This article is part of an ongoing collaboration between the Colorado School of Public Health, the Denver Museum of Nature & Science, and the Institute for Science & Policy. Watch the full recording of this session and find all of our previous COVID-19 webinars and recaps here.

In our previous episode, we gained some national perspective on lessons learned during the first six months of COVID-19. On Monday, October 12, we’re turning our focus to Colorado. Since March, state public health officials have been grappling with the challenges of the pandemic and implementing new policies at a rapid pace. How has Colorado’s response evolved since that first wave of cases in the spring, and what insights have we gained about communication, testing, tradeoffs, and the efficacy of public health guidance?

The Institute’s Kristan Uhlenbrock chatted with Jill Hunsaker Ryan, MPH, Executive Director of the Colorado Department of Public Health and Environment, and Kacey Wulff, Senior Advisor for COVID-19 Response, Resilience, and Recovery in the Office of Governor Jared Polis, about the latest caseload data and got an assessment of the state’s current level of infrastructure and preparedness.

This transcript has been edited and condensed for length and clarity. View the full recording and download the presentation slides.

JILL HUNSAKER RYAN: I'll start out with the early timeline of COVID-19 in Colorado. You know, we started seeing it in the United States in January and our state laboratory didn't actually get the reagents from the Centers for Disease Control until February. Then it turned out that the reagents weren't working very well and so for a while in February, although we knew the virus was here, we weren't able to test for it. Initially, the CDC provided the reagents just to state labs, so there weren't any commercial labs able to do testing for COVID-19.

We got our first case on March 5 in Summit County, and within a week, we started having more cases that we were identifying through testing. At that point, we really didn't know how many cases were out there or how widespread this was in Colorado. We were flying a little bit blind. We had our first COVID death on March 9 and that told us that indeed we had cases in February. The very next day, the governor declared a state of emergency, that point we weren't detecting 18 cases around the state. We were really seeing cases up in the mountain communities, and it seemed like it was so many tourists coming in. That was probably one major source of the initial spread in Colorado.

So on March 15 and 16^th, some of the ski resorts closed on their own and then the governor did an order to close ski resorts and also restaurants and bars. On the same day, we gave the message that we wanted tourists to go home, to go to their final destination. At this point, we knew we probably had widespread community transmission and just wanted to get people out of the state. Contact tracing, isolation, and quarantine become much more complex for visitors.

On March 18, we closed the schools. Like I said, it was evident that we had widespread community transmission. We're also starting to understand how incapacitating and deadly the virus could be.  On March 25, just looking at how quickly our cases were increasing, the governor issued a statewide stay at home order. At that point we had detected 1,385 cases. We were operating on an epidemiological model called the hammer and the dance that was basically saying, if you could close everything down and suppress the virus below a reproductive number of one, you'd have a lot easier time maintaining the suppression through regular containment efforts like testing, isolation, and quarantine without having to do these major restrictions of everybody in the state. On April 27, the statewide order was lifted and we went into a safer at home order. At that time, we're detecting over 1,400 cases [daily].

So I'll show you what the epidemic looks like today. When we say daily case count, what we mean is, these are the number of cases that the state health department has been able to detect mostly through testing. We are indeed into our third wave in Colorado. The wave that happened in July was really due to the Fourth of July weekend. We were able to knock it back down before got away from us. We knew that with kids going back to school and universities starting, we're probably going to see an increase. And then we also knew that with the Labor Day holiday, we might see a jump like we had seen in July. And that is, in fact, the case.

So we are as we sit here today, we are still on an increasing trend in Colorado. Our goal is to knock this down before the holidays start, and the winter season comes and the flu comes in. So we're working with counties to try and put some additional mitigation efforts in place. Remember going back to March where I said that the CDC had only given states the testing reagent and commercial labs didn't have it? What that meant was, we weren't able to test very many people. Initially we were testing about 160 people per day. Now we're testing over 20,000 people per day. They now think that in March we actually had as many as 4,500 new cases per day. And it's why we really needed to do the stay-at-home order.

