2.2 Study design, objectives and measurements
We analysed procedure timing and levels of ventilatory support at time of tracheostomy (PEEP ≤ 10 or Fi02 ≤ 0.5). We also documented surgical approach and duration, procedural complications, outcomes, mortality, and staff safety.
 Patient demographics included age, gender, ethnicity, BMI, very severe co-morbidities, and APACHE II score.  We compared our cohort to the ICNARC one that underwent advanced respiratory support during their critical care admission [1].  We measured ventilator settings (PEEP, Fi02) between 7am and 9am on the day of the procedure.
 Outcome measures were recorded from day of tracheostomy until the day when:
·       Sedation stopped: Intravenous sedative infusions (Propofol, Fentanyl, Alfentanil or Midazolam) successfully held for 24 hours
·       Ventilation stopped: No mechanical ventilation required (including BiPAP / CPAP) for at least 24 hours
·       Patient discharged from ITU
·       Decannulation of tracheostomy
·       Survival
We separately analysed the outcomes of those who underwent tracheostomy on day <14 vs ≥14.
 For the study duration, all team members involved were surveyed and asked to report COVID-19 symptoms [8], and volunteer whether they had COVID-19+ PCR test. Results were included if this occurred within 5-14 days after a procedure.
 Descriptive data is presented and analysed. Tests for significance were performed using a Chi-Squared test and Mann-Whitney U Tests.