Skip to content

CT Should Not Be Used to Guide Ventilation in COVID-19 Respiratory Failure

The FLARE Four

  • Rapid division of COVID-19-related acute respiratory distress syndrome (ARDS) into "H" and "L" subphenotypes has been proposed based on qualitative descriptions of patients at a single center
  • A new study rigorously tested "H" and "L" subphenotypes and found no relationship between respiratory system compliance and degree of consolidation/non-aerated tissue as depicted by CT
  • Bedside measurement and titration of ventilator settings should be based on individually measured mechanics

Early in the COVID-19 pandemic, rapid division of acute respiratory distress syndrome (ARDS) into "H" and "L" subphenotypes was proposed based entirely on qualitative descriptions of patients at a single center. Subsequent publications have failed to validate these subphenotypes, but discussions persist. In part this is due to a perception that imaging abnormalities in patients with COVID-19 are unexpectedly mild compared with prior experience, despite evident clinical severity.

A recent study, published in the Annals of the American Thoracic Society, comprehensively explored the relationship between respiratory system mechanics and findings on chest CT. The results are reported and put in context by the fast literature assessment and review (FLARE) team at Massachusetts General Hospital, with advisory review by Kathryn Hibbert, MD, director of the Medical ICU and physician in the Division of Pulmonary and Critical Care, in an update posted on May 14, 2020.

Subjects and Methods

Bos et al., retrospectively analyzed 38 critically ill, intubated COVID-19 patients who had chest CT images available. Patients were ventilated using low tidal volume (6 mL/kg) and positive end-expiratory pressure of approximately 10 cm H2O.

The fraction of consolidated lung by CT scan was determined by classifying all areas above −500 Hounsfield units as consolidated, and patients were categorized according to quartiles of the fraction of involved lung (0–25%–50%–75%–100%). The radiographic phenotype was further classified as focal or non-focal. Respiratory system compliance was calculated for each patient immediately after intubation.


The results showed heterogeneity in lung morphology, consistent with prior reports on chest CT in COVID-19 (reviewed in a previous FLARE).

Respiratory system compliance was heterogeneous, also consistent with prior reports (covered in another previous FLARE). After matching patient respiratory system compliance with CT findings, the authors were not able to divide patients consistently into the proposed "H" and "L" subphenotypes.


As the authors acknowledge, there were several limitations to the study:

  • CT analysis is semi-quantitative
  • There was no assessment of recruitability
  • The total number of patients was relatively small

Still, based on this study and prior research, clinicians should not assume that respiratory system mechanics correlate with CT scan patterns. Bedside measurement and titration of ventilator settings should be based on individually measured mechanics.

View all COVID-19 updates

Visit the Division of Pulmonary and Critical Care Medicine

Related topics


Massachusetts General Hospital researchers find that there's reasonably strong evidence for low tidal volume ventilation in non-ARDS respiratory failure, although it is not without risks.


According to Massachusetts General Hospital clinicians, observations of hypoxemia in COVID-19 are entirely consistent with known mechanisms and do not require the invocation of novel processes specific to SARS-CoV-2.