A forecast issued by Oregon Health & Science University found that the impact of the delta variant could be grim for Oregon. Peter Graven is the lead data scientist at OHSU, and Dawn Nolt is a professor of pediatrics and infectious diseases there and the OHSU Doernbecher Children’s Hospital. They join us with details on the forecast and what Oregonians should know as the delta variant continues to spread.
The following transcript was created by a computer and edited by a volunteer.
Dave Miller: This is Think Out Loud on OPB. I’m Dave Miller. Two big things happened on Tuesday: new modeling by data scientists at OHSU found that nearly 1100 Oregonians could be hospitalized with COVID-19 by early September. That would be almost double the record set last fall and would put an enormous strain on the state’s hospitals. In response to that announcement, Oregon Governor Kate Brown made an announcement of her own. She is reimposing a statewide mask mandate for all Oregonians, vaccinated or not in indoor public spaces. The order will take effect tomorrow. We’re going to talk about all of this right now, about the modeling that led to this big change and the situation in Oregon hospitals right now, Peter Graven is the Director of the Office of Advanced Analytics and Lead Data Scientist at OHSU. Dawn Nolt is a pediatrician at OHSU who specializes in infectious diseases. Thanks very much to both of you for joining us.
Dawn Nolt: Great to be here.
Peter Graven: Me as well.
Dave Miller: Peter Graven, first, for more than a year now, it seems that your projections are the ones that lead the governor to act. If you say things are going to get really, really, really bad, then she seems to respond with mandates. Can you describe the basics of how you do your work?
Graven: Thanks, Dave. Yeah, so the basic projection models are, there is a kind of established approach which is called SIR Model, which we’re not going to give you the details here but essentially it’s looking at each person and figuring out if they are still susceptible to infection and whether or not they’ve been infected or maybe been vaccinated. And then we combine that with what we understand about how fast the virus will spread in the population. And so with those assumptions and some additional ones about what kind of effect we’ve had on policy, we can project forward where we’re at, we know how many people are still vulnerable and then we can anticipate how many are yet to get infected. Of course, the big element is not just the infections but in our case and what drove a lot of the discussion is the impact on hospital capacity. People being sick at home is not a problem for us, assuming they’re not having long-term impacts. But if they are showing up at the hospital, that’s where it causes, we have a real fixed amount of space for that. And so the projections have always been something that every week I produced them, I take in as much data as I can and I make revisions and then put them out every week.
Miller: Has the delta variant drastically changed the way the modeling works. If it’s much more infectious, does that mean that we’re going to get very different numbers when you put in, say the number of people who are vaccinated, the number of people who have already been infected and then crunch the numbers? Does the delta variant really change your model?
Graven: It doesn’t change the framework of the model but there is a parameter for how fast the current circulating strain of virus spreads and that dramatically changes the results. So we were dealing with a virus that had an R-Naught, which is a basic way of understanding how fast the virus spreads, that’s the term, R-Naught of 3. The new estimates and we kept getting new data about this in the last few weeks in how it’s impacting Oregon was that the delta variant was spreading with an R-Naught of 8. And you can literally think of that as 2-3 times stronger or faster transmission. That drastically changed the results. It drastically changed where we thought we were. The reality is the immunity level that we had coming out of June at around 70% was probably strong enough for the original virus. And I know that was part of why we shot for that number. But with the delta variant, it became so much faster that we were no longer protected by our kind of herd-protected status of having a number of vaccinated people and people who have been previously infected.
Miller: What is the new number that you think would head off the kind of upward hockey stick that we’re looking at now? If 70% or so made approximate sense pre-delta variant, what percentage do we need to see now?
Graven: I think we’re looking at 90% which I know is a very high number. We’re probably, if you consider the vaccinations and previous infections, not just people who were tested positive, but ones that we think were also infected that never was tested as a case. We think we’re probably around 73% or so now. So we have quite a ways to go and that’s what’s driving this. This big surge is about 17% of the population needs to be infected or vaccinated before this thing really slows down and unfortunately it’s so transmissible and without any restrictions, it can spread extremely fast. And that’s kind of what we’re observing.
Miller: What went through your mind when you first saw the results that the public became aware of earlier this week?
Graven: Well, I’ll be honest, I was shocked and I was, I actually had to go and check other modeling resources. I had to go and kind of triangulate against other numbers. I looked at other states and countries to see if that’s truly a plausible number. And the math held up and indeed we’ve actually seen census levels like this in other states. Oregon’s been lucky because we have taken action, so we haven’t seen these extreme situations, but the reality is in other states, the census has gotten three or four times as high as ours per capita, certainly in other countries as well. So I knew it was possible, the math showed it was happening and...
Miller: So can I interrupt you because I want to make sure that, I always assume if I don’t understand something that many listeners don’t either. When you say the census numbers, you’re talking about hospital numbers?
