DISCUSSION

Surge capacity in mass critical care is the ICU’s ability to rapidly expand care to as many patients as possible in a disaster. An important component of surge capacity is the staff.^{Reference Seda and Parrish6} A contagious epidemic, such as COVID-19, produces an ongoing stream of new patients that may go on indefinitely. In this setting, the number of critically ill patients rapidly overwhelms the capacity to care for them and may require any individual hospital to double or triple its ICU capacity.^{Reference Parker7} Furthermore, the demands on the ICU are prolonged because it is the end destination for critically ill patients with hospital stays that last for weeks.^{Reference Hossain, Ghazipura and Dichter4} In addition, health-care workers in the ICU are at risk of acquiring the infection and potentially becoming critically ill themselves.^{Reference Parker7} Consequently, surge planning needs to be able to account for a prolonged duration of ICU functioning at a given surge capacity.^{Reference Hossain, Ghazipura and Dichter4}

In our study, the ICUs increased their bed capacity by an average of 160%, with an increase of 400% in some centers. Medical staff surge capacity depended on both critical care and noncritical care physicians. ICM specialists recruited to increase the staff during the pandemic were either former residents from the departments, physicians who were previously hired to take calls and/or were working as first responders, or retired physicians. In some cases, senior residents in their last year of training assumed the role of specialists. It was more difficult to hire intensivists in smaller units because of the competition for them with bigger hospitals, and their reduced residency positions.

The critical care staff was extended with noncritical care physicians working as “force multipliers.” A 2-tiered staffing system was established to provide care to more critically ill patients, with critical care staff (intensivists and critical care nurses) supervising noncritical care staff.^{Reference Seda and Parrish6-Reference Wax8} There was an average increase in the number of physicians present inside the ICUs of nearly 90%, including non-ICM specialists. This allowed for an increase of beds managed by each intensivist (from 1 to 2 patients to 3 to 4 patients per ICM specialist). Patients were also managed by pneumologists using noninvasive mechanical ventilation in intermediate care units.

Each center organized their shifts. One of our centers organized 12-h work shifts with 2 intensivists and 1 anesthesiologist during the night shifts, and the rest of the team during the day shifts. Each physician did 3-d shifts and a 1-night shift every 5 d, with the remaining time of the week to rest. Another center increased the number of physicians on call (from 3 to 4 physicians) and, during the worse moments of the pandemic, added 2 physicians supporting the call team until 8 pm and 10 pm, respectively. Some centers were able to maintain their usual 8-8 shifts due to the support of the pediatric intensivists and the anesthesiologists. Others combined work shifts of 8 and 12 h, depending on the number of patients inside the ICU.

Most of the non-ICM specialist help was provided by anesthesiologists, followed by pediatricians and cardiologists. A possible explanation for this is that these 3 specialties manage their own ICUs (postoperative, pediatric-neonatology ICUs, and coronary ICUs, respectively). Furthermore, anesthesia specialists are ICU specialists in some centers. Generally, anesthesia either cooperated with ICM specialists in the care of COVID-19 critical patients inside the ICU or handled COVID-19 patients inside their own units independently. Another likely reason for the presence of pediatricians in the adult ICUs is that SARS-CoV-2 affects children less commonly and severely in comparison to adults.^{Reference Liguoro, Pilotto and Bonanni9} Pediatricians, in particular pediatric intensivists, were not as likely to be as overwhelmed as adult intensivists during the pandemic. In fact, in our survey, 1 center described that 10 beds inside the pediatric ICU were used for adult COVID-19 critical patients.

In the ICUs of the authors of this study, most of the help with the direct management of the patients was provided by anesthesiologists, and occasionally also by pediatric intensive care specialists, under the supervision of intensivists. Most intensivists managed 3 or 4 patients and supervised residents and other specialists, generally anesthesiologists who were seeing another 3 patients themselves. However, at a given point, some intensivists were taking care of 6 patients each.

The roles were determined depending on the previous training and experience of the volunteers present in the ICU. Non-ICM specialists were not always involved in the direct management of patients. For example, in some centers ENT specialists performed tracheostomies; orthopedic surgeons, gynecologists, and general surgeons made prone teams (teams to turn the patients onto their front to improve their oxygenation)^{Reference Kulkarni10}; pediatricians helped with note-writing and asked for labs, and psychiatrists gave information to the families. In addition to the direct care of patients, anesthesiologists also helped with airway management. Nevertheless, they occasionally lacked specific training in intensive care pathologies. The majority of the other specialists had to be supervised by the ICM specialists, in particular with the respiratory management of the patients.

