Editor’s note: The Salt Lake Tribune is providing free access to critical stories about the coronavirus. Sign up for our Top Stories newsletter, sent to your inbox every weekday morning. To support journalism like this, please donate or become a subscriber.
It was just a bit awkward.
At this week’s coronavirus briefing, Utah Gov. Gary Herbert bounced some questions he frequently receives off of his top epidemiologist, Dr. Angela Dunn. It’s a good idea — the old-fashioned FAQ is a useful way of delivering easy-to-understand information, a trope I’ve used repeatedly in my columns too.
But the delivery was a little disjointed. Every time Herbert asked a question, he had to go to the podium, which then meant Dunn had to socially distance from him off screen, and then walk back up there when he was done. It’s difficult to verbally answer questions like that fully, yet succinctly and clearly. And Dunn — who has been a reassuring voice of reason during all of this, as Herbert noted — didn’t really have any visual aids, which generally help. She did as well as she could with this format.
But I was left thinking there was some nuance that could have been added if it was a written FAQ rather than an on-the-spot verbal one. While I’m not as smart as Dr. Dunn, I do have Utah’s public coronavirus info, data visualization tools and a bunch of hyperlinks at my disposal. Let’s give it a shot. Here are the governor’s questions.
“What are the differences and similarities between COVID-19 and our traditional flu?”
There are four types of influenza — A, B, C, and D. The first two, A and B, are the common ones, and within A and B, there are different subtypes and strains. For example, swine flu, or H1N1, is a type of Influenza A. Every year, to develop the seasonal flu vaccine, we make informed guesses on which ones are going to spread most.
Still, compared to COVID-19, they all have significant differences.
COVID-19 is more contagious. The seasonal flu typically has a contagion rate of about 1.3, which means the average infected person passes it along to 0.9 to 2.1 others. The contagion rate for the 1918 Spanish flu pandemic has been estimated to be about 1.4-2.8, while the H1N1 epidemic of 2009 had a contagion rate of 1.4-1.6. COVID-19′s default contagion rate appears to be between 2 and 3.
Now, that may not seem like a big difference, but it is. One point of contagion rate makes a huge difference in how quickly a disease spreads, because the process is exponential.
COVID-19 symptoms are generally more severe and less predictable. COVID-19 symptoms usually come on slower, which means that a person is likely the most contagious right before they start feeling sick. The disease is also much more likely to attack the circulatory system: patients with COVID-19 were nine times more likely to have blood clots in their lungs than patients with H1N1.
However, for children, the flu is more likely to lead to severe sickness than COVID-19. Some kids get quite sick from the coronavirus and a rare few have gotten Kawasaki-syndrome-like illness, but usually, for kids, the flu is the worse disease to face.
Flu treatments are well established. We have lots of good treatments for the flu, however, only a couple for COVID-19 have been discovered and approved so far.
COVID-19 is more deadly than the flu. This was Herbert’s next question.
“How much more deadly is COVID-19 than is the flu?”
There are two ways to look at deaths: there is the case fatality rate (CFR) and the infection fatality rate (IFR). The case fatality rate is simple division: you divide the number of deaths by the number of people who tested positive.
The infection fatality rate divides deaths by the total number of infected people, not just people who were fortunate or proactive enough to get a test. Thanks to antibody surveys from all over the world, we have a much better idea of the IFR than we did before.
Using both methods, the death rate for the flu is much lower than that for COVID-19. For the 2009 pandemic flu, the CFR in the U.S. was 0.6%, with an IFR of 0.02%. In the U.S., COVID-19′s CFR is at 3.4%, about six times higher. And the Centers for Disease Control and Preventions’ current best estimate of the IFR is 0.65% in the U.S, and that feels about right given serological studies done in Indiana, Spain, Iceland, and the U.K. That’s about 30 times higher than the flu.
We should note that Utah’s case fatality rate right now is 0.76%, though remember that some people who currently have the virus will die and that may boost that somewhat in the final calculus. Dunn noted in her answer that the number of Utahns who have died of the coronavirus in the past five months has already topped the number of Utahns who died of the flu during the most recent flu season, which spans nine months.
Why is Utah’s COVID-19 fatality rate so low? Utah’s version of the coronavirus isn’t different than anywhere else’s, but Utahns have been more likely to get tested — about one in six Utahns have been. Moreover, Utah’s population is younger and healthier than anywhere else: A recent University of Utah study found that about a quarter of Utah’s lower fatality rate was due to our younger population.
