clock menu more-arrow no yes mobile

Filed under:

How to make sense of the case and hospitalization data as omicron takes off

Both metrics are important, but all of our data doesn’t matter if we don’t do anything with it.

Local residents are tested for Covid-19 at a free outdoor testing site at Farragut Square in Washington, DC, on December 21 as coronavirus cases surge in the city.
Alex Wong/Getty Images
Dylan Scott covers health care for Vox. He has reported on health policy for more than 10 years, writing for Governing magazine, Talking Points Memo and STAT before joining Vox in 2017.

A lot of Americans are going to get infected with the omicron variant of the novel coronavirus in the coming days and weeks. The United States is currently averaging more than 143,000 new cases every day, already nearly matching the peak of the delta wave over the summer and well on its way toward the record of 250,000 daily cases seen last winter.

Record case numbers seem likely in the near future. Hospitalizations and deaths will increase to some degree as well; they always do. But how much they increase remains to be seen, since more Americans are vaccinated now than were in previous waves.

This has led to a debate among some public health experts and media commentators: Which metrics — case numbers, hospitalizations, or deaths — should we be paying the most attention to? In the Biden administration, some advisers are urging the federal government to focus more on serious cases than on overall infections, according to CNN.

Both metrics have value; they just tell us different things. Hospitalizations and deaths are the most serious outcomes of Covid-19, and the ones many experts are tracking most closely. They quantify the most important toll of the pandemic and tell us when and where hospital systems are coming under strain.

“The major public health impact has always been severe disease and death,” Natalie Dean, assistant professor of biostatistics at Emory University, told me this week. “One of the goals of vaccinations is to decouple that relationship” between cases and severe illness.

During prior waves, cases and deaths moved in sync. Cases rose and, soon after, so did hospitalizations and deaths, in proportion to the severity of the dominant variant at the time. The omicron wave should, in theory, be different, due to the immunity granted by vaccination, previous infections, and booster shots.

But case numbers are still important, Dean and others emphasized. They are still a signal of how many people are sick and unable to work or otherwise go about their lives. Infected doctors and nurses who must then isolate for 10 days can add to the strain on hospital systems and lead to the worst-case scenario: people dying because they cannot get medical care. While the vaccines may reduce the chances of long Covid, people who get infected still incur some risk of developing those long-term symptoms. And, to some degree, cases will likely be a predictive indicator of a wave of hospitalizations and deaths to come.

Some of the basic facts about omicron — how much severe illness it causes on its own, how well it can evade prior immunity, and so on — are still murky. But several of the experts I spoke to expect this wave to be different from the ones that preceded it, and said that the measurements that we focus on should change too.

“The mindset has been to equate cases with deaths. That’s going to change, potentially quite a bit,” said Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota. “Hospitalizations and deaths are going to be the marker of what’s going on in your community.”

Ultimately, though, the biggest question might not be which metric to watch, but how those numbers are used. These data points are supposed to inform which measures are taken to try to control the spread of Covid-19 and to relieve any stress on the health care system. But many experts worry the appetite for public health interventions is waning, meaning these numbers could continue to pile up without any meaningful change in how the nation responds.

Why experts are focusing on hospitalizations in the omicron wave

As Dean pointed out, the goal of the vaccination campaigns has been to break the link between cases and hospitalizations and deaths, to give people protection against severe illness so that fewer and fewer people who get infected end up so sick they are admitted to the hospital. Even if the denominator (case numbers) stayed the same, the numerator (hospitalizations) would shrink.

“Some places sort of succeeded in doing that,” Dean said. The United Kingdom’s delta wave this summer is instructive.

The UK actually matched or exceeded the US in the number of daily cases, adjusted for population, as delta became the pandemic’s dominant strain. The virus ebbed and flowed over the course of several months; the UK saw peaks of 690 new cases per million people in mid-July, 565 in early September, and 693 in late October. The US barely reached more than 500 cases per million across that time frame.

Our World In Data

(A reminder: Case numbers depend on sufficient testing, and the UK has consistently and significantly outpaced the US. That is another reason hospitalizations and deaths are seen as a more reliable metric, because they are easier outcomes to measure.)

Yet over the same period, the US consistently saw more deaths per capita than the UK, sometimes enduring three times the daily death rates during the worst of the delta wave.

Our World In Data

Dean’s best guess for why the UK saw more decoupling between cases and serious outcomes than the US during the delta wave is vaccinations. The UK surpassed the US in its overall vaccination rate in late June and continues to outperform America.

The UK has done a particularly good job of vaccinating its most vulnerable residents: seniors. As of today, the United States has gotten two doses to 88 percent of its over-65 population. England specifically has reached more than 95 percent of that population with two doses, and more than 85 percent of those people have received a third dose. The US, by comparison, has boosted a little more than 50 percent of its over-65 population.

The UK’s experience demonstrates that a country can make meaningful progress toward decoupling cases and severe outcomes when more people get vaccinated.

The US has continued to vaccinate more people since the summer, just not as many as experts would like to see. So while we could see more infections with omicron than ever before, the hope is that a smaller share of those cases will end in hospitalization or death because more people have some immunity.

