Abstract
Introduction
The severe acute respiratory syndrome corona virus-2 (SARS-CoV-2) pandemic has resulted in significant changes to the use of personal protective equipment (PPE) in healthcare institutions worldwide. It is thought transmission is primarily via droplet spread 1 but there are fears that aerosolized particles (generally <5 micron diameter) 2 could cause airborne transmission. 3 Many organisations have advised the use of PPE for those undertaking certain activities that have been designated as aerosol generating procedures (AGPs) to minimise this risk. Examples include percutaneous tracheostomy and bronchoscopy 4 but the quantitative evidence base for AGP designation is weak and the relative risk of each procedure unknown. 5 To date aerosols generated during percutaneous tracheostomy have not been measured. Our previous study identified that background particle counts in most settings are too high to allow reliable aerosol detection. We had a unique opportunity to measure aerosol generation during percutaneous tracheostomy insertion in an ultra-clean operating theatre.
Methods
A prospective environmental monitoring study was in progress at North Bristol NHS Trust to quantitate aerosol generation during operating theatre procedures. 6 Ethical approval for this study was given by the Faculty of Life Science and Science Research Ethics Committee at the University of Bristol and this waived the need for individual patient consent. Our ICU’s COVID-19 percutaneous tracheostomy insertion guideline mandated insertion in an operating theatre allowing us to measure aerosols during the procedure.
A lightweight, portable Optical Particle Sizer (TSI Incorporated, model 3330, High Wycombe, UK), connected to a plastic funnel (custom 3 D printed from PLA by RAISE3D Pro2 Printer, 3DGBIRE, Chorley, UK, max. diameter 150 mm) and sampling through silicone tubing (TSI Incorporated, model 3001788, length 2 m, internal diameter 4.8 mm) was used to quantify aerosol generation, with a sampling resolution of 1 s. Observations were made at a distance of 20–30 cm from the airway and surgical site within an operating theatre with an ultraclean laminar flow ventilation system with the airflow temporarily set to standby.
The tracheostomy procedure was performed as per our standard practice and was divided into 7 separate phases for analysis (Figure 1). The funnel was above the mouth in phases I and IV, and above the neck for phases (II, III, V and VI). The end-point was specified as inflation of the tracheostomy tube cuff.
(a) Concentration of aerosol particles in the diameter range 0.3 – 10 µm sampled during a tracheostomy procedure. The five phases indicated correspond to; I. the “anaesthetic phase” including pharyngeal suction, cuff deflation and repositioning II. the “preparation phase” including skin cleaning, swabbing and draping; III. the “seldinger phase” involving needle, cannula and guide wire; IV. bronchoscopy including port opening; V. dilatation and tracheostomy insertion; VI. ventilation, initially with cuff down and including disconnection.
Results
We sampled continuously during the 18 minutes and 41 seconds taken to complete the procedure (Figure 1). The average concentrations of particles recorded during the procedure were as follows (concentrations quoted as an average sampled over a 10 second period): Phase I – 0.061 particles/cm3; II – 0.27 particles/cm3; III – 0.019 particles/cm3; IV – 0.02 particles/cm3; V – 0.02 particles/cm3; VI – 0.11 particles/cm3.
Discussion
We have measured particles generated during percutaneous tracheostomy insertion in an ultraclean theatre environment. The background particle count was extremely low: 0.03 particles/cm3 averaged over a 10 s period.
Highest particle counts were detected during the preparation phase (II – 0.27 particles/cm3) measured above the neck, after the tube was repositioned and cuff re-inflated before tracheostomy insertion. We have previously identified that use of a dry woven swab generates a large particle count and it is likely that these particles account for the count seen.
We measured very low particle counts during the airway components of the procedure (I – 0.061 particles/cm3, IV – 0.02 particles/cm3, VI – 0.11 particles/cm3). This included using a Yankauer sucker in the oropharynx and a period of endotracheal tube cuff deflation. In addition we carried out bronchoscopy via a suction port and a number of ventilator circuit disconnections, all with minimal particle counts.
During the tracheostomy phase there was an audible leak from the tracheal site with visible blood droplet formation however this did not produce significant particle counts (III – 0.019 particles/cm3; V – 0.02 particles/cm3).
Our patient was deeply anaesthetised and paralysed and therefore unable to breathe or cough during the procedure. We detected very low particle counts during tracheostomy and airway manoeuvres. These were between 10 and 100 times less than during a normal cough in the same setting (Figure 1; from previous data, average volunteer coughs (n = 38) generated a total of 8.2 ± 0.8 particles/cm3, with peak concentration of 1.2 ± 0.1 particles/cm3). By an explicit UK definition aerosol generating procedures are those considered to have a greater likelihood of producing aerosol than coughing. 7 Our study shows that on this basis tracheostomy insertion and bronchoscopy in paralysed patients may not need to be considered “aerosol generating procedures”.
This is the first report quantifying aerosol generation during percutaneous tracheostomy. Further measurements in an ultra clean environment are needed to confirm these findings.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We acknowledge support from the Elizabeth Blackwell Institute with funding from the EBI Rapid Response Call (COVID-19).
