Now 2 years old, the unprecedented coronavirus disease 2019 (COVID-19) pandemic produced a medical response that was equally unprecedented. Starting with nearly no knowledge about diagnosis, transmission, prevention, treatment, or predictors of outcome, physicians and hospitals were forced to develop real time answers for a new and unfamiliar disease. Particularly, challenging medical aspects of COVID-19 included multisystem involvement, extremely high transmissibility, rapid mutation, and an extraordinarily wide spectrum of clinical presentation from asymptomatic carrier to fatal respiratory failure.
COVID-19 placed a particularly heavy strain on critical care services. In addition to renal and coagulation abnormalities,1 patients with COVID-19 were characteristically and often severely hypoxemic. Respiratory failure would persist for days to weeks, and those who progressed to intubation were often exhausted and prone to cardiorespiratory collapse. In many cases, intubated patients would require aggressive ventilatory support for weeks and progress further to extracorporeal membrane oxygenation (ECMO) support. The combination of rapidly growing caseloads during COVID-19 “waves,” infected providers unable to work, and prolonged courses of respiratory failure in patients with advanced disease severely strained intensive care unit (ICU) resources even at large medical centers.2
As critical care and airway management experts, anesthesiologist intensivists found themselves at the vanguard of many hospital responses to waves of critically ill patients with COVID-induced respiratory failure. Reductions in surgical activity, a need for high-risk airway management, and expertise with complex respiratory failure made anesthesiologist intensivists an ideal option for expanding ICU services. Multiple publications, many in Anesthesia & Analgesia, testify to robust anesthesiologist involvement in COVID-related critical care.3–5
|1. A hospital-wide critical care response to a new disease|
|2. Creating ICU space out of thin air|
|3. From easy to very, very hard: intubating the COVID patient|
|4. A new therapy for an even newer disease: ECMO for COVID-associated respiratory failure|
|5. Staying sane in unprecedented times|
|Pathogenic evolution can be rapid and unpredictable|
|Clinical guidance and best practices may evolve more slowly than empiric adaptation|
|Seek out opportunities for collaboration and discovery|
|Anticipate and manage uncertainty|
|Be mindful of provider stress|
Although the origins of COVID-19 remain unclear, many public health officials worry that the likelihood of a “next pandemic” is increasing.6 To imagine how the critical care response to the next pandemic might be informed by experience with COVID-19, members of the Society of Critical Care Anesthesiologists (SOCCA), each involved in the COVID-19 response in their institutions, were asked to describe what they learned and how that knowledge might inform their response to the next pandemic. What follows are 5 diverse perspectives (Table 1) that describe the critical care response to COVID-19 in terms of organization, space, clinical care, and wellness. In each, we asked the authors to describe not just their response to the COVID-19 pandemic, but what they learned that might inform the “next next wave” (Table 2).
A HOSPITAL-WIDE CRITICAL CARE RESPONSE TO A NEW DISEASE
Dr Mark Nunnally is director of adult critical care services at NYU Langone Health in New York City and participated in their critical care response to the COVID-19 pandemic at NYU. Here he describes the elements of the hospital-level response to COVID-19 and what lessons to take away for the next pandemic.
For health care systems, the first casualty of the COVID-19 pandemic was time. After news about the novel coronavirus began to reach hospitals in New York City in late February 2020, daily admissions grew rapidly over the next several weeks, and reached an unprecedented level by the middle of March. New York hospitals (including ours) speculated, made plans and implemented them, only to see the volume and severity of cases quickly exceed their most pessimistic projections. Under such circumstances, leadership and communications had to (and did) become nimbler.
To organize and adapt, we found that decisions were best made directly through executive leaders. Frequent, regular discussions tackled new issues, and scheduled conference calls provided a forum for rapidly addressing new concerns. Fortunately, we already had an emergency management team in place due to Hurricane Sandy in 2012, and when COVID-19 began, we used that team to run these conferences and manage equipment, space, organizational skill and technologies. Because COVID was an unknown entity, these conferences also provided an essential link with clinical teams who were adapting and innovating on the wards in real time. Anticoagulation protocols and practice modifications such as earlier tracheostomy were examples of locally derived practice strategies.
