The Repurposed Use of Anesthesia Machines to Ventilate Critically Ill Patients With Coronavirus Disease 2019 (COVID-19)

Maurizio Bottiroli; Angelo Calini; Riccardo Pinciroli; Ariel Mueller; Antonio Siragusa; Carlo Anelli; Richard D. Urman; Ala Nozari; Lorenzo Berra; Michele Mondino; Roberto Fumagalli


BMC Anesthesiol. 2021;21(155) 

In This Article


This is the first analysis of a relatively large registry investigating mortality in patients who received care with Anesthesia Machines versus standard ICU ventilators for COVID-19 respiratory failure during a peak surge. Our results indicate that during the emergency response to the initial peak of COVID-19, the care of critically ill patients with repurposed Anesthesia Machines was associated with an increased rate of complications and mortality.

Italy was the first country outside of China to suffer a major outbreak of SARS-CoV-2 infection and the registry population provided unique advantages for studying the impact of Anesthesia Machines use in the care of COVID-19 patients compared to standard ICU-VENT use. The population we described in this study had similar characteristics to that of other published case series.[13–17] The 70% mortality for patients in the AM group however is remarkably high. The overall 60-day mortality though of the analyzed cohort (43.8%) is similar to what has recently been reported by the COVID-19 Lombardy ICU Network in the largest currently available outcome study of Italian cases (48.7%).[14] While greatly varying across the multiple available reports, ICU mortality from COVID-19 is primarily driven by the development of ARDS, with 50% mortality among patients with COVID-19 ARDS generally considered an accepted estimate.[18] Our overall mortality findings are, therefore, in agreement with the currently available literature, supporting the quality and external validity of our data.

Given the dramatic increase in mortality that we observed in patients whose care involved the use of Anesthesia Machine, we aimed to quantify factors associated with their care contributing to lethality. We wish to be explicit in stating that our registry analysis does not allow us to conclude that the Anesthesia Machine ventilator itself is the exclusive culprit. Instead, we believe that our study demonstrates that the clinical care scenarios associated with using Anesthesia Machines are linked to increased mortality. There are several considerations regarding possible changes in clinical care and unique Anesthesia Machine-related challenges that should be discussed.

First, the correct setup of audible alarms on Anesthesia Machines and the ability to respond with a prompt corrective action might prove challenging for any operator, particularly when clinicians are trying to limit proximity to patients and must don and doff personal protective equipment.[5,8] We believe that several non-quantifiable factors related to Anesthesia Machines not being a normal standard of ICU care – even for clinicians comfortable with their operation in the OR setting – could have led to a higher degree of mortality.

Second, clogging of HMEF and filters due to excess moisture or secretion burden is a major problem in patients receiving prolonged ventilation on an AM without a heat source and active humidification, particularly when low fresh gas flow is used.[19,20] HMEs are a passive form of humidification. The device stores heat and moisture from the patient's own exhaled gas which is released during inhalation of fresh gas, which would otherwise be dry and at ambient temperature.[21,22] In our study, frequent HME replacements were required in the AM group to prevent occlusion. The use of a higher fresh gas flow rates reduced this complication and is currently recommended by the APSF/ASA guidelines.[8]

Third, COVID-19 patients often show tenacious and abundant tracheal secretions, whose inspissation might lead to an even higher risk of ETT occlusion. Given the lower temperature in the Anesthesia Machine circuit (without a dedicated heating system or active humidification), we believe Anesthesia Machines could make this risk even higher. In COVID-19 patients ventilated with repurposed Anesthesia Machines, Panchami, KR, et al. reported a 29% incidence of critical airway obstruction requiring emergency Fiber Optic Bronchoscopy or tube exchange.[23] In our study, subtotal tube obstruction due to mucus accumulation was also a common occurrence with Anesthesia Machines. In one case, an exceptionally large mucus plug caused a sudden complete airway obstruction at the level of the carina leading to hypoxia and cardiac arrest. It reasonable to presume that inadequate heating and humidification could also have led to increased secretion burden and decreased secretion clearance in more distal airways.

