ProPublica wanted to know why experts were wrong when they said U.S. hospitals would be overwhelmed by COVID-19 patients.
This article was first published on Monday, June 15, 2020 in ProPublica.
The prediction from New York Gov. Andrew Cuomo was grim.
In late March, as the number of COVID-19 cases was growing exponentially in the state, Cuomo said New York hospitals might need twice as many beds as they normally have. Otherwise there could be no space to treat patients seriously ill with the new coronavirus.
"We have 53,000 hospital beds available," Cuomo, a Democrat, said at a briefing on March 22. "Right now, the curve suggests we could need 110,000 hospital beds, and that is an obvious problem and that's what we're dealing with."
The governor required all hospitals to submit plans to increase their capacity by at least 50%, with a goal of doubling their bed count. Hospitals converted operating rooms into intensive care units, and at least one replaced the seats in a large auditorium with beds. The state worked with the federal government to open field hospitals around New York City, including a large one at the Jacob K. Javits Convention Center.
But when New York hit its peak in early April, fewer than 19,000 people were hospitalized with COVID-19. Some hospitals ran out of beds and were forced to transfer patients elsewhere. Other hospitals had to care for patients in rooms that had never been used for that purpose before. Supplies, medications and staff ran low. And, as The Wall Street Journal reported on Thursday, many New York hospitals were ill prepared and made a number of serious missteps.
All told, more than 30,000 New York state residents have died of COVID-19. It's a toll worse than any scourge in recent memory and way worse than the flu, but, overall, the health care system didn't run out of beds.
"All of those models were based on assumptions, then we were smacked in the face with reality," said Robyn Gershon, a clinical professor of epidemiology at the NYU School of Global Public Health, who was not involved in the models New York used. "We were working without situational awareness, which is a tenet in disaster preparedness and response. We simply did not have that."
Cuomo's office did not return emails seeking comment, but at a press briefing on April 10, the governor defended the models and those who created them. "In fairness to the experts, nobody has been here before. Nobody. So everyone is trying to figure it out the best they can," he said. "Second, the big variable was, what policies do you put in place? And the bigger variable was, does anybody listen to the policies you put in place?"
So, why were the projections so wrong? And how can political leaders and hospitals learn from the experience in the event there is a second wave of the coronavirus this year? Doctors, hospital officials and public health experts shared their perspectives.
The Models Overstated How Many People Would Need Hospital Care
The models used to calculate the number of people who would need hospitalization were based on assumptions that didn't prove out.
Early data from the U.S. Centers for Disease Control and Prevention suggested that for every person who died of COVID-19, more than 11 would be hospitalized. But that ratio was far too high and decreased markedly over time, said Dr. Christopher J.L. Murray, director of the Institute for Health Metrics and Evaluation at the University of Washington. IHME's earliest models on hospitalizations were based on that CDC data and predicted that many states would quickly run out of hospital beds.
A subsequent model, released in early April, assumed about seven hospitalizations per death, reducing the predicted surge. Currently, Murray said, the ratio is about four hospital admissions per death.
"Initially what was happening and probably what we saw in the CDC data is doctors were admitting anybody they thought had COVID," Murray said. "With time they started admitting only very sick people who needed oxygen or more aggressive care like mechanical ventilation."
A model created by the Harvard Global Health Institute made a different assumption that also turned out to be too high. Data from Wuhan, China, suggested that about 20% of those known to be infected with COVID-19 were hospitalized. Harvard's model, which ProPublica used to build a data visualization, assumed a hospitalization rate in the United States of 19% for those under 65 who were infected and 28.5% for those older than 65.
But in the U.S., that percentage proved much too high. Official hospitalization rates vary dramatically among states, from as low as 6% to more than 20%, according to data gathered from states by The COVID Tracking Project. (States with higher rates may not have an accurate tally of those infected because testing was so limited in the early weeks of the pandemic.) As testing increases and doctors learn how to treat coronavirus patients out of the hospital, the average hospitalization rate continues to drop.
New York state's testing showed that by mid-April, approximately 20% of the adult population in New York City had antibodies to COVID-19. Given the number hospitalized in the city and adjusting for the time needed for the body to produce antibodies, this means that the city's hospitalization rate was closer to 2%, said Dr. Nathaniel Hupert, an associate professor at Weill Cornell Medicine and co-director of the Cornell Institute for Disease and Disaster Preparedness.
Dr. Ashish Jha, director of the Harvard Global Health Institute, and his team also assumed that between 20% and 60% of the population would be infected with COVID-19 over six to 18 months. That was before stay-at-home orders took effect nationwide, which slowed the virus's spread. Outside of New York City, a far lower percentage of the population has been infected. Granted, we're not even six months into the pandemic.