But omicron, because it is so transmissible, still poses a serious threat to the US health system. The numerator of cases could still become so large that, even if a smaller share of them lead to severe illness, the raw number of hospitalizations could still exceed anything we have seen so far. Already, during the summer’s delta wave, some US hospitals were so overwhelmed that they didn’t have the beds or staff to care for all of their patients who were experiencing a medical emergency. Some of those patients died.

That could happen again if omicron spreads too quickly, even if the proportion of cases that turn serious becomes smaller. That is why experts say it is so important to keep track of hospitalizations: That metric will tell us when and where the health system is coming under strain.

“I haven’t seen evidence of a full decoupling for these metrics anywhere, though fewer cases are being hospitalized and fewer hospitalized patients are dying compared to earlier time periods in the pandemic,” Spencer Fox, associate director of the University of Texas Covid-19 Modeling Consortium, told me. “I still believe that hospitalizations are the metric to track. Until we can have a pandemic surge that doesn’t threaten health care capacity, we are going to be in a position where we cannot allow unchecked transmission.”

Why case numbers are still important to keep an eye on

The more immunity in the population, the more hospitalizations and deaths become the most reliable metrics of how serious the crisis is. But that doesn’t mean we can ignore case numbers entirely. As more infections occur, the coronavirus will have more opportunities to evolve.

“We are far from everyone being vaccinated,” Eleanor Murray, a Boston University assistant professor of epidemiology, told me. “Given that every infection is a potential opportunity for a new, worse variant to emerge, we really need to still be tracking infection rates as much as we can.”

Case numbers will also give us an idea of the toll the pandemic is taking on society beyond those worst outcomes of hospitalization or death. People who test positive for Covid-19 are still supposed to isolate for 10 days, according to the CDC. That means they can’t go to work or go to school. That is an inconvenience for their lives and, depending on who they are, it could also jeopardize essential services for everybody.

A person in scrubs, gloves, and mask squats on the floor of a hospital room containing an empty bed.
A nurse waits for her next Covid-19 case to be brought from the emergency room shortly after a deceased patient was removed from the same Intensive Care Unit room at Asante Three Rivers Medical Center in Grants Pass, Oregon, in September.
Nathan Howard/Getty Images

Experts are already worried about a massive number of health care workers testing positive and needing to isolate at the same time hospitals are experiencing a surge of omicron patients. Staffing remains as big of a challenge for health systems as physical capacity, and people will not get the care they need if many doctors and nurses get infected at the same time.

“It matters who tests positive,” Osterholm said. “Many health care workers are not going to be able to work. ... I think we are going to see a substantial challenge, above and beyond what we’ve seen so far.”

He also gave the example of police officers or firefighters, essential workers who test positive and then must isolate, as another example of how widespread infections could compromise essential services and put more people at risk beyond the direct effects of the coronavirus itself.

And there remains the unavoidable fact that, even in places with high vaccination rates, increasing case numbers will likely translate into increasing hospitalizations, especially with the omicron variant proving more elusive for vaccine-conferred immunity. The relationship might not be as strong as it has been in prior waves because more people are vaccinated, but it will still be there.

“Case rates are an early warning for where we are headed with hospitalizations, even in the setting of high vaccine coverage,” said Maya Peterson, associate professor of biostatistics and epidemiology at the University of California Berkeley. “If you have a huge number of cases, even with a low hospitalization rate, you can overwhelm hospitals.”

Whatever the metric, what matters is what we do with the data we have

It’s not as simple as focusing on just cases or just hospitalizations. Both of the metrics tell us something important about how Covid-19 is affecting our society and our health systems.

But any of this data is only useful if public health authorities act upon it, said Justin Feldman, a social epidemiologist at Harvard University. There is a widespread recognition among public health experts that, in many parts of the country, no new public health measures are likely to be put in place — certainly not preemptively — and, by the time cases and hospitalizations and deaths are rising, it’s already too late to prevent a crisis.

“I believe the argument goes: Case rates may no longer be closely linked with hospitalization rates because large proportions of people are vaccinated, at least in some areas. If we used case rates to trigger certain public health policies in the past, we should now use hospitalization rates instead,” Feldman told me. “My first question is, ‘What public health measures are we even talking about?’ Politicians at all levels have mostly given up on them, even when necessary.”

Some local governments have started putting more serious measures into place, reinstituting mask mandates and requiring people to be vaccinated to eat in restaurants or attend other indoor events. The Biden White House is announcing this week a plan to make more free tests available and deploy emergency personnel in places hit hard by omicron.

But even with omicron upon us, only nine states have mask mandates in place. Many governors are uninterested in closing businesses again or putting any more restrictive policies into place, including banning large crowded indoor events. Some of them are Republicans who have been skeptical of Covid-19 interventions; others are Democrats who say that they don’t want to punish their vaccinated constituents with new rules.

Individuals can still take precautions, and local data may give them an idea of how seriously they should be taking the Covid-19 threat at any given time. But the will of government leaders to act in the face of omicron appears to be selective.

So without new interventions, all these metrics — case numbers, hospitalizations, and deaths — could continue to pile up, with a limited effect on the government’s pandemic response.

As Feldman put it to me: “These metrics are mostly useless if we’re not going to implement additional public health measures.”