Graven: Correct. Hospital census, that’s the term we use. Not the U.S. Census count of people. This is the number of people in our hospital beds, and I’m adding them up across every hospital in Oregon to come up with these numbers and of course, that’s our critical resource and even more so the ICU within that, that’s the number I’m trying to predict every week.
Miller: Dawn Nolt, can you give us a sense for what the situation is practically, just right now, before we get to the worst days to come, what the situation is like right now in Oregon hospitals?
Nolt: Hospitals are crowded. They’re overflowing right now. And that was happening before delta really took hold in July and August. So we’re seeing people that are coming in not with COVID earlier, but they had delayed care during the past year because of the pandemic. They’re coming in with emergency issues. So that’s what’s been filling our hospitals earlier in the summer and now with the delta variant, we’re adding an increased pressure on the hospitals and they are already strained and to be quite honest, Dave, I don’t know how much more we can take, but Peter’s data showing that we may have to take more and hospitals are really struggling to figure out how to balance that, knowing that people need to come in to get care but we can’t put them anywhere.
Miller: What does it actually mean then? To say that by Labor Day, the state could be short 400 - 500 staffed hospital beds?
Nolt: It means people are languishing in areas that they’re not going to get the best care or that they need to go somewhere or be discharged somewhere, but they are still in a hospital setting. So what we’re saying is that we can’t fit those people in and therefore Oregonians who are having an exacerbation of a chronic medical issue. They come in and they have an injury or trauma that needs emergent care, it’s going to be very hard to get them in. And to be quite honest, Dave, we sit there and we have waiting lists of people who we need to triage, to figure out who’s going to come to a hospital like OHSU or who can actually stay at a different hospital and just be temporized there until we get a bed.
Miller: We’ve heard that the surge so far has been worse outside of the Portland area in terms of per capita hospitalization rates and available beds, worse outside the Portland area, for example, much worse in southwest Oregon in places like Jackson County than it is within Portland. How much slack can Portland area hospitals take up?
Nolt: Not a whole lot. We have prided ourselves in, at OHSU and certainly the other regional health centers in the Portland area of being the tertiary care center,
that we can accept people who need higher levels of care, but we are having a very hard time taking care of people who are already in our system. We don’t have a whole lot of slack for requests that come in from other parts of the state.
Miller: So, are we talking about beds in tents in parking lots or sending people to other states? I mean, what is, I was gonna say the worst-case scenario, but maybe the phrase is, what’s the likely scenario within a month?
Nolt: So I do know that there are some health care systems that are planning on field tents to just make rooms. We may be trying to get some help from outside the state, such as health care workers. And Dave, it’s not just the beds, it’s not just the physical beds. The bed capacity, which is also getting overwhelmed, but it’s the capacity for us to have those beds and patients in those beds be cared for by the appropriate people. And we also don’t have a great supply of healthcare workers because we have a lot of burnout. This has been going on for 20 months. And so there’s a lot of concern. We’re working with different health care systems in the state to figure out if we can move patients whenever a bed opens up. We’re trying to see, as I said, to get maybe healthcare workers outside of the state to help us. But it’s very difficult.
Miller: If you’re just tuning in, we’re talking right now about the latest modeling out of OHSU. It predicts a huge increase in COVID-19 hospitalizations in the coming weeks. We’re talking with Dawn Nolt, the pediatrician at OHSU and Peter Graven, Lead Data Scientist there who’s modeling has been some of the most important numbers that the state has been paying attention to, that state leaders have been paying attention to. Peter Graven, how big a difference could this mask mandate make?
Graven: So we’ve done mask mandates before and last summer, you’ll probably be aware, that’s when mass became much more known actually. Some of the science became a lot more clear and we made them more, they were required in more situations. I saw a direct impact on our census levels when that went into effect. That helped kind of curtail a summer surge. And we’ve also used them at other points throughout most of the pandemic. So we know the value of masks. The issue, of course, is if there are still times when we’re indoors with other people without our mask on, then the policy is not going to be nearly as effective. And you know we can all think of situations like that. I think the mask mandate is hopefully a good first step. Hopefully, it’s enough. But I do think that we are going to be looking for people to follow additional recommendations in order to make sure that we were able to slow the spread.
Miller: Well how much data do you actually have about adherence to masking mandates at this point? I mean do you have real numbers about the extent to which people follow these rules?
Graven: We actually do have some decent data. So there’s a couple sites that are essentially either surveying people on a regular basis and we can see pretty quickly whether or not people are indicating that they’re wearing their mask on a regular basis in certain situations. That rate was around 90% through most of the pandemic. It went up and down a little. After vaccination became very prevalent and many recommendations went away from wearing masks, we saw that drop all the way to 32% in the data source I looked at. More recently the CDC, the state and counties have recommended masks and we only saw an increase from 32% up to 44% in data last week. That clearly was not enough and so unfortunately the recommendation didn’t work. We needed something stronger. I think the governor recognized that. It’s great if vaccinated people are wearing masks, but it’s most important if unvaccinated people are wearing masks and we needed to make sure that was going to be the policy.
