It Will Get Worse Before It Gets Better

For the past two months, I have been warning of a likely 4^th surge of coronavirus cases in Idaho, likely to occur by late March or sometime in April. I have been pleading with our local school districts not to bring back more students to school than can be successfully distanced given the CDC’s prediction that the UK (B.1.1.7 – Kent) variant would be dominant in the US by the end of March. Obviously, it doesn’t make sense to decrease the precautions you are taking when you will be dealing with a new form of the virus that is 40 – 70 percent more contagious, and reportedly causes more severe disease.

It seems that my warnings have largely been ignored and quite surprisingly, one school district decided that my admonitions were inconvenient, leading the board to determine that they would find another adviser who could be more supportive of their plans to bring back all students without proper distancing.

Even more surprising was the fact that the school board never asked me to explain why I was so concerned, nor the even more obvious question, if they were intent on disregarding my warnings: how can we bring all the students back in the safest possible way given the increased risks? I had offered suggestions last fall as to ways we could safely bring back more students and still maintain six feet of distancing, but no one seemed interested, and now it would be much more difficult. Had we done this planning last summer, we likely could have safely had students for full in-person instruction for most of the school year.

So, I thought that I would take this opportunity to explain to my blog followers why I have been predicting this 4^th surge, why I think the evidence has only grown stronger over the past two months to support the likelihood of a 4^th surge and also refute some of the reasons given by others as to why they don’t believe such a surge will occur.

1. Let’s start off with looking at the epi curve for the US (see below).

To understand what I mean by a fourth surge, look at the graph. The first surge began on March 12, 2020. The peak of the first surge came on April 9, 2020 with a 7-day moving average of daily new cases of 9.9 per 100,000 population.

Let’s take a moment to give perspective on what these numbers of new cases mean with respect to the degree of disease transmission. People often ask me when we will achieve herd immunity. I don’t know, other than I know it is not going to be April as a well-known physician author wrote recently (see my prior blog piece for an explanation as to why that is completely unrealistic.) However, I am pretty sure we won’t get there in 2021. While everyone wants to know the percentage of the population that has to have immunity to achieve herd immunity (and no one knows the answer to that), I tell them one way to know when we get there is to look at the 7-day moving average of daily new cases. It should be <1 daily new cases per 100k and remain that low for months. We are nowhere near that.

Community spread is when the level of disease transmission in a community is such that we no longer can identify the source of infection. Early on in the pandemic, there was a period of time that we could identify an exposure from travel or an infected person. When we get to the levels of disease transmission associated with community spread, we no longer can reliably pinpoint sources of infection. Using the metric of the 7-day moving average of daily new cases, community spread is 1 – 9 daily new cases per 100k population. So, you can see that at the peak of the first surge, we were at the highest level of community spread.

When the 7-day moving average of daily new cases per 100,000 is in the range of 10 -24, we refer to that as accelerated community spread and when > 25, we refer to that as the tipping point – generally the point where public health experts would call for serious mitigation measures, e.g., stay-at-home orders, closure of certain businesses, etc.

Now, notice something concerning from the epi curve above. We never came down to that well-controlled level of disease transmission of < 1. The lowest point following the first surge occurred on May 28, 2020 when the 7-day moving average of daily new cases per 100,000 was 6.5 – community spread.

We began the second surge from this elevated level of disease transmission. In other words, we loosened up our restrictions too soon. If you recall, people were celebrating the cases coming down, began getting together in larger groups without masks, believing the worst was over, and surely as we moved into the beginning of summer, the virus would go away, though many of us were warning that there was no evidence that there would be seasonality to this virus, and even if there was, it likely would be back in the fall with a vengeance.

So, the second surge was under way at the end of May/beginning of June and peaked July 22, 2020 at a level of 20.3. Notice that this second surge was higher than the first, and like the first, we again did not come down to levels of disease transmission that were under control before we began the third surge. In fact, the lowest point after the second surge was on September 12, 2020 when the 7-day moving average of daily new cases per 100,000 was 10.3. Note that this low point was higher than the low point following the first surge and represented accelerated community spread.

Having this springboard of accelerated community spread resulted in an even higher third surge that peaked January 8, 2021 with an unbelievably high disease transmission rate reflected by a 7-day moving average of daily new cases of 78.3 per 100,000.