Once identified, teams composed of service directors, unit directors, and managers then disseminated those strategies to other units, and in turn brought concerns and observations back to health system leadership. The communications infrastructure was populated by clinicians to allow rapid movement of information and evolution in decision making. Equally important, the system generated a sense of institutional trust, without which it would have been difficult to be so nimble. When bedside clinicians felt that their needs were being heard and leadership saw their decisions making an impact, this trust grew.
Maintaining a managerial presence in the ICU was an invaluable element of ongoing effective communication. Senior management conducted daily walk rounds, even to the patient bedside in isolation COVID-19 units. Such rounds facilitated information gathering and dissemination, and let clinicians know that the hospital was paying attention. Regular e-mails shared essential details, thoughts and even speculative musings, creating a sense of community among the staff. The sharing of uncertainties elicited a positive response, as did “I don’t know.”
Many key aspects of the pandemic surge response were logistical—getting the right combinations of people to do the work, the space for them to do it, and the materials they needed to get the job done. Complex and unanticipated supply issues, such as lack of sufficient oxygen manifold capacity and transportation issues when sourcing new supplies, were common. Paradoxically, one advantage of the intensity and size of that first COVID-19 wave was that displaced most other hospital services. Organizational adaptations could then focus almost exclusively on the COVID-19 pandemic. At the peak, it really was “All COVID, all the time.”
Ultimately, successful elements of our COVID-19 response—and those we would start with in a “next pandemic”—included making executive decisions more efficient, bridging with bedside providers to identify and disseminate local innovations, escalating concerns and providing feedback, enhancing communications through presence and transparency, paying close attention to logistical concerns, and emphasizing coordinated, multi-specialty solutions to emerging problems.
CREATING ICU SPACE OUT OF THIN AIR
A striking aspect of the COVID-19 pandemic in New York City was an acute, massive influx of large numbers of critically ill patients beginning in March 2020. Starting with a baseline capacity of 117 ICU beds, New York Presbyterian/Columbia University Irving Medical Center (NYP-Columbia) reached a peak census of 257 COVID-19 patients requiring ICU care during the Spring 2020 wave. Drs Jonathan Hastie and Aaron Mittel were part of a novel response to this challenge at New York Presbyterian/Columbia University Medical Center, repurposing operating rooms (ORs) as ICUs.7 Here, they describe what they learned from managing an unprecedented transition, which could inform a response to a future pandemic.
In March 2020, NYP-Columbia was nearly overwhelmed by a massive surge of critically ill patients with COVID-associated respiratory failure. To create the additional ICU space and capacity to accommodate that surge, we converted unused ORs into the “Operating Room ICU” (ORICU), an 82-bed ICU that cared exclusively for COVID-19 patients. At its peak, the ORICU managed 79 patients and represented 34% of the ICU census at NYP/Columbia.7
Our experience with transforming ORs into ICUs during the COVID-19 pandemic taught us to focus on 3 main themes: a clear vision of how care will be delivered, an intentional approach to managing the change, and the necessity of considering the emotional, physical, and psychological strain on health care workers (HCWs) delivering care in the ORICU.
Defining the Capability of the ICU Space
During initial plans for what was an unprecedented reconfiguring, we found that defining the nature of care that will take place in that space was crucial. To clarify how needed care would be delivered, we conducted pre-use walkthroughs and training exercises for high-risk interventions such as prone positioning for mechanically ventilated patients before actual deployment. Operationally, housing multiple patients per OR necessitated the near-constant presence of a HCW in each OR. Because ORs, unlike traditional ICUs, are closed spaces, we used existing windows in each OR to relay information from inside to providers outside each room. Similarly, vital signs were prominently displayed on bedside machines and relayed to terminals grouped outside each OR. This approach permitted high-fidelity communication and monitoring, while limiting HCW exposure.