Our hospitalizations follow the exact same curve. Hospitalizations normally follow about two weeks behind increasing cases because it takes about that long before somebody is hospitalized for the illness. If you have enough testing, then you know you're really able to use isolation and quarantine as containment measures. Because of the asymptomatic nature of COVID-19, we have to be able to detect when somebody has the illness. Otherwise they can walk around for 14 days infecting people and not know. In terms of positivity rate of tests, we want that below 5%. In April, we were up over 25% positivity and it showed that we just weren't conducting near enough tests to be able to be confident that we were capturing most of the cases. And now, I'm going to hand it off to my colleague Kasey Wulff from the governor's office to talk about what we've learned in the past six months.

KACEY WULFF: Thank you, Jill, and thank you so much for having me today. I really appreciate the chance to share. It’s pretty stunning when you look at those slides over the past six months to see how much our trends have changed. And I want to talk a little bit about some of the ways that we've incorporated this information into our policymaking and how that's evolved and changed over these past six months.

I think it's important to articulate that when the virus first came to Colorado, we didn't have tests. Commercial labs didn't have tests. We knew very little about how it was transmitted. We knew very little about, for example, the efficacy of mask wearing. We also knew very little about clinical management of the disease. We had a handful of isolated case studies, but the ability to do effective clinical treatment has also changed. So as much as it is a public health and policy response, it's also been a scientific response. We have been sprinting to do the scientific work, to really ground the decisions that we're able to make in the best available data and evidence, knowing that the best available data and evidence is still evolving and changing pretty profoundly even week over week.

So, as Jill mentioned, early in March we had the ability as a state to be running about 160 tests per day. Last week, we had about three days where we hit over 20,000 tests per day. And that's pretty profound. What this really shows is our ability to actually detect a high integrity estimate of how much transmission we actually have in our communities has profoundly changed. This is really important because without these public health tools – testing, containment, treatment - we really have to rely on things like social distancing in order to do containment. But by able by being able to do this testing and being able to have a high integrity estimate of what transmission looks like, it means that we're able to more and more start relying on these public health tools.

In the last seven days in Colorado, we've had over 70,000 reported cases total. But we still have one of the lowest incident rates and lowest mortality rates in the country. The country as a whole has had about 7.75 million cases. We really do look at other states and labs, we constantly evaluate other state policies, we have a wide variety of approaches on how you do containment and we monitor them all really closely. I just feel really proud that Colorado has consistently been a positive outlier compared to our sister states immediately surrounding us. And, you know, we think that that really stems from the governor's commitment and Jill’s commitment and leadership to ground our decisions in the best available data and evidence.

We know that there are many aspects of this that are heavily politicized and that's understandable. This has been an incredibly disruptive year, and people are impacted profoundly. However, we do really use our science and evidence as our North Star as we're navigating difficult decisions and we think that that's one reason why Colorado has continued to be this little sea of yellow [lower case numbers] surrounded by that orange and pink [higher case numbers in surrounding states].

Similar to how the United States is really diverse, within Colorado our counties are incredibly diverse. Our populations are incredibly diverse. The strategies to do containment in Denver, for example, are probably going to look really different than the strategies in Saguache County, or another county of ours. And so what we've been working toward is building a framework that really balances that need for consistency statewide with an acknowledgement that our communities are really diverse and there needs to be room for local management.

What we're trying to do is make sure that that level of openness is proportional to the level of transmission in a community. The more transmission in a community there is, the less open it should be, and the less transmission in a community, the more open it should be. So we're walking this tightrope between making sure that those things are right-sized. We don't want to be an inch more restrictive than we absolutely have to, from an economic perspective. We also can't be an inch more open than is appropriate, because what we know about this virus is that it grows exponentially. And so once it starts to take off, once it starts to go through that curve that we saw early in March, then we know that our our public health tools like testing and containment may not be able to contain it. And what we then have to do is look at these more these things like extreme social distancing in order to do containment. That's a scenario like we went through in March and April that we never want to have to go through again.