Once again, as happened with each of the two prior surges, cases came down from the peak, and I began hearing the all-to-common refrain by now that things were returning to normal (how soon people forget), but coupled by the even more concerning actions by elected leaders across the country to loosen and often terminate all public health measures in place to contain the spread of the virus.

But, while others saw reason to cheer, I became alarmed in the third week of February when I saw the sharp decline begin to plateau, and like the first two surges, at way too high a level of disease transmission – the lower range of accelerated community spread and the higher end of the range for community spread. I had already seen this movie twice and I knew how it turns out.

• So, now let’s look at Idaho’s epi curve:

Idaho, too has experienced three surges, but there are a few differences. Our first case was on March 13 and our first surge peaked on April 4, 2020 at 6.5 average daily new cases per 100,000. But by May 13, we were down to average daily new cases of only 1.3, almost at the containment level. The difference from the US disease transmission was, due to the brave actions of our governor, we had a stay-at-home order in place and it worked. Recall that the U.S. was almost 6 times higher than this level of disease transmission. We know that the governor’s actions saved lives.

Our second surge took off in the middle of June and hit a peak on July 19 at a level of daily new cases of 31.9 per 100,000, higher than the U.S. peak for the second surge. Unfortunately, our low point after the second surge was only 13.4 on September 13. This time, without the stay-at-home order and with COVID-fatigue already setting in, like the rest of the country, Idahoans began relaxing their compliance with public health measures too soon.

Our third surge would begin almost immediately and peaked on December 10 with a 7-day moving average of daily new cases of 91.7 per 100,000. Although cases declined significantly after hitting this peak, as they had in every prior surge, again, I became concerned in mid-February when our state cases hit a plateau and remained in the community spread range, even while many were celebrating a return to normal and our elected leaders were saying all kinds of ridiculous things, like this did not even qualify as a pandemic and there was no longer any need to wear masks, physically distance or restrict the size of gatherings. Of course, in an interesting twist of karma, the legislature then found itself with an outbreak of COVID in the Capitol and as a result had to interrupt the session to recess in order to stop the spread of infection among lawmakers and their staff members — a reminder that we can deny COVID, we can spread lies about it, we can create alternative facts, we can find some doctor out there to tell us what we want to hear that has already been debunked by the medical community, and we can embrace conspiracy theories, but that merely plays into the virus’ hands resulting in many needless infections, as science will always prevail in the end.

While the U.S. and Idaho epi curves explain why I was concerned about a possible fourth surge by mid-to-late February, why was I concerned even before then that we would experience a fourth wave?

There have been two times during the pandemic that we could get a preview of coming attractions. The first was at the beginning of the pandemic in January of 2020 when we could look at the outbreak in Wuhan to see what we might eventually be dealing with in the U.S. – cases of pneumonia that seemed to disproportionately land older people in the hospital. The second time has been since December 2020 in Europe, particularly the U.K., as the B.1.1.7 (Kent) variant quickly evolved over the course of just two to three months to become the dominant strain in the U.S. causing an increase in cases in all age groups and a new surge that was far greater than anything the U.K. previously experienced and was really a surge upon a surge. Let’s look at the U.K. epi curve:

The first surge in the UK peaked on April 24, 2020 at a 7-day moving average of daily new cases of 7.1 per 100,000. But, unlike the U.S., the U.K. successfully brought their disease transmission down to below 1 by July 4, 2020 and it remained low for weeks.

By August 6, 2020, cases began to very slowly rise. Then a sharp rise in cases began at the very beginning of October. At the end of October, the U.K. implemented a lockdown, but allowed schools to remain open. Cases peaked on November 16 at a 7-day moving average of daily new cases of 37.3 per 100,000. By December 4, 2020, cases had come down to a low point of 21. Keep in mind, this level is still in the accelerated community spread level of disease transmission.

But something else was happening under the radar. A new variant had emerged in the U.K. just outside of London in the village of Kent in October. This variant came to be known as B.1.1.7. We saw that it only took 2 – 3 months for B.1.1.7 to become the dominant circulating form of the virus in the U.K. While cases were declining from U.K.’s second surge, the proportion of isolates from people infected that were B.1.1.7 were increasing. What happened next? A new, much larger spike in cases began reaching a new peak on January 10, 2021 with a 7-day moving average of daily new cases of 87.9 per 100,000, despite a lockdown in November and a new lockdown in January. The difference? One difference was that the cases in this most recent, higher spike were mostly B.1.1.7. Another difference was that during the November lockdown, schools were kept open.