A number of factors go into disease models, including the attack rate (the percentage of the entire population that eventually becomes infected), the symptomatic rate (how many people are going to show symptoms), the hospitalization rate for different age groups, the fraction of those hospitalized that will need intensive care and how much care they will need, as well as how the disease travels through the population over time (what is known as "the shape of the epidemic curve"), Hupert said.
Before mid-March, Hupert's best estimate of the impact of COVID-19 in New York state was that it would lead to a peak hospital occupancy of between 13,800 to 61,000 patients in both regular medical wards and intensive care. He shared his work with state officials.
David Muhlestein, chief strategy and chief research officer at Leavitt Partners, a health care consulting firm, said one takeaway from COVID-19 is that models can't try to predict too far into the future. His firm has created its own projection tool for hospital capacity that looks ahead three weeks, which Muhlestein said is most realistic given the available data.
"If we were held to our very initial projection of what was going to happen, everybody would be very wrong in every direction," he said.
Hospitals Proved Surprisingly Adept at Adding Beds
When calculating whether hospitals would run out of beds, experts used as their baseline the number of beds in use in each hospital, region and state. That makes sense in normal times because hospitals have to meet stringent rules before they are able to add regular beds or intensive care units.
But in the early weeks of the pandemic, state health departments waived many rules and hospitals responded by increasing their capacity, sometimes dramatically. "Just because you only have six ICU beds doesn't mean they will only have six ICU beds next week," Muhlestein said. "They can really ramp that up. That's one of the things we're learning."
Take Northwell Health, a chain of 17 acute-care hospitals in New York. Typically, the system has 4,000 beds, not including maternity beds, neonatal intensive care unit beds and psychiatric beds. The system grew to 6,000 beds within two weeks. At its peak, on April 7, the hospitals had about 5,500 patients, of which 3,425 had COVID-19.
The system erected tents, placed patients in lobbies and conference rooms, and its largest hospital, North Shore University Hospital, removed the chairs from its 300-seat auditorium and replaced them with a unit capable of treating about 50 patients. "We were pulling out all the stops at that point," Senior Vice President Terence Lynam said. "It was unclear if the trend was going to go the other way. We did not end up needing them all."
Northwell went from treating 49 COVID-19 inpatients on March 16 to 3,425 on April 7. "I don't think anybody had a clear handle on what the ceiling was going to be," Lynam said. As of Wednesday, the system was still caring for 367 COVID-19 patients in its hospitals.
As hospitals found ways to expand, government leaders worked with the Army Corps of Engineers to build dozens of field hospitals across the country, such as the one at the Javits Center. According to an analysis of federal spending by NPR, those efforts cost at least $660 million. "But nearly four months into the pandemic, most of these facilities haven't treated a single patient," NPR reported. As they began to come online, stay-at-home orders started producing results, with fewer positive cases and fewer hospitalizations.
Demand for Non-COVID-19 Care Plummeted More Than Expected
Hospitals across the country canceled elective surgeries, from hip replacements to kidney transplants. That greatly reduced the number of non-COVID-19 patients they had to treat. "We generated a lot more capacity by getting rid of elective procedures than any of us thought was possible," Harvard's Jha said.
Northwell canceled elective surgeries on March 16, and over the span of the next week and a half, its hospitals discharged several thousand patients in anticipation of the coming surge. "In retrospect, it was a wise move," Lynam said. "It just ballooned after that. If we had not discharged those patients in time, there would have been a severe bottleneck."
What's more, experts say, it's clear that some patients with true emergencies also stayed home. A recent report from the CDC said that emergency room visits dropped by 42% in the early weeks of the pandemic. In 2019, some 2.1 million people visited ERs each week from late March to late April. This year, that dropped to 1.2 million per week. That was especially true for children, women and people who live in the Northeast.
In New York City, emergency room visits for asthma practically ceased entirely at the peak, Cornell's Hupert said. "You wouldn't imagine that asthma would just disappear," he said. "Why did it go away? ... Nobody has seen anything like that."
Undoubtedly some people experienced heart attacks and strokes and didn't go to the hospital because they were fearful of getting COVID-19. "I didn't expect that," Jha said. A draft research paper available on a preprint server, before it is reviewed and published in an academic journal, found that heart disease deaths in Massachusetts were unchanged in the early weeks of the pandemic compared to the same period in 2019. What that may mean is that those people died at home.
The Coronavirus Attacked Every Region at a Different Pace
Some initial models forecast that COVID-19 would hit different regions in similar ways. That has not been the case. New York was hit hard early; California was not, at least initially.