Although OR and ICU environments both accommodated high acuity care, key differences between the physical capabilities of each space required careful prereview. In the setting of COVID-19, we addressed 3 questions:
- Which patients will be cared for in this space? Clearly defining this patient population was key to facilitating triage and balancing physiologic, ethical, and social concerns. We expected that COVID-19 patients would need mechanical ventilation, sedation, paralysis, and prone positioning. We also did not group patients by gender to allow us to use valuable OR/ICU space as efficiently as possible.
- What physiologic support will be needed? Defining what clinical care can safely be delivered was vital to safe ORICU operation. We recognized that mechanical circulatory support was likely best delivered in a standard ICU setting. Unanticipated limitations to clinical care also arose during operation. A lack of plumbing to drain machine effluent prevented us from providing renal replacement therapy in the ORICU until facility engineers could modify existing drainage systems.
- Which isolation procedures would be needed? To address caregiver infection, we had to modify our ORs. Because ORs are positive pressure spaces where air flows are normally directed from inside to outside the room, their use for COVID-19 care raised the possibility that airborne viruses could contaminate the hallways. We solved this problem by converting each OR into a negative pressure room.
Only after these questions have preliminary, workable answers should a project as large as reconfiguring OR space to care for ICU patients be undertaken.
Intentionally Managing the Transformation
The second lesson is that transforming space requires actively managing the change process to avoid inadvertent misunderstandings. For example, in planning to adapt anesthesia machines for use as ICU ventilators, we realized that respiratory therapists would not be familiar with anesthesia ventilators. Regular discussions included members from involved disciplines and departments empowered to make decisions and perform plans. A set agenda, specific follow up on projects, and clear communication facilitated planning. In particular, we found that having leaders seek out dissenting voices helped identify potential problems.
In the case of the ORICU, attendance by senior hospital leadership demonstrated support for the project and helped secure needed resources. In addition, we sought out patient and family perspectives and input to answer questions such as: How will family members receive updates? What situations, not present in a typical ICU, might generate patient distress?
When adapting care to novel areas, starting small helped us to identify problems. We initially admitted only a few patients and addressed the needs and gaps identified from those first few patients. Additionally, we targeted quality benchmarks, including mortality and rate of caregiver infection, to assess safety during this initial phase. We also maintained a readily accessible pathway for clinical teams to easily report incidents.
Anticipating HCW Strain
Repurposing space under pandemic conditions was (and is) extremely disruptive to those who deliver care in that space. At our peak, we cared for up to 6 patients in a single OR, a starkly abnormal circumstance in which even basic care delivery was difficult. Recognizing the considerable impact on provider wellness, our hospital leaders coordinated with hospital support services to acknowledge the crisis, validate caregiver stress, and orient the team around a shared goal and values.
We hope that in the next pandemic, repurposing ORs into ICUs will not be needed. But if so, shaping a clear vision of how care will be delivered, creating an intentional process to managing the change, anticipating, detecting, and being ready to address unanticipated issues or events, and prioritizing the considerable strain that HCWs experience under such circumstances, are major priorities.
FROM EASY TO VERY, VERY HARD: INTUBATING THE COVID PATIENT
The frequent need for invasive airway management pushed critical care anesthesiologists to the forefront of COVID response efforts at many hospitals. An estimated 50% of SOCCA members report working in an all-COVID ICU during the pandemic. While early airway management protocols focused on minimizing HCW infection, clinicians quickly recognized that many patients could be maintained on high-flow oxygen therapy. Those who failed noninvasive support, however, represented a particularly high-risk group for airway management. Dr Allison Dalton describes her experience as part of an anesthesia intensivist group at the University of Chicago that pivoted to medical ICU care during the pandemic to address one of the largest COVID surges in Illinois.
Few COVID-essential therapies have evolved as much as airway management. Although COVID-19 included other organ systems besides the lung, the predominant clinical course in early 2020 involved cough, dyspnea, and a surprisingly severe degree of hypoxemia. A key unknown during that initial wave of the COVID-19 pandemic was how easily the virus could be transmitted from patient to provider. Early airway strategies thus focused on minimalizing the number of aerosolizing procedures and included limiting the number of staff present at intubation, utilizing rapid sequence induction/intubation techniques to minimize aerosolization, first-look videolaryngoscopy to avoid potential exposure to exhaled virus, and avoiding noninvasive respiratory support devices with the potential for aerosolization of respiratory secretions such as high-flow nasal cannula (HFNC).8,9 As a result, early in the pandemic, most patients were relatively hemodynamically stable and not on maximal oxygen therapy at the time of intubation.
