New data shine a harsh light on what can happen when hospitals become so overcrowded that they have to resort to crisis standards of care, a level of care where practice standards are relaxed under the strain of scarce resources.

In an investigation of the association between surges of patients with COVID-19 and survival in 558 U.S. hospitals from March to August 2020, researchers found nearly one in four deaths was attributable to hospitals that were strained by their surging caseloads.1

“While crisis standards of care were not necessarily invoked in many [regions] on a state level, at the hospital level, many were practicing crisis standards of care,” explained Sameer Kadri, MD, MS, lead author and head of the clinical epidemiology section at the National Institutes of Health Clinical Center. “While that becomes necessary, it comes at a cost. Modifying care standards invariably leads to medication errors, provider fatigue ... a higher bar to hospitalize, an earlier trigger to comfort care, [clinicians] practicing outside their spoken [area of expertise], and also secondary impacts on [patients with] non-COVID illnesses.”

Kadri spoke about the study during a virtual presentation on Aug. 26 sponsored by the Department of Health and Human Services Office of the Assistant Secretary for Preparedness and Response, the National Emerging Special Pathogens Training and Education Center, and Project ECHO.2 Kadri and colleagues developed a metric called the surge index to fully capture how severely hospitals were straining from the volume of COVID-19 patients.

“We felt like just using counts of patients might not be sufficient. We also needed a metric of severity because a more severe case requires more attention, requires more personnel, and requires more resources — and that leads to more strain on the hospital,” Kadri said. Consequently, patients who needed more intensive care, such as ICU patients and patients requiring ventilators, were given more weight in the metric than patients who required no intubation.

Kadri and colleagues mapped all of these data from the Premier Healthcare Database, a system that has de-identified information from more than 800 hospitals, accounting for roughly 20% of hospitalizations across 48 states. Of these, investigators focused on 558 hospitals that reported more than 15 cases of COVID-19.

Over the course of the study period, it became clear in the data that clinicians improved with more experience treating COVID-19, Kadri observed. However, he cautioned that at hospitals with the most intense surges, such improvements in clinical care were less visible. “Under ordinary circumstances, we had improved survival. When we got to a point where the learning curve was done — the second wave — it became apparent that non-surge mortality went down the most, but surge-related mortality remained high,” Kadri said. “There aren’t many things in medicine where there is one factor to which so many deaths can be attributed.”

The lesson from the data is clear: To save lives, healthcare providers must find ways to relieve capacity issues before they reach a severe level of strain. Unfortunately, this message arrives when hospitals in several states are managing yet another COVID-19 case spike. “Currently, hospitals are in a terrible crisis, and we have to really try to lead in terms of showing these hospitals what they need to do in order to decompress, and how important it is to decompress in terms of the outcomes for their patients,” Kadri said.

Kadri emphasized the need for load-balancing — transferring patients to other hospitals for care when an institution approaches the kind of surge levels that can compromise care. However, that technique is difficult when entire regions are overwhelmed with COVID-19 cases.

“It is very difficult for [hospitals] to discharge patients, particularly those who are COVID-19-positive, to nursing homes and other institutions,” says Sandra Schneider, MD, FACEP, associate executive director for clinical affairs at the American College of Emergency Physicians (ACEP). “[Hospitals] are becoming crowded with patients that they can’t send anywhere.”

Further, Schneider emphasizes that too often, efforts designed to decompress ICUs and inpatient floors come at the expense of emergency providers. In many cases, this happens when hospitals try to maintain specific nurse-to-patient ratios on the upper floors.

For example, Schneider notes that in California, ICU nurses take care of a maximum of two patients. On an inpatient floor, nurses may take care of four to six patients, depending on the type of floor. Also, while mandated nurse-to-patient ratios may be different in other states or regions, the circumstances are similar when hospitals reach capacity.

When there are too many patients and not enough nurses to maintain the required nurse-to-patient ratios, hospitals have just two options, according to Schneider. “They can either hire more nurses, or they can keep patients in the ED until an inpatient [or ICU] bed opens up,” she says. “That means right now, in many cases, there are actual physical beds, but there are not enough nurses to cover those beds. Those beds have to be closed down.”

