The severe acute respiratory distress syndrome coronavirus 2 (SARS-CoV-2) pandemic has put an unprecedented strain on healthcare systems worldwide. Even well-resourced countries with comparably high numbers of intensive care beds per capita not only reached but quickly went beyond capacity with supersurge arrangements being implemented in places. After the original major outbreak in Wuhan and other regions in China (1), Italy followed by France, the Benelux countries, Spain, and the United Kingdom (UK) became particular disease hotspots in Europe. The epicenter of the pandemic subsequently shifted to the United States followed by Latin America and the Indian subcontinent. Despite the very disparate distribution of healthcare resources and spending across nations and continents, the logistics and pressures in relation to health infrastructure are similar around the globe, such as provision of adequate quantities of mechanical ventilators, medication, and even most basic concerns around hospital oxygen reserves, supplies, such as personal protective equipment (PPE), and not the least manpower.
Many interventions at the beginning of the pandemic were based on the experience gained from recent but limited global outbreaks of novel viral respiratory diseases such as severe acute respiratory syndrome (SARS), influenza A virus subtype H1N1, Middle Eastern respiratory syndrome (MERS), or on expert opinion. Where these previous viral outbreaks were associated with a higher case fatality rate, transmission rates for coronavirus disease 2019 (COVID-19) far exceed those of MERS and SARS due to the high incidence of minimally or asymptomatic carriers and an at the beginning uncertain incubation period, during which patients were able to transmit the disease. In addition, acute respiratory distress syndrome (ARDS) induced by COVID-19 appears to present clinicians with unique challenges due to a combination of factors leading to and perpetuating lung injury. Pulmonary vascular and endothelial pathology with abnormal adaptive vasomotor responses and perfusion, a relatively high incidence of pulmonary micro- and macrovascular thrombosis, neurotropic and pathic effects of the virus affecting central control of breathing, inflammation triggered increase in basal metabolic rate and oxygen extraction and increase in ventilation perfusion mismatch from high cardiac output states and finally iatrogenic or spontaneous breathing-induced lung damage (2). Unlike other forms of acute lung injury and ARDS, recovery from COVID-19-induced ARDS appears to be delayed in the range of weeks and sometimes months (1).
As Europe and the United states are entering a second wave of the pandemic, the experience gained during the last 8–9 months will undoubtedly improve the way we are treating in particular the sickest cohort of COVID patients, yet health infrastructural and capacity challenges do remain.
Tracheostomy for patients who are anticipated to not readily wean from mechanical ventilation is a common procedure performed in ICU. It may allow for faster weaning of sedation in patients receiving invasive mechanical ventilation; in some studies, this was associated with shorter ICU stay and it reduces anatomical dead space, facilitates airway toilette, and improves patient comfort (3,4). It is also associated with a reduced risk of subglottic and tracheal stenosis. In addition, some authors (including our own experience) have reported an increased incidence of laryngeal edema (5) and problems with tenacious secretions and airway bleeding occurring more frequently in COVID-19 patients; also prolonged duration and frequency of prone positioning can lead to significant facial edema making extubation attempts hazardous. The risks associated with both percutaneous and surgical or hybrid methods of tracheostomy insertion are generally low and even in patients with coagulopathy, obesity, or under other anatomically challenging situations. A number of studies including randomized controlled trials have tried to answer the question as to whether early tracheostomy performed within 7 days is associated with improvement in outcomes compared with tracheostomy after 7 days or no tracheostomy, with until now persistent clinical equipoise (3).
Both surgical and percutaneous tracheostomies are aerosol-generating procedures (AGP), a particular concern in the context of high-consequence infectious diseases transmitted by droplets and aerosols. At the outset of the pandemic and based on the experience from, in particular, the SARS outbreak in 2003, a number of professional societies advocated against performing tracheostomies all together or delay until viral polymerase chain reaction (PCR) test turned negative (6–8). A number of early fatalities among healthcare workers particularly in China were attributed to droplet exposure and AGP, mostly in situations when infection status was unknown or not suspected and at the very beginning of the outbreak when adequate PPE may not have been utilized. In turn, the presence of prolonged viral shedding and low levels of RNA detected by PCR do not necessarily equate with ongoing infectivity (9); healthcare workers may not be exposed to a particular high risk when performing AGPs despite persistence of positive viral RNA. In turn, delaying tracheostomy until RNA turns negative on PCR testing will expose patients to unnecessary long duration of endotracheal intubation and complications.