As the number of patients affected by COVID-19 increased, 3 clinical observations drove a pivot to a new strategy. The first was recognition that the duration of intubation was often long in patients with COVID-19 lung disease, and that ICU capacity would be rapidly depleted if early intubation practices continued. The second was that patients could often be maintained on noninvasive respiratory support and recover without intubation and mechanical ventilation. The third was that with appropriate personal protective equipment, the risk of contracting COVID-19 during airway management was low.
Consequently, many clinicians shifted to a strategy where noninvasive modalities, such as HFNC, noninvasive ventilation (bilevel positive airway pressure [BiPAP], helmet), awake prone positioning, and diuresis, were used to “prevent the vent.”10 However, such strategies were limited by the high oxygen flow rates needed for noninvasive ventilation and discomfort with helmet devices. In addition, we realized that patients who failed noninvasive ventilation were often exhausted, intravascularly volume depleted, and prone to severe peri-intubation hemodynamic instability. Due to the severe gas exchange abnormalities characteristic of COVID-19, even brief periods of apnea during airway management led to severe, persistent oxygen desaturation, hypotension, and even cardiac arrest.11,12 In addition, partly due to the high risk of hemodynamic instability and infection control constraints, the risk of failed intubation was similar to that reported for prehospital trauma.13 In light of the heightened risk of hemodynamic instability, hypoxia, and failed intubation with hypnotic techniques, we again pivoted to topicalized awake intubations toward the end of the first wave—despite the perceived infectious risk to avoid patient harm.
If early intubation represented inefficient use of ICU resources and was unnecessary in many patients, and if intubation only after failing noninvasive ventilation was associated with severe cardiopulmonary deterioration, then was there a middle ground? A yet unresolved issue with COVID-19 is the optimal timing for intubation and mechanical ventilation. Decision making regarding the timing of intubation was (and remains) challenging in part because patients with COVID-19 pneumonia may exhibit different phenotypes.14
One form of COVID-19 pneumonia was characterized by low elastance, low ventilation-to-perfusion ratio, low lung weight and recruitability, and a surprising lack of dyspnea. In contrast, another clinical scenario was characterized by high elastance, high right-to-left shunt, high lung weight, and high lung recruitability. Although we empirically found that patients with the former phenotype could often be treated with noninvasive strategies, identifying risk factors for failed noninvasive therapy and defining the optimum timing of invasive mechanical ventilation remains an important research goal. A better understanding the pathophysiology of COVID-induced respiratory failure and clinical risk factors for progression to intubation and mechanical ventilation will hopefully help intensivists avoid high-risk intubations in patients who fail noninvasive support.
For the next pandemic, retaining the flexibility to adjust to clinical circumstances was the primary lesson for us at the University of Chicago. Within a few weeks of the first COVID-19 wave, elaborate protocols detailing the thresholds that should trigger intubation were retooled to target a completely different noninvasive approach. Staying flexible, listening to frontline caregiver experiences, being open to different approaches, and sharing experiences with other institutions were all important elements. Although we have not yet identified the ideal timing and intubation technique for COVID-19 patients, our adaptability and willingness to adjust allowed us to find a balance between keeping our providers safe, prioritizing scarce resources, and providing safe care and represents an important lesson for the next next wave.
A NEW THERAPY FOR AN EVEN NEWER DISEASE: ECMO FOR COVID-ASSOCIATED RESPIRATORY FAILURE
The severity and prolonged duration of respiratory failure with COVID-19 led many to wonder whether ECMO could be used to support patients until their gas exchange improved. Controversial even in conventional acute respiratory distress syndrome (ARDS), the use of a cutting-edge therapy in a new disease with an unknown pathophysiology and no effective treatment was a big step into the beyond. Dr Craig Jabaley is medical director of a regional ECMO referral center for severe respiratory failure (including COVID) at Emory University Hospital in Atlanta, Georgia. Here, he describes the deployment of ECMO in the treatment of COVID-19.