Miller: So let’s say that a really high percentage of Oregonians do either start, starting tomorrow, or continue to wear masks when in public situations indoors. How much of an increase in hospitalizations is already baked into the system?
Graven: I think we see another 200 to 250, maybe even 300 either way.
Miller: Increase in people with COVID using hospital beds in Oregon?
Graven: Yes. And that’s because, as you can imagine, you do something today and that’s going to prevent transmissions that start today, but to have showed up to the hospital for actually about 12 days. So we kind of put in a two-week buffer there between any action you can do in the time that will actually help our hospitals. So all the actions from two weeks ago, those have already been taking place. We know that those patients are going to keep showing up. Stuff that we do today though can help. And so that’s the part that became very urgent is whenever you’re dealing with exponential growth or anything that grows this fast is you have to act so soon. There’s no time to wait. And that’s part of why the modeling, I think, did help people understand the urgency.
Miller: Dawn Nolt, as I mentioned, you’re a pediatrician who focuses on infectious disease. How much of an increase have you seen in young Oregonians with COVID-19?
Nolt: Well, we have been quite fortunate in pediatrics in general that COVID-19 really hasn’t caused a high degree of infection nor a high degree of serious illness. And we are not seeing any changes with that sort of baseline knowledge between delta and non-delta variants. At this time…
Miller: I have to say that that goes against what I had thought I’d been seeing, which is more reports of more kids with delta, with the delta variant and more sicker kids.
Nolt: So that is likely a reflection of the sheer number of kids that are being infected. And therefore you will see the small 1% of those who end up in the hospital or in the ICU. But at this time, the data is fairly limited about the impact of delta on children. But it appears that it’s not causing more severe disease, but just more disease because there are more children being infected.
Miller: Which in its own dark way, actually could be seen as good news, for parents out here who have been, who may have gotten the message, not just that delta is more transmissible, but it’s more dangerous. You’re saying the data’s out, but there is at this point, no definitive data to say that it actually is more dangerous per se.
Nolt: That is correct. But the chance of your child who’s not wearing a mask and or who is unvaccinated, the risk of infection is much higher because this variant is so much more transmissible.
Miller: How much have you been talking to parents who don’t want to vaccinate their 12- or 13- or 14-year-olds?
Nolt: A few, I’ve spoken to a few.
Miller: What are those conversations like for you right now?
Nolt: There is some hesitancy or lack of confidence in the authorized vaccines, perhaps around the perception that they are not fully licensed and therefore the safety data is not all in. There are some concerns about long-term effects from vaccines, particularly the platforms, the MRNA platforms and these are conversations that you need to be ongoing. People have specific concerns, but we have to weigh that against the real possibility that their child could be infected with COVID-19. And we also know that COVID-19, even though it may not be severe, could have some long-term effects that we’re not entirely aware of in children. So even though parents may worry about long-term effects of vaccines, certainly if they are not vaccinated, they’re much more likely to get infected and perhaps experience the long-term effects of COVID.
Miller: Peter Graven, what’s the endpoint that you see right now? I was struck by one of the things that you said in the press conference a couple of days ago, which is that the way we were heading, it was sort of towards full herd immunity. But the worst way to do it, meaning just out of control infection. I mean in the big picture, is the main variable right now, the rate at which unvaccinated people are going to get infected?
Graven: Yeah, the main issue right now is obviously the capacity crunch. What we want to do is slow it down so that you can buy time to get vaccinated. Unfortunately, if you do not. there is a very good chance you’re going to get infected in the next 2-3 months and probably in the next month, it is spreading that fast. So this is kind of the last chance we have in order to lessen the medical impact of the virus. I do think we are on a path to herd immunity but like I said, it’s the worst, the worst possible path and there’s, I have to say, people, we talk with each other when situations get bad we share information, people start hearing about things that are going on and make changes to the life to avoid danger and what’s happening right now is there’s a whole bunch of people who aren’t gonna know what hit them. It’s coming that fast, and unfortunately, that’s what we’re dealing with in our hospitals.
Miller: Peter Graven and Dawn Nolt, thanks very much for joining us today.
Nolt: Thank you.
Miller: Dawn Nolt is a pediatrician at OHSU who specializes in infectious disease. Peter Graven is the Director of the Office of Advanced Analytics and Lead Data Scientist at OHSU.
Contact “Think Out Loud®”
If you’d like to comment on any of the topics in this show or suggest a topic of your own, please get in touch with us on Facebook or Twitter, send an email to thinkoutloud@opb.org, or you can leave a voicemail for us at 503-293-1983. The call-in phone number during the noon hour is 888-665-5865.