Thus, an ED nurse may be caring for more patients awaiting admission to an ICU bed than an ICU nurse is caring for upstairs. The same dynamic is taking place for patients in the ED who are waiting for inpatient beds to open. “There are no ratios for care in the ED, so the ED takes care of whatever patients it has,” Schneider says. “The ED takes care of patients who might not be in the ED for an emergency reason. It takes care of patients who are inpatients who don’t have a place yet on an inpatient floor. It does that, by and large, with the nursing staff that it has.”

If administrators seek to maintain specific nurse-patient ratios on the upper floors by pushing all the patients awaiting beds onto the ED, that will not improve outcomes. “We have study after study after study that shows when [the ED] is crowded, we also have medical errors. We also have delays in care, and we also have increased mortality,” Schneider says.

Schneider adds there is a critical shortage of nurses in the ED at a time when more patients are presenting. She urges administrators to carefully consider how to achieve what is best for patients under difficult circumstances and to look at the entire health system, not just inpatient and ICU beds.

During the Aug. 26 virtual presentation, presenters discussed developing regional transfer centers capable of working with several hospitals to facilitate transfers and load-balance between facilities so the burden is shared. Schneider agrees transferring patients to other hospitals can be beneficial in easing crowding — but only if there are open beds in the region.

“The current outbreak has hit a much larger geographic area. Transfers out of the ED in particular have been much more difficult,” Schneider says. “Instead of making one call or maybe two calls to get a patient transferred, in some cases our members have been saying they have been making an average of four to five calls, and in some cases as many as 10 or 15 calls.”

Additionally, patient transfers are taking much longer to complete. “It used to take an hour or so; it is now taking four hours, in part because we can’t find ambulances to take patients [to other hospitals], and in part because we are having to spend so much time trying to find a bed,” Schneider reports.

While large health systems often operate transfer centers to manage such tasks, clinicians usually have to carry out the legwork at smaller institutions. “The emergency physician, the emergency nurse, and maybe a clerk are the ones who are making the calls to try to find a bed,” Schneider notes.

Regardless of who is conducting the search for an available bed, he or she typically has to look well beyond the geographic region, especially in hard-hit regions.

“I have heard of transfers from Texas that have gone to Montana, I have heard of a transfer that went from Texas to Connecticut, and I have heard of a transfer from Louisiana to Illinois,” Schneider shares. “In Florida, the likelihood is that you will not find an ICU bed readily available in the state.”

Under such circumstances, how can anyone better manage surging caseloads? Schneider notes Washington created a dashboard that shows everyone where available ICU beds are, at least within the state. That has greatly simplified the search process for busy clinicians and administrators, although they cannot see bed availability their borders.

Schneider favors the creation of a national dashboard to identify open ICU beds. She notes ACEP has begun to talk with stakeholders about developing such a resource. Schneider says there is a tool to identify available extracorporeal membrane oxygenation machines.

Schneider also would like to see many more hospitals leverage full capacity protocol, an approach that smooths out surging caseloads so not all patients who are awaiting a bed on the upper floors are boarded and cared for in the ED.3 Under such a scenario, the burden is shared between multiple hospital units or floors in a prescribed way. “There are a few hospitals that have made arrangements so that an intense backup doesn’t occur in the ED,” she explains. “For example, the ED might take the first four inpatients, and then every other hospital unit takes one extra patient. Or, the ED will take the first two ICU patients, and then each ICU will take one additional patient.”

Considering much of the crowding that has occurred during the fourth wave is a result of a severe nursing shortage, Schneider wonders how the industry could streamline tasks. This would mean nurses are not weighed down with certain screening questions or other tasks that may fall into the “nice to have” category, but are not critical to maintain at a time of extreme patient surges. Schneider says she plans to talk with nursing organizations soon about this issue. 


  1. Kadri SS, Sun J, Lawandi A, et al. Association between caseload surge and COVID-19 survival in 558 US hospitals, March to August 2020. Ann Intern Med 2021;174:1240-1251.
  2. Grand rounds for EMS, critical care, and emergency department: Load balancing. Aug. 26, 2021.
  3. Alishahi Tabriz A, Birken SA, Shea CM, et al. What is full capacity protocol and how is it implemented successfully? Implement Sci 2019;14:73.