“What’s the most important data that the state uses for making our decisions? And what should the public use?”
OK, Guv, the state has not been particularly data driven in making its decisions, especially as it relates to the color-coding system. Utah’s government set up clear data benchmarks and then chose not to abide by them because elected leaders wanted to open businesses earlier. That’s a defensible decision, actually, but let’s be honest about it.
What should the public use in making their decisions? Dunn pointed to coronvirus.utah.gov, and I do check that site every day. And beyond that, here’s what I look at:
The current state of the coronavirus in my county. While the state’s site has the case rates broken down by each county, and even a map of cases per 100,000 persons, that doesn’t tell me what I really want to know: how many of those cases happened recently? The easiest spot I’ve found to check this every day is The New York Times, which tells you county-by-county new cases for the past seven days per capita.
For example, Salt Lake County had the seventh most infections per capita last week among Utah counties. San Juan, Juab, Garfield, Utah, Duchesne, and Weber had higher infection rates.
The risk of running into a contagious person. Georgia Tech created a risk assessment tool for everyday people to figure out how dangerous a given event is in their community. You select an event size, estimate how many coronavirus cases aren’t being caught in testing (I recommend choosing “5″ in their current settings), then find your county on the map. The tool then calculates the approximate risk that someone who has been infected in the past 14 days would be at that event.
An example: let’s say I’m thinking of going to a movie in Utah County. I look at the theater’s website and see it is allowing 50 people per theater, and this particular showing looks pretty likely to be near that capacity. I put the numbers in the calculator, and it estimates that there’s a 51% chance that someone with COVID-19 is going to be there. Yikes! I reconsider.
The effective contagion rate. Naturally, I want to know about the trends: whether cases are rising or falling. Of course, I can watch case counts, but how about a snapshot? The effective contagion rate — how many people on average get the virus from an infected person — is my best shot.
Rt.live is still my favorite place to check the contagion rate in Utah and the nation. The site created by Instagram co-founder Kevin Systrom takes into account both case numbers and the number of tests in its model. Right now, it says the contagion rate in Utah is 0.95, which means the virus is declining. Masks are working! (Note this is very close to what Dunn said the rate was and she relies on state data that is not public.)
“We hear about case numbers. We hear about infection rates. We hear about transmission rates. Can you explain the difference between those three?”
Case numbers: The total number of people who have tested positive for the coronavirus in Utah. Note this is different than the number of positive tests, so people who get two positive tests are not counted twice.
Infection rate: The percentage of people in any population who were infected over a given time period. Salt Lake County is bigger than San Juan County. Adjust for that.
Transmission rate: Also known as the contagion rate, this is the number of people each infected person passes the virus onto on average.
“How do we, in fact, determine if somebody died based on a COVID-19 infection?”
This was my favorite answer of Dunn’s, so here it is:
“We at the health department are notified of every individual who has COVID-19 and then passes away. We work with clinicians and our medical examiner to determine the cause of death.”
“Just because someone has COVID-19 when they pass away does not mean that they will be counted as a death due to COVID-19. We actually investigate those cases to make sure that the cause of death is related to COVID-19, so that if that individual had not had COVID-19, they would not have passed away.”
“So all of our deaths are specifically due to COVID-19, and I’m very confident that we’re capturing the deaths accurately.”
I’ve also written about and investigated Utah’s deaths and believed the reporting to be accurate. So far, Utah has had 310 COVID-19 deaths. As of July 22, the CDC estimated that there had been 489 excess deaths in Utah compared to a normal year since Feb. 1. The state’s reported death count is within the margin of error of the estimation, though.
“We’ve heard a lot of talk about mask wearing and certainly in Salt Lake County.... has Salt Lake County really gone down, or has it plateaued? We know that they have a higher — about 57% — death rate. How does those juxtapose and is Salt Lake County getting better, or is the rest of state getting worse?”
The above was actually edited for clarity, and still is pretty jumbled. I see three sub-questions:
Has Salt Lake County really gone down, or has it plateaued?
It pretty clearly has gone down. Here’s the 7-day average case graph since the middle of March.
What’s the deal with Salt Lake County’s death rate vs. the rest of the state?
Of Utah’s 310 deaths, 176 have been people from Salt Lake County — about 57%. But Salt Lake County is only 36% of Utah’s population, and 47% of Utah’s coronavirus cases. What’s going on?