The U.K. had previously had the same experience with schools that we had in the U.S. Before the variants, we saw few infections in kids and little spread within schools. So, the U.K. locked down essentially everything but schools at the end of October. However, it soon appeared that the transmission characteristics of B.1.1.7 were different than those of the variant we have dealt with around the world for much of the last year. The U.K. started seeing more infections in children and more outbreaks in schools and daycare. And, lest people think that this was something unique to the Brits, we were also beginning to see the same thing happen in Italy, Germany and Denmark where B.1.1.7 was growing in the percentage of isolates causing infection. Then an interesting fact emerged from the data. While B.1.1.7 tended to increase infections in all age groups, the age group with the highest increase in secondary attack rates (i.e., rates of infection of others who are infected) was 30 – 39-year-olds. Why would that be? In the U.K., adults of that age range should mostly be working from home due to the lockdown. My speculation was that this is likely the age of the parents of children in school. In other words, my concern was that they are having in-school transmission among children, many of those infections might have been mild or asymptomatic and not detected, yet the children were going home from school and infecting their parents.

Obviously, at the beginning of February, we still didn’t have all the answers, but this was enough for me to sound the alarm with our local school districts, but one that fell on deaf ears. As one board member told me as I pleaded with the board not to make a decision in February to bring all children back at the end of March when we knew the best physical distancing we could achieve in schools with students in the hybrid model was 4.5 feet and when I assured them that we would know much more about the variants and how vaccinations were going in Idaho by the end of March: “we don’t care about what might happen then, we are just looking at what is happening today.” I was shocked.

Further myopic was the board and administration’s assessment that things would be fine because children seldom get serious illness and by the end of March, teachers would have had an opportunity to be vaccinated. I tried to point out that students and teachers are not the concern based on what we were seeing in Europe. It is the parents in their 30s and 40s who were not expected to be vaccinated in Idaho by the time they were bringing all students back just as the CDC was predicting that B.1.1.7 would be the dominant circulating form of the virus in the U.S. This didn’t seem to influence them in the least, I presume because they assume that 30- and 40-year-olds don’t get seriously ill. Again, I tried to explain that studies were coming back suggesting that B.1.1.7 was 40 – 70 percent more transmissible and early reports out of the U.K. were that this variant caused more severe illness. I also tried to point out that there are many parents in this age range who have serious underlying health conditions that place them at increased risk, and there are many who are taking chemotherapy for cancer, have immune deficiencies or are taking immunosuppressive medications and may not do well if they become infected with a more transmissible and virulent form of virus. I also tried to point out that some students, especially those in socioeconomically disadvantaged situations may be living in multi-generational households with older family members who may not be vaccinated, and will be at very high risk due to age and potentially underlying medical conditions. It was if I was talking to myself.

By this time, it was not only the UK, but Italy, Germany, and Denmark that would give the world warnings that B.1.1.7 transmission characteristics were different. While children did not seem to get sick as often as adults and did not seem to spread infection within schools or contribute to community spread prior to the rise in B.1.1.7, the countries and others were now indicating that there seemed to be significant spread in schools that was contributing to community spread and these countries were now closing schools. They warned the world that B.1.1.7 is more transmissible, causes more severe disease and that it is much more difficult to contain its spread. The U.K. warned that maintaining physical distance was the only hope to controlling the spread of this variant.

So, for all these reasons, I was very concerned that this was exactly what would happen in the U.S. Why would we be different? Some dismissed my concerns believing that what happens in Europe has no bearing on what will happen in the U.S., let along Idaho. Then it started happening in Canada. That still did not persuade people. But, recently, it started happening in the U.S. While I had hoped that I was wrong and overreacting, all the evidence kept affirming my fears. So, let’s look at a few examples of this additional evidence.

1. Outbreak of B.1.1.7 infections in Minnesota in late January into February 2021. From the Minnesota Department of Health – Since late January, at least 84 cases of COVID-19 have been linked to participants in both school-sponsored and club sports activities, including hockey, wrestling, basketball, alpine skiing, and other sports. In addition, health officials have seen increases in cases in Carver County gyms and fitness centers, with many of the cases linked to the sports-related cases. The Minnesota Department of Health (MDH) has observed a concerning increase in the rate of growth of cases in Carver County over the past month (a 62% increase between the week beginning Jan. 27 to the week beginning Feb. 24). Case rates in Carver County are now approaching rates seen in October 2020, and a high proportion (35%) are under age 20.

