Data collection and statistics
Data were recorded on a Google Sheets spreadsheet (Google LLC, Mountain View, CA, US), anonymised and extracted. Descriptive statistics were performed by transferring data into an Excel 2010 spreadsheet (Microsoft Corp, Redmont, WA, US). Due to the small sample, data were considered as nonparametric.
RESULTS
The English translation of the full questionnaire with responses is available online as supporting information to this paper. Each director from the 42 Otolaryngology departments in Lombardy answered the questionnaire. Among them, two belonged to cancer centres and one to a paediatric hospital. All the hospitals, minus the two cancer centres, admitted COVID-19 patients into their wards and ICUs. The 40 hospitals accounted for almost 9000 hospital and approximately 1000 ICU COVID-19-dedicated beds. The two cancer centres were devoted to managing all complex cancer cases from the region, and patients were transferred whenever possible.
At the time of interview (six weeks after Italy’s outbreak), over 500 tracheostomies were reported for intubated critical COVID-19 patients in Lombardy. In 38 of 40 involved hospitals, at least one procedure was performed. However, the number of performed procedures was inconsistent among centres, regardless of the number of long-term invasively ventilated patients. Notably, nearly half reported fewer than ten procedures.
In all cases, the anaesthesiologist determined a tracheostomy need, but we noticed tremendous time and preferred technique discrepancies among centres. We found the median time between symptom onset and ICU admission was 10 ± 3, 5 days (range 1–18 days), and since then, the median time between intubation and tracheostomy was 12 ± 4 days (range 3–18 days). Pronation cycles were tried in each centre before prescribing tracheostomies but significantly differed among hospitals. A median of 3,5 ± 1,5 pronation cycles was performed before tracheostomy (range 1–18 cycles).
Concerning the preferred technique, 27 of 38 centres that reported at least one tracheostomy performed exclusively or mostly PTs. Among the remaining hospitals, in five cases PTs and STs were equally performed, whereas in six cases STs were preferred to PTs. No association was found between the number of performed procedures and preferred techniques.
Reasons for the chosen techniques are reported in Table 1. Operator preference was the primary determiner for choosing PTs. Additional reasons frequently cited were the superior postoperative management of the cannula and lower risk of complications. Open STs were mainly utilised for unfavourable anatomic conditions but also from a lack of experienced PT staff or PT kits.
DISCUSSION