Four, we estimate that each patient on an Anesthesia Machine had to be disconnected roughly twice per day on average, to either change filters, or perform startup self-tests. On these occasions, we had no other choice than to ventilate the patient with a manual resuscitator. Disconnections from the mechanical ventilator in ARDS might result in loss of PEEP and lung collapse and should be avoided at all costs.[24,25] The use of manual bag ventilation might lead to hyperventilation with excessive rate, pressure, and tidal volume, all critical determinants of VILI.[26,27]

Finally, one must theoretically consider that ventilator-induced lung injury of increased severity is also a potential explanation for the decreased survival in patients in the AM group. The accumulation of excess condensation in the circuit of AMs often hindered the accuracy of flow sensors and increased resistance, leading to inconsistently delivered tidal volumes. The importance of an accurately set tidal volume within a lung-protective ventilatory strategy is a mainstay of ARDS treatment.[28,29]

All the points discussed above might also explain themselves a lower amount of Ventilator free-days observed in the AM-group, however we believe that it mainly reflect a higher mortality rate in this group.

Our registry demonstrated a significantly increased use of volatile anesthetic in patients receiving Anesthesia Machines. We believe the OR ICU in this study represented the safest possible setting for volatile anesthetic use given the presence of experienced personnel, proper monitoring systems and adequate machines. For these reasons we largely utilized Sevoflurane as the main sedative drug the critically ill phase of our patients undergoing mechanical ventilation with the AMs. Therefore, the AM group cohort had significantly more volatile anesthetic use compared to the ICU-VENT group. We do not have a high suspicion that the use of halogenates could explain the difference in outcome between the two groups for the following reasons. The use of sevoflurane as a first-line sedative drug in ARDS patients has shown shorter awakening and extubation times compared to IV sedatives.[30] Additionally, Jabaudon et al. observed that when Sevoflurane was compared with midazolam in ARDS patients, oxygenation improved, and alveolar/systemic inflammation and lung epithelial injury were reduced.[31] For these reasons some authors have even suggested that inhaled volatile agents may provide important pulmonary benefits for COVID-19 patients with ARDS.[32,33] We were more concerned that there could be a potential connection between lower blood pressures and inhaled anesthetic use. However, an analysis of blood pressures amongst groups during the first week of ventilation did not reveal concerning trends (Figure 1 in Supplemental Digital Content 2).

The allocation to repurposed AM was not the only factor associate with higher risk of 60-day mortality. Higher age, history of diabetes mellitus and a severer organ damage (measured by creatinine and bilirubin) at ICU admission are also predictors for worst outcome. These findings are in agreement with currently available literature on critical COVID-19.[13,14,17]

Although our initial experience using Anesthesia Machines in the COVID-19 pandemic saw an increased incidence of mortality, we do hope that there were learned lessons that we can share with other clinicians currently experiencing COVID-19 surges. In the event that Anesthesia Machines are required to keep hospital capacity afloat, a summary of the issues we encountered using Anesthesia Machines and related proposed solutions is provided in Table 5.


Our study presents several limitations. Our analysis sought to limit confounding factors and to study the use of Anesthesia Machines as the only difference between groups. However, our sample size is relatively small. It is possible because of the small group size there are differences that persist between groups that were not identified as statistically significant in our analysis, but may be clinically meaningful. As in any retrospective study there is the possibility of residual confounding or bias in our interpretation. It should be noted however that patients who received AMs did so because of bed availability, not patient acuity or other factors, thus these are not likely biasing our results. Detailed data on bed assignments are not available. Our study took place during the very early stages of the pandemic. At the time, we witnessed a shifting emphasis on using certain drugs (e.g., antivirals, hydroxychloroquine, or immunomodulators). Their use in our cohort has been fragmented and based on institutional indications and local availability, rather than supporting evidence. Globally, a better understanding of the disease became available in the summer months of 2020. Our group gained growing experience in the management of COVID-19 patients as time went by, raising the possibility that the results were influenced by secular trends. Finally given that the nature of the study is an analysis of a registry, we are not able to draw causal interpretations to our results. Rather, our results are suggestive of a pattern increased mortality associated with the clinical care of COVID-19 patients whose management involved Anesthesia Machines.