate” or treatment dose for prophylaxis was used, with some measuring D-dimer to guide their approach. An evidence-based argument against this practice, published in the New England Journal of Medicine, failed to sway readers. (Look at the poll at the end of the article and you’ll see how readers answered.)

Two articles recently published online in CHEST attempted to bring clarity to the debate over COVID-19 and VTE prophylaxis. The first study evaluated critically ill patients in France, and researchers found that higher doses of anticoagulation reduced thrombotic complications without an associated increase in bleeding events. The study is well done but certainly has its flaws. It is observational and retrospective, and it essentially uses a before-after comparison technique. Such an approach is particularly prone to bias during COVID-19, given that practice patterns change quickly.

The second paper is a systematic review looking at VTE and bleeding rates among patients hospitalized with COVID-19. The authors found high rates of VTE (17.0% overall), with screening, admission to the ICU, and the prospective study design all being associated with increased rates. Of importance, unlike the retrospective trial cited in the previous paragraph, the authors of the systematic review found treatment-dose anticoagulation was associated with higher bleeding rates.

I admit, the title of this piece is a bit of a misnomer. The “late adopters” would truly have their revenge if deviation from guidelines for COVID-19–related ARDS and VTE prophylaxis proves to be harmful. It’s not clear that’s the case, and at least for VTE prophylaxis, results from several randomized, controlled trials (REMAP-CAP, ATTACC, and ACTIV-4a) will be released soon. These are sure to provide more definitive answers. If nothing else, the COVID-19–related ARDS and VTE data reinforce how difficult it is to obtain high-quality data that yield clear results. Until something more definitive is published and released, I will remain a “late adopter.” Standard non–COVID-19 guidelines for ARDS and VTE prophylaxis are good enough for me.

Dr. Holley is program director of the Pulmonary and Critical Care Medical Fellowship at Walter Reed National Military Medical Center, Bethesda, Md.

A version of this article first appeared on

Next Article:

   Comments ()