In recent weeks, hospitals in Montgomery, Alabama, saw a lot of patients. Arizona's health director has told hospitals in the state to "fully activate" their emergency plans in light of a spike in cases there. The Washington Post reported on Tuesday that hospitalizations in at least nine states have been rising since Memorial Day.
Dr. Mark Rupp, medical director of the Department of Infection Control and Epidemiology at the University of Nebraska Medical Center in Omaha, said his region hasn't seen a tidal wave like New York. "What we've seen is a rising tide, a steady increase in the number of cases." Initially that was associated with outbreaks at specific locations like meatpacking and food processing plants and to some degree long-term care facilities.
But since then, "it has just plateaued," he said. "That has me concerned. This is a time when I feel like we should be working as hard as we can to push these numbers as low as possible."
Rupp's hospital has been caring for 50 to 60 COVID-19 patients on any given day. The hospital has started to perform surgeries and procedures that had been on hold because "elective cases stay elective for only so long."
The hospital's general medical/surgical beds are 70% to 80% filled, and its ICU beds are 80% to 90% full. "We don't have a big cushion."
Even in New York City, the virus hit boroughs differently. Queens and the Bronx were hard hit; Manhattan, Brooklyn and Staten Island less so. "Maybe we can't even model a city as big as New York," Hupert said. "Each neighborhood seemed to have a different type of outbreak."
That needs further study but could be attributable to both social and demographic conditions and the type of jobs residents of the neighborhoods had, among other factors.
What We Can Learn From Coronavirus "Round One"
While hospitals were able to add beds more quickly than experts realized they could, some other resources were harder to come by. Masks, gowns and other personal protective equipment were tough to get. So were ventilators. Anesthesia agents and dialysis medications were in short supply. And every additional bed meant the need for more doctors, nurses and respiratory therapists.
In early February, before any cases were discovered in New York, Northwell purchased $5 million in PPE, ventilators and lab supplies just in case, Lynam said. "It turned out to be a wise move," he said. "What's clear is that you can never have enough."
Northwell has spent $42 million on PPE alone. "We were going through 10,000 N95 masks a day, just a crazy amount," he said. "One of the lessons learned is you have to stockpile the PPE. There's got to be a better procurement process in place."
If there's one thing the system could have done differently, Lynam said, it's bringing in more temporary nurses earlier. Northwell brought in 500 nurses from staffing agencies. "They came in a week later than they should have."
Dr. Robert Wachter, chair of the department of medicine at the University of California, San Francisco, agreed. "I've helped run services in hospitals for 25 years," he said. "I've probably given two minutes of thought to the notions of supply chains and PPE. You realize that is absolutely central to your preparedness. That's a lesson."
Experts and hospital leaders agree that everyone can do better if another wave hits. Here's what that entails:
- Having testing readily available, as it now is, to more quickly spot a resurgence of the virus.
- Stocking up now on PPE and other supplies. "We definitely have to stockpile PPE by the fall," Gershon of NYU said. "We have to. … [Hospitals and health departments] have to really get those contracts nailed down now. They should have been doing this, of course, all the time, but no one expected this kind of event."
- Being able to quickly move personnel and equipment from one hot spot to the next.
- Planning for how to care for those with other medical ailments but who are scared of contracting COVID-19. "We have to have some sort of a mechanism by which we can offer people assurance that if they come in, they won't get sick," Jha said. "We can't repeat in the fall what we just did in the spring. It's terrible for hospitals. It's terrible for patients."
- Providing mental health resources for front-line caregivers who have been deeply affected by their work. The intensity of the work, combined with watching patients suffer and die alone, was immensely taxing.
- Coming up with ways to allow visitors in the hospital. Wachter said the visitor bans in place at many hospitals, though well intentioned, may have backfired. "When all hell was breaking loose and we were just doing the best we could in the face of a tsunami, it was reasonable to just keep everybody out," he said. "We didn't fully understand how important that was for patients, how much it might be contributing to some people not coming in for care when they really should have."
Lynam of Northwell said he's worried about what lies ahead. "You look back on the 1918 Spanish flu and the majority of victims from that died in the second wave. … We don't know what's coming on the second wave. There may be some folks who say you're paranoid, but you've got to be prepared for the worst."
“All of those models were based on assumptions, then we were smacked in the face with reality.”
Robyn Gershon, NYU School of Global Public Health
ProPublica is an independent, non-profit newsroom that produces investigative journalism in the public interest.
The models overstated how many people would need hospital care.
Hospitals proved surprisingly adept at adding beds.
Demand for non-COVID-19 care plummeted more than expected.
The coronavirus attacked every region at a different pace.