Could U.S. ICUs Handle 45,000-Bed Coronavirus Load?

In 2009, I was staffing an intensive care unit where 10 of the first 40 confirmed cases of a novel influenza A (H1N1) ended up as patients in our ICU. Everyone was mechanically ventilated, three ended up on ECMO, or extracorporeal life support. I can’t recall who lived and who died. I can recall the feeling that if H1N1 was to have a 25% ICU admission rate, our healthcare system would quickly be overrun.
As an ICU doctor, I am not interested that much in pandemics. My focus is on the patient in front of me and how they will manage. ICU care is a team game and apart from me, critically ill patients require materials and people to create a chance at survival. This has been a bad flu season, and we have a steady diet of sick flu patients, all seemingly unvaccinated, that have kept up our steady ECMO business.
A recent paper described 138 hospitalized patients with novel coronavirus infection in a single hospital in Wuhan, China, claiming an ICU admission rate of 26% and a mortality rate of 4.3%. Upon review, it appears that the overall mortality rate is about 2%-3%. Based on these numbers and additional information, we can create a model for a worst-case scenario here in the U.S.
Let’s assume the total number of cases of novel coronavirus to be 18,000 at the time of this published series, with 138 patients admitted to a single hospital. This corresponds to a 0.76% hospital admission rate. The population of the U.S. in 2020 is estimated to be 331 million. Let’s assume that 10% of the entire population of the U.S. contracts coronavirus, the same rate as the annual influenza infection prevalence.
If we assume a 1% hospital admission rate, that would correspond to roughly 300,000 patients admitted to hospitals. If we assume 25% of those admitted patients end up in an ICU, that corresponds to 75,000 patients.
According to the American Heart Association annual survey in 2015, the U.S. had 4,862 acute care registered hospitals and 94,837 ICU beds with a 68% ICU occupancy rate. My own experience is that ICU occupancy rate is more than 80% but let’s assume 70% occupancy. If we assume the total number of ICU beds in the U.S. is 100,000, this corresponds to an available bed rate of 30,000. This means 45,000 patients over and above the available ICU bed rate.
Imagine now a 20-bed unit with five open beds under normal circumstances and instead, you need 10 beds. To add five beds, you need a physical place to add them, including all of the equipment and all of the staff. You will need roughly 10 nurses to staff these new beds or an additional 60,000 nurses around the country.
An ICU room costs about $400 per square foot to build and the room size should be at least 250 square feet. This corresponds to each new ICU bed costing $100,000.
So adding 45,000 new beds to the U.S. healthcare infrastructure would therefore cost $450 million.
These 45,000 new beds would need to be staffed by new or existing ICU physicians, pharmacists, respiratory therapists, dietitians, physical therapists, and occupational therapists, and supported by additional medical specialists including a variety of medical and surgical specialties.
Coronavirus management will be impacted by benefits and liabilities. It is not a mass casualty event and hospitals won’t likely need to face the prospect of many patients at one time. An ICU patient has various trajectories. In an ICU stay, the day associated with the highest mortality is the day of admission. To be frank, a dead patient does not consume resources and some patients will be transferred to the ICU and not need ECMO.
The liability here is not only the rate of rise of new cases but also the number of healthcare workers that become ill. If many healthcare workers fall ill, this will impact the healthcare workforce in negative ways. A mass casualty event does not generally sicken the healthcare worker, but coronavirus has killed at least one physician in Wuhan.
What is our reasonable response to novel coronavirus? A vaccine for this virus might be years away and the occurrence rate appears to be increasing in more than linear fashion. Our response will depend on the actual rate of transmission and our ability to control the spread here in the U.S.
I can only manage the patient in front of me. In order to combat coronavirus, we may need coordination and cooperation much beyond what we currently achieve.
Joel Zivot, MD, is at Emory University School of Medicine in Atlanta. His clinical expertise and research interests include care of critically ill patients in the OR and ICU, education, and scholarly work in bioethics, the anthropology of conflict resolution, pharmaco-economics, and a variety of topics related to anesthesiology/critical care monitoring and practice.
https://www.medpagetoday.com/hospitalbasedmedicine/generalhospitalpractice/84845