Frequency and outcomes of critically ill COVID-19 patients with tracheostomy, a retrospective two-center cohort study.

BACKGROUND

The optimal use of tracheostomy in COVID-19 patients is debated, and considerable uncertainties on the frequency, timing, and outcomes of tracheostomy remain. The objective was to study the frequency and timing of tracheostomy in a real-world population of critically ill COVID-19 patients. The secondary aim was to study whether early tracheostomy was associated with days alive and out of intensive care unit (ICU), days free of invasive mechanical ventilation (IMV), 60-day mortality, ventilator weaning rate, and ICU discharge rate compared to late tracheostomy.

METHODS

The study is a retrospective two-center cohort study. All COVID-19 patients admitted to critical care in the Region Östergötland County Council, Sweden, between March 2020 and September 2021 were included. Early (≤10 days from tracheal intubation) and late (>10 days) tracheostomy were compared. Through the Swedish intensive care registry, 249 mechanically ventilated COVID-19-positive patients ≥18 years old with respiratory failure were included. The pre-defined primary outcomes were the frequency and timing of tracheostomy. Secondary outcomes were days free of mechanical ventilation and intensive care, ICU discharge rate, ventilator weaning rate, and 60-day mortality.

RESULTS

Of 319 identified patients (70% men), 249 (78%) underwent endotracheal intubation. Of these, 145 (58%) underwent tracheostomy and 99 (68%) were performed early. Tracheostomy patients (vs. non-tracheostomy) had fewer IMV-free days and ICU-free days (27 [0-43] vs. 52 [43-55], p < .001, and 24 [0-40] vs. 49 [41-52], p < .001). Late (vs. early) tracheostomy patients had fewer IMV- and ICU-free days (16 [0-31] vs. 36 [0-47], p < .001 and 8 [0-28] vs. 32 [0-44], p < .001). Early tracheostomy (vs. late) was associated with a significantly higher ICU discharge rate (adjusted HR = 0.59, 95% CI [0.40-0.86], p = .006), but not with the weaning rate (adjusted HR = 0.64, 95% CI [0.12-3.32], p = .5) or 60-day mortality (adjusted HR = 1.27, 95% CI [0.61-2.67], p = .5).

CONCLUSIONS

Tracheostomy is common in critically ill COVID-19 patients. In patients predicted to need a tracheostomy at some point, early, rather than late, tracheostomy might be a means to reduce the time spent in ICU. However, we do not have sufficient evidence to suggest that early tracheostomy reduces mortality or weaning rates, compared with late tracheostomy.