Rebuttal From Drs Pandian, Murgu, and Lamb.

Brenner and colleagues identified several reasons why tracheostomy should be done before 14 days of mechanical ventilation in patients with coronavirus disease 2019 (COVID-19): 1) reduced prevalence of post-ICU syndrome (PICS); 2) earlier participation in rehabilitation; 3) decreased risk of tracheomalacia and tracheal stenosis, ventilator-associated pneumonia, and mortality; 4) lack of high rates of infection in health care workers performing tracheostomy; and 5) increased ICU bed capacity.

These reasons for performing early tracheostomy must be viewed with caution. Although prolonged mechanical ventilation is a known risk factor for PICS, no data suggest that tracheostomy reduces this risk. We expect that PICS will be prevalent in COVID-19 ICU patients who survive, but whether the timing of tracheostomy affects this risk is not yet known. The data supporting an earlier return to walking, talking, and eating and the debatable evidence favoring early tracheostomy in terms of ventilator-associated pneumonia and mortality portrayed in Figure 2 come from studies in non-COVID-19 patients.

The data to support the claim that tracheostomy decreases the risk of tracheomalacia and tracheal stenosis are lacking. We agree that earlier extubation may lower the risk of these complications, but a higher prevalence of stenosis is reported with tracheostomy than with intubation alone. In fact, the stomal, suprastomal, cuff- or tracheostomy tube tip-related strictures or granulation tissue are well-documented complications, and no clear evidence indicates that early tracheostomy reduces the prevalence of these complications. Although we agree that high infection rates in health care workers performing tracheostomy have not been reported, the existence of asymptomatic carriers should not be ignored, and is not without consequence in the health care environment. To date, no study has reliably performed timely COVID testing of the clinicians involved in the tracheostomy procedure.

Some argue that early tracheostomy may improve ICU bed capacity because of reports of a shorter length of stay (40 vs 49 days); however, the timing of the tracheostomy has not been shown to decrease the duration of mechanical ventilation or the time to decannulation. In our opinion, transferring a patient out of the ICU to a long-term assisted care unit days earlier is not necessarily a patient-centered outcome. We also recognize the preexisting limited number of long-term care or subacute rehabilitation facilities (398 in United States) that can accept patients with a tracheostomy.

A systematic review and meta-analysis of 18 studies exploring 3,234 COVID-19 patients showed that only 5.2% of tracheostomies were performed within 7 days (early), and 21.2% were performed between days 8 and 13, whereas most (71.5%) of the tracheostomies were performed 14 days or later post-intubation. The meta-analysis did not reveal the benefit of early tracheostomy in terms of duration of mechanical ventilation or time to decannulation, nor was late tracheostomy associated with increased mortality. ,

We believe that the existing published data suggest that the historical distinction of early vs late tracheostomy may not be applicable in patients with COVID-19. Pre-COVID-19 studies of tracheostomy were not necessarily related to critical illness after an infectious cause but rather to an exacerbation or progression of a pre-existing condition. Even in New York City during the first few months of the pandemic, the mean time from intubation to tracheostomy was 12.23 days (SD, 6.8), despite a deliberate effort to perform an early tracheostomy. These data suggest that despite the potential need to increase ICU bed capacity in the United States epicenter of the COVID-19 pandemic, on average, clinicians proceeded with tracheostomy after day 10, likely because they accounted for the virology of SARS-CoV2 as well as the context of the natural history of the disease. The decision to proceed with a tracheostomy should be multifaceted and multidisciplinary. It should account not only for one variable of days since intubation but also for the estimated duration of required mechanical ventilatory support, peak infectivity period, disease severity markers, use of approved anti-COVID therapies, and overall goals of care. We suggest that future algorithms should avoid the traditional dichotomy of early vs late and consider a three-tiered approach of early (<7 days), intermediate (8-13), and late (>14 days) tracheostomy. Such an algorithm could account for these factors and may help clinicians individualize the approach to tracheostomy in COVID-19 patients. We believe that in the past year, we have learned that we do not need to wait for day 21 to perform a tracheostomy in a patient with COVID-19. However, published evidence and understanding of this disease do not justify an early (<7 days) tracheostomy if we consider safety of the health care team and patient-centered outcomes.