Although definitive evidence supporting the use of venovenous ECMO for hypoxemic respiratory failure was lacking in early 2020, it was a logical treatment for a disease whose primary presentation was severe hypoxemic respiratory failure. However, early reports suggested remarkably poor clinical outcomes, with mortality rates approaching 90%,15 a stark contrast to contemporary ECMO mortality rates of 35% in patients with non-COVID ARDS.16 These initial observations initially called into question the utility of ECMO in this COVID-19 patient population. Over the course of the pandemic, however, pooling of experiences with delivering ECMO for COVID-19 patients allowed intensivists to address management challenges common to all ECMO patients, develop strategies specific to this population, and address the challenge of rapidly scaling ECMO capacity in the face of resource strain. This global journey demonstrated that with rapid dissemination of empirically derived care strategies, ECMO could be adapted to COVID-19 with short-term and longer-term ECMO outcomes comparable to other causes of respiratory failure.17
COVID-19 added an additional layer of complexity to the already challenging task of safely delivering ECMO in conventional respiratory failure. For example, managing the balance between bleeding and coagulation—an existing dilemma in non-COVID-19 patients—was made even more challenging by coagulation abnormalities caused by COVID-19.18 Severe COVID-19 caused both fibrinolysis and microangiopathic pulmonary thrombosis,19 requiring clinicians with limited knowledge of mechanism to adjust anticoagulation to hyper and hypocoagulable states. Despite early reports suggesting a high incidence of thrombotic complications,20 hypercoagulable complications were ultimately not increased in the Extracorporeal Life Support Organization (ELSO) registry with COVID-19 patients,17 suggesting that clinicians adapted rapidly and effectively to additional complexity associated with COVID-19.
COVID-19 also complicated other routine aspects of critical care for ECMO patients. Empirically, intensivists learned very quickly that patients with COVID-19 had unusually high sedative requirements.21 Coughing, patient-ventilator dyssynchrony, and agitation were surprisingly common, increasing metabolic oxygen demand and destabilizing extracorporeal circuit flows. Because the median duration of ECMO support for COVID-19 respiratory failure was almost 2 weeks and its use was complicated by infection in over 30% of patients,17 intensivists learned to adjust, delivering full ECMO support early in the course when patients were metabolically unstable and respiratory failure was severe, and then targeting cannulation strategies that facilitated awakening and mobility during convalescence. Because long-term functional impairment after ECMO for COVID-19 is common, a focus on optimizing function post-ECMO became an important aspect of limiting the impact of prolonged critical illness.22
As a resource intensive treatment modality, the use of ECMO for COVID-19 respiratory failure also raised questions as to its cost-effectiveness relative to conventional management.23 Directing resources toward ECMO during a pandemic wave where scarcity limited the resources available to other patients, both directly (circuits, beds) and indirectly (staff, services), added additional complexity to the decision to initiate ECMO support. Compounding this complexity was wide variability both in the duration of time considered reasonable to continue ECMO support and in the amount of time patients required for recovery.24 In contrast to non-COVID ARDS where the utility of continued support is often assessed at 30 days, clinical experience with ECMO for COVID-19 suggested that patients could recover or be successfully transplanted even after prolonged ECMO support.25
Recognizing that lack of access to ECMO was associated with non-survival,26 expanding ECMO capacity to meet surge demands became an important priority in many centers. However, pandemic-related staffing, space, and equipment constraints all prevented substantial expansion. In subsequent waves and in light of resource/outcome constraints, many centers tightened their selection criteria, aligning them with recently published ELSO guidelines.27 Studies suggest that mortality during ECMO support has increased over the course of the pandemic, which should prompt critical examination of modifiable contributors. While we cannot directly combat the impacts of viral evolution, such as severe illness from the Delta variant (B.1.617.2), we can modify factors like strained staffing, expansion of ECMO within less experienced centers, and the potentially deleterious impact of lengthy noninvasive ventilation.
