On Thursday, the Centers for Disease Control and Prevention (CDC) logged over 2,000 deaths from COVID-19 for the first time since March. With the delta variant dominant across the nation, cases were at record levels in states like Hawaii and Oregon, which had been among the best at handling and tamping down past waves. However, the bulk of the new cases—and new deaths—continued to be in the Southeast. A band of states from Louisiana to South Carolina all have rates of disease that exceed 500 cases per 100,000 citizens over the last week. Those same states are generating stories of too many COVID-19 patients in hospital beds, too many patients in intensive care units (ICUs), and rates of positive tests at record highs.

One thing several of those states are not producing in record quantity: public data. As Kaiser Health News reports, multiple states have pulled back on the amount of data they’re sharing about the COVID-19 crisis. Florida has been a leader in obscuring data from the outset, having ordered medical examiners to stop publicly listing cause of death and famously firing and eventually prosecuting the data scientist who built their state COVID-19 dashboard. That dashboard has been gone from Florida’s state website for months now as the state has repeatedly scaled back its reporting. Official data is now only available once a week, and even then Florida has adopted a new method of reporting deaths that makes each current day seem like a triumph (Florida officially notched just eight new deaths for Thursday) while the actual numbers are much higher (it really logged over 400 deaths spread across the last few weeks). The Florida agriculture commissioner and the states’ only Democratic official, Nikki Fried, has been providing some updates on a daily basis, but there are limits on what Fried can offer simply because the state is not making the data easy to obtain.

Florida isn’t alone. In Georgia, the state has pulled the plug on reporting of cases and deaths in both prisons and nursing homes. Information is still being passed along to federal officials, and some of the information the state used to provide gets rolled up into the numbers available on the CDC’s data tracker. However, those coming into Georgia’s state dashboard will find that if they want to know what they’re not seeing, they’ll have to go to another site and whip out the calculator. 

Nebraska took its dashboard down completely, then put it back—but only with weekly data. Iowa is also down to reporting data once a week. South Dakota has often been a hole in the map for weeks, even in the CDC data. As in Georgia and Florida, data is still moving to the CDC on a daily basis from hospitals, but it provides a limited view of what’s happening. 

It’s almost as if these states are more interested in pushing a narrative, rather than providing the truth.

As Kaiser notes, data from prisons has become particularly hard to come by. Even when states do provide such data, it has grown more limited in details and in some states is only available on a monthly basis. Considering the way COVID-19 has spread massively through some prison systems and how guards and other prison workers have acted as a conduit carrying disease both into and out of prisons to the broader community, this is a big concern. 

Data from nursing homes is also a concern. Despite the increase in deaths for younger victims of COVID-19, the majority of deaths continue to be among those over 60. Almost 19% of Americans over 65 are still not fully vaccinated, and multiple states have refused to require workers at nursing homes to be vaccinated. As a result, federal data shows that over a third of nursing home workers—540,000 out of 1.5 million—remain unvaccinated. The rate of vaccination tends to mirror the general state population, meaning that in several states a majority of these workers are not vaccinated. President Biden has informed facilities that all employees must be vaccinated this fall if they want to keep getting federal funds, but for the moment in many states it’s difficult or impossible to determine the actual level of COVID-19 within nursing homes.

With the decreased levels of data provided from many states, it can be difficult to see what’s really happening. But one thing is clear: What’s happening is bad. As CBS reports, Georgia has joined those states where hospitals are so full that emergency rooms can’t take any new patients. Multiple hospitals across the states have declared themselves on “diversion,” meaning that ambulances are directed to take patients elsewhere. As might be expected, this kind of situation can easily cascade into a widespread outage, with ambulances either shuttling back and forth or waiting in lines for a bed to become open.

As for Florida, back on Aug. 11, Gov. Ron DeSantis assured everyone that the state was on its way to recovery—they just couldn’t see it yet. That’s because, as Florida Politics reported at the time, DeSantis claimed that there were “leading indicators” in the data he was seeing—data not available to everyone else—that showed the delta wave coming to an end. According to DeSantis, areas of the state were already “nosing over” when it came to their rates of COVID-19. Over two weeks later, Florida is still leading the nation in both cases and deaths. 

New cases of COVID-19 in Florida are still setting records.

Florida’s cases aren’t just the highest in the nation, they’ve been consistently higher than they were at the top of the midwinter peak, which came before widespread vaccination was available. Florida actually has 51.9% of its population fully vaccinated, but DeSantis’ policies—forcing schools to open in person no matter what, banning mask mandates, forcing businesses not to check for vaccination—have completely negated the protection that vaccination should have afforded his state. 

But for Florida—and, strangely enough, for Sen. Rand Paul—there is some good news. That’s because a study published this week in Science shows that those who have been through a SARS-CoV-2 infection are left with immunity that appears to be “much greater” than that afforded by vaccines. The data shows that those who have previously been infected with the virus are much less likely to experience a second or “breakthrough” infection, and that includes infection by the delta variant. Those previously infected also matched the vaccinated when it came to having a lower chance of serious illness if they did become infected. So on this one very, very limited thing, Paul was right: His past infection with COVID-19 is providing him with protection against a repeat.

The study was conducted in Israel, where the vaccination rate is high. That allowed researchers to see that while both vaccinated and previously infected people were much less likely to be infected by the delta variant than those who were unvaccinated, those who had been infected had a significant edge. In one part of the study, the chances for the vaccinated to become infected were actually 27 times higher. Remember that number, because Republicans are sure to cite it when they misuse this data. Which they will. However, that’s a bit misleading because both the vaccinated and the previously infected had a huge edge over the unvaccinated.

All of this is good news for everywhere like Florida (or the still-spiking U.K.), where infection rates have been high. It means that the infected population is very unlikely to be “repeat customers” for COVID-19 in the near future. Combine the rates of infection with rates of vaccination, and some locations really should be getting near herd immunity, even if reaching that level requires over 90% of the population be protected. The Science study is a very good sign that eventually, this thing should genuinely burn itself out, at least regionally, and at least for a period of months … and that was not a sure thing.

However, one thing the scientists involved in the study want to make clear is that people should not go out and intentionally try to get infected. Let’s do that one again: People should not go out and try to get infected. No COVID-19 infection parties.

That’s because while getting infected does appear to convey lasting immunity, it does so at a very high price. That price is around 15% of people becoming seriously ill, a still-unknown percentage left with lingering damage, and—getting back to where this article started—about 2,000 people a day dying from infection.

What the scientists want everyone to know is that the vaccine also imparts strong, lasting protection, and unlike the disease, the vaccine will not kill you, send you to the ICU, or leave you dragging around an oxygen tank for life. When choosing a method of protection from COVID-19, choose the vaccine. Every time.

Eventually, areas will gain enough protection, either through vaccine or infection, to limit community spread of COVID-19. The question is: at what price?