We have a framework that defines phases via a set of metrics around the amount of transmission happening in the community. So the positivity rate is really important because you could have an incidence of zero in a community if you are doing zero testing. So these two things they have to balance each other, you need to be doing enough testing to feel like there's a high integrity estimate of the amount of disease in the community. And the third metric is around hospitalizations, which is really a good estimate of the amount of severe disease in a community. So county is a measure on all of these three metrics, and each level sort of has the defined zone of what those metrics are. And if a county exceeds one of those metrics, that’s a check engine light. That county has about two weeks to pull that metric back down with whatever local tactics they want to do. There may be something really specific about that community, they can actually look at their local circumstances and say okay, what's going to get me back into compliance with that with that number. If they get that number back down, the check engine light goes off and they don't have to do anything else. But if they don't get that number back down, then it's the equivalent of having to take the car to the shop. And what that means is do a consultation with Jill and her colleagues at CDPHE to talk about what state and local partnership can help that community with targeted measures.

And so this is our management system with our counties to balance this, the critical importance of local control, and really customizing the response in our diverse communities knowing that there's a statewide interest because while we certainly govern ourselves based on county boundaries or municipal boundaries, the virus doesn't care. So we need there to be that balance and interplay. That's what our policy framework has evolved to be: to strike that balance.

Now I'm going to talk to you about how we've advanced our ability to really protect our most at risk populations. In March, April, and even in May, but particularly in April, residential healthcare was by far the majority of our outbreaks. And this is really worrisome because that's things like senior facilities or nursing homes. And so this is where our Coloradans really at some of the highest risk of having a severe outcome with COVID are. That amount has gotten smaller and smaller, and what it really reflects is the incredible amount of investment and effort we have made to do containment in those settings and really minimize those settings.

Some other areas have increased [recently], including camps and education and childcare. That makes sense: school was out, kids weren't in school prior to about September. What we can see is where our outbreaks are happening is where populations are interacting. And what we really focus on is making sure that we're minimizing groups and possible outbreaks happening in places with our most at risk individuals.

We can also give an estimate of the average outbreak size. There's been this incredible success story of the outbreak size being really large - around 25 people per outbreak - and that has fallen to actually under five. So what that means is, we're catching those outbreaks a lot faster and we're interfering a lot faster before they can spread to impact, you know, 10, 20, 25 people.

We know that COVID spreads exponentially. The way I like to think about this is that if you're driving in a car, and you see a car stop ahead of you, and you're going five miles an hour or you're going really slowly, you're not going to be worried about crashing into that car. But if you're going 90 miles an hour down the road, and there's a car stopped in front of you, you're going to be worried about getting into an accident. So the rate of change matters profoundly. As you can see, at the end of April we went into suppression and we had week over week decreases. And then we started to creep up gradually with some pretty profound increases right after our holidays begin. Right after the Fourth of July and then right after Labor Day. And so what we're really trying to do is change the trend to get back into suppression and get back into a decrease in the number of cases, knowing that we have holidays coming up.

The number of hospital admissions are something that's much less sensitive to the amount of testing. If you need help, that’s something we're going to be able to measure regardless of how many tests are out there. Each week, we make an estimate of what the long term curve is going to look like. Right now, hospitalizations are going up at a faster rate than our models originally predicted. So that's something that we need to pay really close attention to. And then this slide projects this information out over a longer period of time, so we can start to estimate what's going to happen over the next few months. We don't want to just be reacting. Day to day, we actually want to be thinking about our strategy over a longer period of time.

If we continue on our current path. we're not going to be in good shape for the holidays. So we need to evaluate what kinds of changes need to be made at the state level, the local level, or at the individual level. Halloween is coming up, and we know consistently every time we have a holiday, we see people interacting with each other and we see a decrease in social distancing and we see transmission. We just want to encourage everyone to have fun for Halloween and do it as safely as possible. Get creative with it, think about how you can have a fun celebration while really minimizing the number of people you're interacting with and keeping your activities outside where it is much safer than being inside.

KU: A question from Meg Wingerter, a reporter with the Denver Post, who was commenting on the fact that we saw that second wave tied to Fourth of July and then we're kind of in this third wave. With holidays coming up which Kacey was speaking to, is this a pattern we're expecting around every kind of major holiday event and so how do we get on with that pattern? Jill, could you give us the final word before you leave?