In this issue of the journal Critical Care Medicine, Rosano et al (10) from Brescia in the Lombardy region of Italy, the epicenter of the first wave of the pandemic in Italy and the initial area to be severely affected in Europe, report on their experience of early tracheostomy in patients undergoing invasive mechanical ventilation for SARS-CoV-2. Of more than 2,000 patients admitted to their hospital, 181 required invasive mechanical ventilation. After exclusion of 17 patients who either died n = 13 or were discharged during the first 3 days of admission, the authors analyzed the remaining 164 patients in relation to outcomes and based on whether they underwent early versus no tracheostomy. From day 4 onward, patients were assessed and categorized as to whether they were likely to need invasive ventilation for longer than 7 days. A total of 121 patients subsequently underwent percutaneous tracheostomy at an average of 6–7 days of mechanical ventilation. Compared with patients managed with translaryngeal ventilation throughout their stay, tracheostomized patients appeared to have a lower mortality on logistic regression analysis and despite similar baseline demographic parameters and underlying severity of illness. More than half of the tracheostomized patients survived to hospital discharge, younger age and female gender being the only predictors of survival. In a secondary analysis, earlier spontaneous breathing following the tracheostomy procedure was the only parameter associated with successful decannulation prior to ICU discharge; successfully decannulated patients also stayed in the ICU for a significantly shorter period of time.
So what are strengths of this study? This is thus far the largest single center experience of percutaneous tracheostomies performed in COVID-19 patients. The authors used a very pragmatic approach relating to timing and choice of tracheostomy—surgical versus percutaneous—and based on local protocols that existed prior to the pandemic. The sheer number of patients admitted to ICU and in need of mechanical ventilation led to a decision to very early identification of patients who were unlikely to wean within the subsequent week and after 3 days on the ventilator. Early weaning to spontaneous modes of ventilation, even in the context of modest oxygenation levels and at moderately high positive end-expiratory pressure (PEEP), as well as daily spontaneous breathing trials were used to assess patients’ readiness to successfully wean and extubate. The 3-day cutoff was chosen arbitrarily in order not to subject patients who are on a fast improving trajectory or those that are likely to die, to unnecessary procedures. The only exclusion for early tracheostomy was if patients were deemed to be moribund by the treating clinician or if ongoing need for prone positioning delayed the procedure. Like all other studies published on this topic, results need to be interpreted with caution in view of the retrospective nature of the investigation and the resultant inherent selection bias, especially what indications and decisions in relation to timing of the procedure is concerned. This is particularly true when comparing survival between the groups. Therefore, the authors’ conclusion was that early tracheostomy was safe and did not increase mortality or subjected patients to futile interventions, rather than claim an improvement in survival and despite the statistical effect that favored the intervention.
This is one of only two studies comparing survival between the patients managed with or without tracheostomy. The second largest single-center study to date published by the group from the Queen Elizabeth Hospital in Birmingham (11), United Kingdom, and in 100 patients showed that tracheostomy (75% percutaneous) was independently associated with a 30-day survival benefit and adjusted for severity of illness. Tracheostomy was performed later and at a mean of 14 days after intubation. Thirty-day survival was 85% in the tracheostomy patients who were on average 10 years younger than this trial.
This study differs from other larger size single-center reports (11–16) on various levels: The number of tracheostomies reported is the largest thus far published from a single institution—20% more cases than in the second largest study (11). All tracheostomies were performed percutaneously and on the ICU. Other groups reported a mixture of percutaneous and surgical techniques (11–16); therefore, a large number of patients in other centers required transfer to the operating theatre with all the complex logistics associated with transfer of critically unwell patients. Tracheostomies were performed early and on average on day 6–7 after ICU admission, which is a significantly shorter interval than reported in most other trials, with a duration of on average 9–14 days. Finally, this is the only larger sample size study reporting on complete outcome figures to ICU discharge with 55% of tracheostomized patients surviving to ICU and hospital discharge, more than 70% of who were successfully decannulated on leaving the ICU.
The percentage of patients tracheostomized, of all intubated patients admitted to ICU, was significantly higher than in most other reports at 67%. The main reason for this was the early decision-making, which also meant that a number of patients may have been subjected to unnecessary interventions, could have extubated successfully after 6–7 days, or died regardless. The authors defend this very proactive approach with the perceived high risk associated with failed extubations, a point one may disagree with.
Reassuringly and like in many other studies, there appeared to be no major safety concerns, untoward incidents or deaths resulting directly from the procedure. It is feasible that a number of incidents may have gone unreported or were not documented, as this was not a controlled trial setting.