Salt Lake County has had 104 of their deaths in long-term care facilities, which means they’ve had 72 of the state’s non-long-term facility deaths. There have been 165 non-long-term care deaths. So 43% of those have happened in Salt Lake County, right about what you’d expect.
In other words, Salt Lake County’s higher death rate is caused by the deaths of people in nursing homes or other long-term care centers. There are more long-term care facilities in urban areas in Utah, though it’s difficult to see if those in Salt Lake County have a higher fatality rate than elsewhere.
How does Salt Lake County’s cases trend compared to those from the rest of the state?
Here’s the graph:
You’ll notice that the rest of Utah peaked higher and later than Salt Lake County, but is declining quickly too. There’s a reasonable hypothesis that Salt Lake County’s mask mandate is the reason for the earlier peak, and then it took a while for the rest of the state to catch on with the importance of mask wearing — perhaps as national stores started their own requirements. But it could be a number of other factors, too. Correlation is not necessarily causation.
“Can you explain what the truth is about the recovery rate of the people of Utah?”
Utah’s recovery number confuses a lot of people: the state says 28,747 people have recovered from the coronavirus. What does that mean?
It only means that they tested positive at least three weeks ago and haven’t died yet. It’s kind of a made up number, to be honest, and I’m relatively surprised the state reports it — there are plenty of people who tested positive three weeks ago and are still sick, either in the hospital or just experiencing minor symptoms. Of course, some people take significantly less than three weeks to recover, too.
This doesn’t take into account longer-term symptoms, which was the topic of the governor’s next question.
“What are some of those [long-term] uncertainties that people are looking at with contracting COVID-19?”
As Dunn pointed out, recovery is a relative term. What we’re still really figuring out is the number of people with long-term consequences of the virus, as well as the severity of those effects.
A more substantial conversation about that is here, but I’ll note the effects can range from minor to major. For example, Jazz center Rudy Gobert has a common long-term symptom: his sense of smell and taste isn’t all the way back yet, despite testing positive on March 11.
Some patients who were hospitalized for the original SARS in 2003 still had diminished exercise capacity years later when compared to the population as a whole, though it’s unclear how much that finding applies to COVID-19. And some long-term COVID-19 patients, hospitalized or not, are still experiencing effects like confusion and headache. Immune cell levels are out of whack in some patients for weeks or months afterward.
Finally, blood clotting triggered by the disease can cause real long-term effects, like diabetes, kidney failure, and even stroke.
Does this happen to all COVID-19 patients, or even a majority? No. But there is a substantial risk of it, one that is larger than the fatality rates mentioned above.
“How many hospital rooms are being utilized? ICU beds? Are we in danger of overwhelming the health care system?”
Dunn said the real limitation is internal care unit beds. We have “between 500-600″ of those beds in Utah, and right now, according to Utah’s coronavirus dashboard, 67.2% are occupied. And 82 of those are confirmed COVID-19 patients, or about 20-25% of the state’s ICU patients.
We are in trouble well before we fill 100% of the ICU beds. Dr. Mark Briesacher of Intermountain noted that 80-85% capacity was considered stressing the system. But the good news is that ICU bed occupancy rates have gone down slightly, off recent highs at 70%.
Meanwhile, Dunn wasn’t worried about the non-ICU beds. She says the state has “robust capacity” there.
Is there anything Herbert should have asked about, but didn’t?
I wish he would have asked about testing rates in Utah, because it’s clear they’ve declined at the same time as case counts have dropped. Here’s the 7-day average graph:
The rate of people who test positive has stayed largely constant over the last month at about 10%, but why the recent decline in the number of tests? State officials pointed to the July 24 holiday weekend, but by now, that would have normalized. Those who work at testing facilities have a simpler explanation: fewer people are coming in to get tested.
Quite frankly, the simplest explanation for that is that fewer people are sick. But there are others: fewer doctors sending their patients to get tests, perhaps decreased awareness of the importance of testing. We don’t really know what’s going on here, but it’s definitely worth noting.
Of course, there’s a ton else that Herbert could have asked about: preexisting conditions, the disease’s impact on minorities, its impact on young vs. old, etc. At the risk of sounding like a Dunn-booster or a state official, the Utah Department of Health’s website really does have a lot of good information about the disease. Check it out!
Andy Larsen is a Salt Lake Tribune sports reporter who covers the Utah Jazz. During this crisis, he has been assigned to dig into the numbers surrounding the coronavirus. You can reach Andy at email@example.com or on Twitter at @andyblarsen.