Should patients in the next pandemic require ECMO support, several lessons learned during the response to COVID-19 will likely make implementation smoother. First, mortality with ECMO for COVID-19 decreased with experience, suggesting that clinicians need not be deterred by high mortality early on. Second, intensivists should take heart in how rapidly a cumulative, shared worldwide experience drove clinically relevant treatment paradigms. That a cutting-edge technology like ECMO could be adapted to an unknown disease is itself testament to the flexibility of ECMO support. Third, experience with resource deployment and access to care will likely improve access to therapies like ECMO in the future. Finally, over time, ECMO itself will become safer, more accessible, and easy to deliver, facilitating its use in the next pandemic.
STAYING SANE IN UNPRECEDENTED TIMES
Already an issue pre-COVID-19, provider wellness was further adversely affected by the COVID-19 pandemic. In addition to large waves of severely ill patients, intensivists combating COVID-19 faced a disease with unknown parameters, delivered care in atypical environments, and risked infection themselves. Dr Shahla Siddiqui is an anesthesiologist intensivist at the Beth Israel Deaconess Medical Center in Boston, Massachusetts, who has studied wellness among critical care anesthesiologists. Here she describes lessons learned from COVID-19 that might help preserve wellness during the next pandemic.
Prepandemic, no one could have imagined the enormity of the physical, emotional, and moral stress caused by a novel disease like COVID-19. The impact of COVID-19 on critical care providers was particularly large, because case volumes were high, risk factors for severe disease were unknown, transmissibility to caregivers was uncertain, the mortality rate for intubated patients was high, isolation procedures were extensive, and no effective therapy was available.
The many recurring waves of the pandemic and the cumulative death toll also led to an increase in burnout among anesthesia-trained intensivists. Even before the COVID-19 pandemic, burnout was an issue among anesthesiologists in general28 and critical care anesthesiologists in particular.29 Post-COVID-19, surveys identified markedly high levels of moral distress30 and startlingly common thoughts of depression, anxiety, posttraumatic stress, and even self-harm.31 Although anesthesiologists caring for COVID-19 patients felt they were contributing more to critical care than before the pandemic, they also described an increase in burnout-related symptoms like muscle tension, fatigue, and sleeplessness. These symptoms were more common in women, possibly due to the added burden of home-related responsibilities.32
The consequences of physician burnout due to COVID-19 care have not yet finished reverberating through the postpandemic medical system. Early retirement or shifts to less stressful jobs increase the burden of caring for subsequent COVID-19 waves, or another pandemic, on those physicians who remain.33 Evidence that medical students consider the risk of burnout in choosing a residency specialty34 suggests that if improperly managed, burnout may affect the future supply of physician specialists. In particular, anesthesia intensivists are on the front lines of COVID-19 care, and medical students concerned about anesthesia critical care–related burnout may choose other specialties, placing pressure on current intensivists.
As the above narratives attest, physicians and hospital administrators were aware of high rates of burnout and posttraumatic stress disorder (PTSD) due to COVID-19 care. Unfortunately, knowledge regarding how best to optimize wellness among physicians is incomplete.35 Health care organizations countered early in the first COVID-19 surge by providing resources and perks.36 Existing literature also suggested that one factor that may potentially mitigate burnout is “care and humanistic” overtures—in the form of a kinder work environment, improved resources and an increased presence from leadership and midmanagerial levels.37
Many studies of staff wellness during the COVID-19 pandemic support the concept that peer and managerial support makes frontline personnel feel valued and helps to reduce the stress of pandemic critical care.38 Favorable media coverage of health care providers during COVID-19 was associated with increased personal fulfillment among anesthesia intensivists, which suggests that attention to expressed and perceived respect, as an aspect of wellness, is an underappreciated aspect of physician well-being. For the next pandemic, wellness should be as much a focus as logistics and physical space, and greater attention to societal, community, and organizational support may reduce the mental toll of pandemic care.