JHR: Yeah, absolutely. Meg, thank you for the question. Yes, we do believe that this is a pattern around holidays. People tend to get together instead of socially distancing. We will be providing messaging and media campaigns starting very shortly around asking people during the holidays to continue to maintain their social distancing and stay home, and, of course, wear a mask. We've heard a lot of times that people when they go to visit their families, even if their family's not in the same household, they're not wearing a mask because they're so comfortable around their family. We still want you to wear a mask for sure. Consider if you have family members that are older, or that are otherwise at risk for COVID-19, that you don't visit them. And certainly if you're having any symptoms of COVID-19, stay home and continue with the hand washing and other good hygiene practices. But yeah, we want Coloradans to understand that people's behavior does change around the holidays and then what that does is increase transmission, and it makes it much harder for us to suppress it and get it back under control. So we need people to be extra vigilant starting with Halloween.

KU: We appreciate your time this morning and we know you have to pop off, but we've got a number of questions we're going to continue on with. [Editor’s note: Jill Hunsaker Ryan had to depart for another meeting at this point in the presentation.] Kacey has graciously decided to handle all of those, so Kacey’s here in the hot seat.

I do want to ask you a little bit about the check engine light, and the various metrics used and looking at the county level map for Colorado. You said that they got about two weeks to kind of pull that metric up (or down) to kind of get back in compliance. What are some of the state level resources, when you see a local county struggling, that are available?

KW: The state does a really extraordinary job of customizing how they can support different communities. We get the vantage point of seeing all of our counties and seeing them tinker with what works and what doesn't work, and it allows us to be able to say, hey did you see what they did over here or did you see what they did over there and really be able to draw from those best practices.

The state also has an enormous amount of testing resources that were able to surge, particularly around outbreaks. And I think this is one of the key things, which is that if you're seeing an increase, first and foremost you want to figure out where it's coming from. And the best way to do that is to surge testing, because that's going to help you just continue to get a more and more accurate count of where the cases are, how many cases there are, and where they are.

When the estimated number of cases is really high and then the detected number of cases is low, you want to close that gap as much as possible. The more you close that gap, the more you're going to be able to target interventions. Same with contact tracing. That's a similar tool that allows you to understand who's impacted and why, to really get that root cause. And so, for example, we've seen a handful of communities with high numbers of college students have an increase in transmission. The strategy you're going to use to do containment in that population is going to be really different.

KU: Could you speak specifically to universities, and CU in particular, which has been in the news?

KW:  Sure. So, the Boulder County Public Health Department has done a really extraordinary job in turning that trend around. There was one week where we saw almost an 80% increase in that college age population. I mean, it was shooting up. What Boulder County has done is through a series of local public health orders targeting that 18 to 25 year old population, that student population. Keeping them in sort of a modified form of a stay at home, not quite a full stay at home, but really modifying that level based on if they were in a residence where we knew there was a lot of transmission.

We also know that Boulder City instituted a last call order. Essentially, last call orders are a proxy for making sure that people are, you know, as it gets later into the night, particularly with this population, you may see a little bit of less compliance. I think that's something that we all experience a little bit.

No one single strategy changes a trend. You have to layer the strategies on each other because each strategy has pros and cons and works better or worse, depending on the scenario, but when you use them all together in aggregate they really change a trend. Boulder’s epi curve has come down, and it's not quite, all the way down to the level that we wanted to get to, but it is well on its path. And it is, you know, it's really that check engine light that went on and then check engine light is going off. It really is incredible how they've been able to change that trend.

KU: At this point, who should be getting tested, and who is getting tested generally speaking?

KW: Just think back to when we had 100 tests a day, and are now doing 20,000 tests a day. The other thing that's changed is the kinds of tests that we can do. Almost 40% of new cases are acquired from somebody without symptoms. Almost half the time, somebody is getting sick from someone who has no idea that they're sick. They don't have symptoms. Either they are asymptomatic and they're not going to have symptoms, or they're asymptomatic, but still contagious.