The authors give brief information regarding the technical details of the procedure. In essence, a standard percutaneous approach was chosen and under fiberoptic guidance. Ventilation was not stopped unlike in other trials, when on opening of the airway ventilation was ceased and tubes clamped (8,11) or modified tracheostomy methods used such as described by Angel et al (12) in a large series from New York. Similarly, in our and others’ experience, COVID-19 patients receiving extracorporeal membrane oxygenation support can be safely tracheotomized early and potentially under apnoeic conditions by increasing ECMO flows; special care must be taken if doing so, to mitigate against the risk of air embolism (14,15). As increasingly reported in other trials, standard PPE precautions were used, procedures performed outside negative pressure environments and without powered respirators and full protective suits. For many practitioners, this coupled with the early timing of the procedure may seem a high risk approach in relation to viral transmission. Yet, the percentage number of practitioners testing positive for COVID-19 who were members of the tracheostomy team was lower than the rate of transmission in the rest of the ICU staff. Similar incident rates with no or a very small number of viral transmissions have been reported by other groups (11–16). Even when tracheostomies are performed very early, patients will have been ill for more than 14 days and without taking incubation periods into consideration. Despite ongoing positive viral RNA on PCR, viable viral load is probably limited and so is the infection risk for the operator.
Sharing of real-life observational data is extremely important during a pandemic and in order to learn from experiences others have gained, especially when regions and countries are affected at different time points. This and other larger sample size studies have shown that our indications and practice of performing tracheostomies in the ICU may not differ greatly in SARS-CoV-2 patients compared with other pathologies. As shown by Rosano et al (10), percutaneous tracheostomy performed as early as after 1 week of intubation is in most instances safe, both for patients as well as caregivers. It is very unlikely that we will ever be able to gain randomized controlled trial evidence of adequate sample size to confirm or disprove the findings gained from observational trials, so this may well be “As Good as It Gets.”
1. Zhou F, Yu T, Du R, et al. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: A retrospective cohort study. Lancet. 2020; 395:1054–1062
2. Sherren PB, Ostermann M, Agarwal S, et al. COVID-19-related organ dysfunction and management strategies on the intensive care unit: A narrative review. Br J Anaesth. 2020 Sep 8. [online ahead of print]
3. Young D, Harrison DA, et al. Effect of early vs late tracheostomy
placement on survival in patients receiving mechanical ventilation: The TracMan randomized trial. JAMA. 2013; 309:2121–2129
4. Szakmany T, Russell P, Wilkes AR, et al. Effect of early tracheostomy
on resource utilization and clinical outcomes in critically ill patients: Meta-analysis of randomized controlled trials. Br J Anaesth. 2015; 114:396–405
5. McGrath BA, Wallace S, Goswamy J. Laryngeal oedema associated with COVID-19 complicating airway management. Anaesthesia. 2020; 75:972
6. Miles BA, Schiff B, Ganly I, et al. Tracheostomy
during the SARS-CoV-2 pandemic: Recommendations from the New York head and neck society. Head Neck. 2020; 42:1282–1290
7. McGrath BA, Brenner MJ, Warrillow SJ, et al. Tracheostomy
in the COVID-19 era: Global and multidisciplinary guidance. Lancet Respir Med. 2020; 8:717–725
8. Takhar A, Walker A, Tricklebank S, et al. Recommendation of a practical guideline for safe tracheostomy
during the COVID-19 pandemic. Eur Arch Otorhinolaryngol. 2020; 277:2173–2184
9. Surkova E, Nikolayevskyy V, Drobniewski F. False-positive COVID-19 results: Hidden problems and costs. Lancet Respir Med. 2020 Sep 29. [online ahead of print]
10. Rosano A, Martinelli E, Fusina F, et al. Early Percutaneous Tracheostomy
in Coronavirus Disease 2019: Association With Hospital Mortality and Factors Associated With Removal of Tracheostomy
Tube at ICU Discharge. A Cohort Study on 121 Patients. Crit Care Med. 2021; 49:261–270
11. Queen Elizabeth Hospital (QEH) Birmingham COVID-19 Airway Team. Tracheostomy
in COVID-19 e safety and 30-day outcomes of the first 100 cases from a single tertiary UK hospital: A prospective observational cohort study. Br J Anaesth. 2020 Aug 28. [online ahead of print]
12. Angel L, Kon ZN, Chang SH, et al. Novel percutaneous tracheostomy
for critically ill patients with COVID-19. Ann Thorac Surg. 2020; 110:1006–1011
13. Picetti E, Fornaciari A, Taccone FS, et al. Safety of bedside surgical tracheostomy
during COVID-19 pandemic: A retrospective observational study. PLoS One. 2020; 15:e0240014
14. Yeung E, Hopkins P, Auzinger G, et al. Challenges of tracheostomy
in Covid-19 patients in a tertiary centre in inner city London. Int J Oral Maxillofac Surg. 2020; 49:1385–1391
15. Tornari C, Surda P, Takhar A, et al. Tracheostomy
, ventilatory wean, and decannulation in COVID-19 patients. Eur Arch Otorhinolaryngol. 2020 Aug 1. [online ahead of print]
16. Chao TN, Harbison SP, Braslow BM, et al. Outcomes after tracheostomy
in COVID-19 patients. Ann Surg. 2020; 272:e181–e186