Physicians face limitations when called for to assist in mass casualty situations without the benefit of formal education in disaster medicine.^{Reference Gowing, Walker and Elmer11,Reference D’Alonzo12} Currently, there is no standardized disaster curriculum for health-care professionals to create unified disaster medicine core competencies.^{Reference Kommor, Hodge and Ciottone13} An education strategy to prepare noncritical care physicians to assist in disasters as ICU physician extenders would be useful.^{Reference Seda and Parrish6,Reference Wax8} Noncritical care staff would benefit from a basic adult and/or pediatric critical care course, providing basic concepts essential for critical care, such as recognizing critical illness, airway management, ventilator management, antimicrobial therapy, management of shock, neurocritical care, and basic management of trauma patient.^{Reference Seda and Parrish6} Examples of these courses include the Fundamentals of Critical Care Support and the Fundamentals of Disaster Management.^{Reference Parker7} Education for nonpediatric critical care providers to manage critically ill children and vice versa and education for pediatric intensivists to comanage critically ill adults with adult intensivists should also be considered.^{Reference Wax8}

The mortality rate in the ICU of the authors of this study usually ranges between 10 and 13%. During the pandemic, the mortality rate of COVID-19 patients was higher and varied between 25 and 35%. However, when we compare these rates with the mortality rates for acute respiratory distress syndrome (between 35 and 40%),^{Reference Bellani, Laffey and Pham14} the increase in ICU capacity did not seem to play a major impact on the outcomes.

Studies to determine willingness to work in disaster scenarios have found that the 3 most important factors influencing disaster response are concern for the safety of the family, belief in the physician’s duty to provide care, and availability of personal protective equipment (PPE).^{Reference Snipes, Miramonti and Chisholm15} The ability and willingness have been reported to be lowest for fatal infectious agents and radiation events.^{Reference Qureshi, Gershon and Sherman3,Reference Masterson, Steffen and Brin16} However, in our study, most of the centers kept their usual number of ICM specialists and residents, and in some cases, they hired more specialists.

During the epidemic of SARS that started in 2002, 40% of the people who became infected in Toronto were health-care workers, many of them critical care professionals.^{Reference Parker7} During the 2014-2016 West African epidemic of Ebola Virus Disease (EVD), health-care workers in West Africa were shown to be at 21-32 times higher risk of contracting the disease compared with the general population, and by the end of the outbreak, 881 health workers had become infected and 513 had died from the virus.^{Reference Gee and Skovdal18} At a time when the demand for ICU support was increasing, the available staff was decreasing.^{Reference Parker7} Similarly, some respondents described that their ICUs had a shortage of their ICM specialists because they were infected by SARS-CoV2. Strategies such as health-care worker education, provision of adequate PPE, availability of proper treatment (or prophylaxis) for various dangerous agents, and assurance of effective environmental control are necessary.^{Reference Qureshi, Gershon and Sherman3,Reference Gowing, Walker and Elmer11,Reference Masterson, Steffen and Brin16}

Our study has several limitations. First, because it is survey-based, it is subject to response bias and individual willingness to report. Second, the results may not be generalized to other health-care systems in other countries. In Spain, Intensive Care Medicine is a recognized medical specialty with a 5-y duration residency.¹⁹ Most critical adult patient beds (71%) depend upon the departments of ICM, and the rest are distributed in variable numbers among different specialties (particularly postoperative, coronary, and cardiac surgery units).^{Reference Martin, Leon and Cunat20} The nursing personnel, which may or may not have specific training in critical care medicine, are also an important part of the staff of the services of ICM.¹⁹ However, we do not have advanced nursing providers, such as nurse practitioners and nurse anesthetists, or other providers, such as physician assistants and critical care paramedics. Moreover, we did not study what happened in centers in which the ICUs are managed solely by anesthetists and other specialists.

A third limitation is that most of the responses came from hospitals located in Madrid and Catalonia, with less participation of centers from other regions. This occurred because our survey was distributed by means of WhatsApp groups of physicians who came mainly from these 2 communities. Furthermore, larger hospitals from Madrid answered our survey, but this did not happen with most of the large hospitals in Catalonia. The reason why some of the referral hospitals in Catalonia did not answer our questionnaire is unknown. A possibility is that there was a drag effect within the WhatsApp group in Madrid because the chiefs of the bigger hospitals commented inside the chat group that they had answered the questionnaire. On the contrary, smaller hospitals in Catalonia were probably easier to organize; therefore, it was easier for their chiefs to know the data to answer directly to the questionnaire. Fourth, we did not study the gender, age, or marital status and number of children of the physicians who came to the ICU. In some studies, studying willingness to respond to disasters caused by biological agents, for example, men were significantly more likely to respond than women (67.9% vs 38.3%).^{Reference Masterson, Steffen and Brin16} Physicians may have childcare and eldercare obligations that might affect their ability to report to work in several ways.^{Reference Qureshi, Gershon and Sherman3} Finally, we have only studied the physician response in the ICU. In all of our units, nurses and other clinical support staff from other hospital areas worked in the newly improvised ICUs. Moreover, numerous physician specialists and residents supported the internal medicine department in the management of less severely affected COVID-19 patients and we did not include them in the study. Other medical and surgical specialty demographics were likely present in that setting.