Most people are contagious for about two days before they experience symptoms. I believe that that day before you get symptoms is actually when you may be peak contagious. This is so different from any other disease we have. If you have the flu, you know you have the flu. There is no ifs, ands, or buts. You are in bed, you feel like you've been hit by a truck. Whereas this disease, a large driver of transmission is asymptomatic.

So what that means is that we need to figure out how we're catching those cases. And one way we catch those cases is through testing. For people who have symptoms, we absolutely want them to get tested. For people who have been exposed to somebody, we absolutely want them to get tested. You could have been exposed. You may not be getting sick. We want you to get tested, we want to try to catch that case.

Now, there is a little bit of a push and pull. We don't want everyone to get tested every day. We just don't have the resources for that. What we really want to do is be testing the right people at the right time. That's going to get easier and easier as we have more tests on the market. You guys may have started to hear about some of these tests called antigen tests. They're different than a PCR test, which is the most accurate test, where you get the nasal swab and it gets into a lab. There are some tests that work more like a home pregnancy test, so there's the instant sort of point of care test. And those tests are really good. They're a little bit less accurate than the test that gets in the lab. What we really want to be thinking through is, how do we use the right test for the right use case, how do we use the right test for the right person?

People who are workers in nursing homes, for example, we want them getting tested really regularly, because we if we have a nursing home worker who is asymptomatic, they're working with that population that is really at risk and be really negatively impacted. What we’re trying to do is really use our testing resources to be testing those populations really regularly, that are interacting with those vulnerable at risk populations. And then eventually, starting to build up the availability for that cheap rapid quick testing for individuals that maybe have no high risk, have no reason to think that they've been exposed.

The PCR tests look to test to see presence of the virus in your body. What the antibody test does is it tests to see if your body has had a reaction to the virus and so it's looking for your body's reaction. It's not actually finding a piece of that virus itself. These tests are a little bit harder to interpret. Just because you have antibodies, it doesn't necessarily mean you had it nor does it necessarily mean that you're immune. There's just not enough understanding yet of what that will mean for our population. I don't think it's quite yet at the point where you could interpret that yourself.

KU: Could you speak a bit about your experience with how policy adapts rapidly and maybe something you’ve learned or want to reflect on?

KW: The Governor and the state, we were really early adopters of mask wearing. And actually, the governor started wearing a mask and started doing press conferences talking about the importance of wearing masks before the CDC recommended wearing masks. I mean, we were really early adopters. But if I could hit rewind and go back, that would have been something from day one, because we really do know that masks are incredibly effective at preventing transmission.

There was a really good study that the CDC put out of these two hairdressers in Missouri. They were wearing masks, they saw all their clients over the course of a weekend, maybe like 100 clients. Both of the hairdressers came down with COVID. And that means that the clients they saw, they saw them within the time period when they would have been contagious and they would have been able to transmit to the client. Not a single one of their clients got COVID, and that is just stunning. I mean, masks are incredibly effective. They're the best economic tool we have. There was a paper done by Goldman Sachs that estimated that if everyone wore a mask every time they were in public, we can prevent a $5 trillion GDP drop. I mean, it is the best economic tool we have.

And so we talk sometimes about the tension between some of our public health measures and some of our economic measures. We really want to be maximizing our public health impact by minimizing any negative economic impacts and masks are the tool at the intersection. The more we wear masks, the more we're able to keep our businesses open because the more we're able to halt transmission. So I think if I could hit rewind, and on February 1 have everyone really understand the importance of masks, I bet that we [would be in better shape].

How do we break this boom and bust cycle of the holidays? We really use all the tools we have. So, wear a mask. Wear it while you're traveling, wear it while you're indoors, with people who aren't part of your immediate household and keep your activities outdoors to the greatest extent possible. Yes, it gets a little bit colder over the holidays, but here in Colorado, we know that there's no such thing as bad weather. There's only bad gear. So bring the right gear with you to sit outside. Wash your hands and continue to follow the fundamentals. We know how this disease is transmitted. And we know how to really get ahead of that and do prevention around the people we love.