Data snapshot: What is our capacity for measuring hospital capacity?
Data snapshot: What is our capacity for measuring hospital capacity?
Respiratory virus season is coming. How are things looking?
Long story short: Covid-19 levels are low. RSV levels are low. Influenza levels are low. Obviously this won’t last, but here’s the CDC’s latest summary of the national situation:
How ready are hospitals for what’s coming?
The question of how hospitals are doing is a separate matter. During the Covid-19 emergency, we had a better idea about hospital capacity than usual, because the federal government required that hospitals report relevant data on a continuing basis. How many inpatient beds were being used? How many beds could they fill? These reports helped officials understand how our nation’s hospitals were coping with waves of Covid. However, after the official emergency period ended, hospitals were no longer required to report their capacity data. To me, this was a huge step backward because it meant we were flying blind. Hospital capacity data is not trivial. It’s actually a matter of national security.
Fortunately, the federal government announced earlier this year that starting in November, hospitals would again be required to report hospital capacity information, as well as information about the major respiratory viruses (Covid-19, influenza, and RSV). It will be interesting to see if the data improve next week. Here’s a readout from the Inside Medicine dashboard where we track these things:
You’ll notice there have been a few moments where the number of hospital beds suddenly changed. These reflect changes in reporting not actual changes in beds. So, there was a huge drop in the number of hospitals reporting data when the official Covid-19 public health emergency ended in May 2023. Then in the fall of 2023, the numbers went up, apparently because the CDC made a push for voluntary reporting as the 2023-2024 respiratory virus season began. That seemed to work. Then in May 2024, the number of hospitals reporting capacity data tanked again, this time coinciding with the end of the reporting requirement (apparently some places stopped reporting in May 2023, thinking that they didn’t have to anymore. It’s unclear why they thought this.) But note that in the last two weeks (the arrow without any text on the right), reporting has once again plummeted. We are not sure why this is, but we hope it is due to hospitals preparing to go back to mandatory reporting as of November 1st.
The most recent drop in hospital capacity reporting is also somewhat different from the previous ones (see the arrows in the graph above) due to something the graph does not show: the percentage of inpatient beds being occupied. The other two times reporting fell suddenly, the hospitals that kept on reporting seem to have been representative of the nation overall. I know this because occupancy (that is, beds occupied divided by beds available) did not change suddenly, even though the denominator did. But this last time, occupancy suddenly decreased, by around 10% (from around 75% to 65%). That tells me that a bunch of hospitals that are much busier than average dropped out of the data. I hope that all of this will be fixed with the coming of the resumed requirement. We’ll know in a week or two.
Why comprehensive reporting is needed.
The American Hospital Association is not thrilled with the revived requirement, mainly because it takes administrative time to complete. And I tend to be sympathetic to the AHA’s pushback against the federal government’s ever-growing list of requirements on hospitals. But this one is really important for a few reasons.
First, it applies to nursing homes (“long-term care facilities”). That is absolutely critical. In 2023, 22% of all Covid-19 -attributed deaths occurred in nursing homes—and likely that understates the real number, because people who lived in a nursing home but died in a hospital don’t show up in those data.
Second, sampling is not enough. Sampling is when epidemiologists use a subset of hospitals (say 10%-15% of them) as a proxy for the national situation. That works well in polling. But the problem is that it fails to identify regional pain points, where outbreaks are happening. Looking at wastewater for various viruses and realtime hospital capacity data is powerful. It can help hospital and public health officials anticipate the need to increase staffing in the face of a wave of Covid, flu, or RSV.
Third, hospital capacity data is useful for understanding our national threat readiness for any threat, be it a virus, an environmental disaster, or a human-made problem like bioterrorism. That’s why when Benjy Renton and I rolled out our dashboard that showed which counties in the US were at risk of having more inpatients than available beds, people at the Department of Homeland Security and Veterans Affairs were interested. So, while the regulation requires information about Covid, flu, and RSV, the most important part, in my view, is the capacity data. (That said, the viral data are important too.)
What capacity data can do.
Hospital capacity data has some obvious applications. At the extreme, it can and should guide public policies. If a county is on the brink of exceeding its hospital capacity, staffing models need to be amended to accommodate more patients safely. When that is not possible, government officials should take actions to decrease the spread of any pathogens that might be driving increased demand. That might take several forms—from handing out masks at train stations to quietly decreasing restaurant capacity by 10% to 20%—enough to slow down a virus, but not enough to grind economies to a halt (in fact, during big waves like early Omicron, people stayed home more often anyway; encouraging that to happen a little sooner can actually mean that behaviors can return to normal sooner as well). Remember, anything that decreases contact rates in the community decreases infections. That leads to fewer illnesses and therefore eases the strain on the healthcare system. Overall, our capacity to know our hospital capacity has a lot of power—if we choose to harness it.
Thanks to Benjy Renton for managing the Inside Medicine dashboard that powers posts like this one.
Thanks for sharing this info