Background

Tracheal intubation for patients with COVID-19 is required for invasive mechanical ventilation. The authors sought to describe practice for emergency intubation, estimate success rates and complications, and determine variation in practice and outcomes between high-income and low- and middle-income countries. The authors hypothesized that successful emergency airway management in patients with COVID-19 is associated with geographical and procedural factors.

Methods

The authors performed a prospective observational cohort study between March 23, 2020, and October 24, 2020, which included 4,476 episodes of emergency tracheal intubation performed by 1,722 clinicians from 607 institutions across 32 countries in patients with suspected or confirmed COVID-19 requiring mechanical ventilation. The authors investigated associations between intubation and operator characteristics, and the primary outcome of first-attempt success.

Results

Successful first-attempt tracheal intubation was achieved in 4,017/4,476 (89.7%) episodes, while 23 of 4,476 (0.5%) episodes required four or more attempts. Ten emergency surgical airways were reported—an approximate incidence of 1 in 450 (10 of 4,476). Failed intubation (defined as emergency surgical airway, four or more attempts, or a supraglottic airway as the final device) occurred in approximately 1 of 120 episodes (36 of 4,476). Successful first attempt was more likely during rapid sequence induction versus non–rapid sequence induction (adjusted odds ratio, 1.89 [95% CI, 1.49 to 2.39]; P < 0.001), when operators used powered air-purifying respirators versus nonpowered respirators (odds ratio, 1.60 [95% CI, 1.16 to 2.20]; P = 0.006), and when performed by operators with more COVID-19 intubations recorded (odds ratio, 1.03 for each additional previous intubation [95% CI, 1.01 to 1.06]; P = 0.015). Intubations performed in low- or middle-income countries were less likely to be successful at first attempt than in high-income countries (odds ratio, 0.57 [95% CI, 0.41 to 0.79]; P = 0.001).

Conclusions

The authors report rates of failed tracheal intubation and emergency surgical airway in patients with COVID-19 requiring emergency airway management, and identified factors associated with increased success. Risks of tracheal intubation failure and success should be considered when managing COVID-19.

Editor’s Perspective
What We Already Know about This Topic
  • IntubateCOVID is a large, multinational, multispecialty, voluntary, self-reported database of healthcare workers who have performed intubations on patients with known or suspected COVID-19 established shortly after the widespread onset of the pandemic in March 2020. Data collection focuses on practitioner and hospital level characteristics related to the intubation, and no patient identifiable characteristics are collected. Practitioners record any subsequent symptoms suggestive of COVID-19 or positive tests for it.
What This Article Tells Us That Is New
  • The authors report a secondary analysis of associations of intubation and operator characteristics related to the primary outcome of first-attempt intubation success in 4,476 intubations among 1,722 clinicians at 607 institutions across 32 countries, also considering differential rates of success between high-income and low- and middle-income countries.
  • Although successful first-attempt intubation was noted in 89.7% of intubations, 0.5% required four or more attempts, an emergency surgical airway was required in 0.2%, and a composite variable of failed intubation occurred in 0.8%.
  • Multivariable analysis demonstrated that successful first attempts were more likely with rapid sequence intubations, when operators used powered air-purifying respirators, and with increasing operator experience.
  • Intubations performed in low- and middle-income countries were nearly half as likely to be successful on first attempt than in high-income countries.
  • These results provide potentially useful information for global and local policy-making related to this and future pandemics. However, the observational nature, along with lack of patient level characteristics, leave room for substantial residual confounding of these associations.