“We are increasingly concerned about dramatic increases in cases, particularly in Carver County where the B117 variant has been confirmed.”

Through recently completed whole genome sequencing of a portion of the sample specimens from the outbreak cases, health officials confirmed 27 cases of the B117 variant, a highly contagious and potentially more virulent variant first identified in the United Kingdom. Cases have occurred in athletes, coaches, students, and household contacts. Multiple schools, both public and private, have confirmed cases tied to the variant strain cases. Many of the people with B117 variant COVID-19 attended school or sports activities while infectious. The concern is this highly contagious variant has the potential to spread exponentially and quickly across youth who are largely unprotected by vaccines.

• In one week in early March, Duke University had an outbreak of COVID cases among students. The number of new cases in that one week exceeded the total number of cases that occurred during the entire Fall semester.

• We have seen a large outbreak of new COVID cases in Eastern Idaho. Looking at just one of the involved counties, Bonneville, the peak occurred on March 14 and reached a 7-day rolling average of daily new cases of 72.8 per 100,000.

• Although we do not do enough sequencing in the United States or in Idaho to know what percent of cases are currently caused by B.1.1.7, we do know that B.1.1.7 is significantly increasing and has become the most frequently identified variant of concern in samples that are sequenced. European countries have warned that once B.1.1.7 surpasses accounting for half of the circulating virus, we can expect a soon upcoming surge.

• The testing positivity rate in Idaho had fallen to 4.5% by the third week of February, but has increased to 5.5% by mid-March.

• What is happening now?

1. New cases in the U.S. are up 20% from the prior two weeks.
   1. Cases are in the accelerated community spread category, increasing and staying high in 28 states plus the Virgin Islands and Washington D.C.
   1. Deaths are increasing in 8 states plus Washington D.C.

I could keep on going with examples, but let’s look at one more local indicator that I find fascinating.

This is the wastewater testing for SARS-CoV-2 virus done by the City of Boise. It shows that we got to a low on February 14 of about 33,000 virus copies per liter, but it has been increasing since then, signifying more virus (i.e., more infections). By March 17 it had peaked at about 293,000 virus copies per liter. As of the last date of data available (March 25) the count was about 160,500, i.e., almost a 5-fold increase in viral levels from the low just the month before.

So, I will stop here and address some of the counter-arguments. But, hopefully you can get a sense of why I have been predicting this upcoming 4^th surge.

Arguments:

1. The vaccine roll-out will stem this surge. My response – I hope so, but don’t think so. We simply do not have enough people vaccinated to get us anywhere near herd immunity. For more on this, see my prior blog piece.
2. We don’t have to worry about Boise or West Ada schools because the CDC said we can operate schools safely with only 3 feet of distancing. Response – Wrong. The CDC has indicated that 3 feet of distancing is not recommended for schools in communities with the highest levels of disease transmission (which is what the CDC considers Ada County to be in) and the CDC indicated that in all cases there should always be at least 6 feet of distancing when students have masks off (for example, the cafeteria), but in all the schools I inspected, I never found a school that could get more than 4.5 feet of distancing in the cafeteria with only half the students in attendance.
3. We have been able to prevent transmission among health care workers in hospitals where they often work less than 3 feet apart. Response – Two big differences. First, the majority of health care workers have been vaccinated. Almost no children have been vaccinated. Second, this was the experience with the wild-type virus, not B.1.1.7. In fact, one of the Duke hospitals just reported an outbreak of COVID cases on a cancer unit involving 20 staff and patients.
4. We don’t have to be concerned because it has warmed up and everyone will be outside and so we won’t have much spread of the virus. Response – (1) There is absolutely no evidence of seasonality of this virus. (2) Why do you think that will be the case this year when it wasn’t last year? Note from the epi curves for the US and Idaho above that the surges in the summer were far greater than the first surges in March/April of last year. People don’t spend all day, every day during the warmer months outside. In fact, in many places in the country, the heat and humidity cause people to move indoors during the middle of the day for air conditioning.

Well, we will know soon enough who is correct. I pray that I am wrong, but as I review the activity around the world and in the U.S. and even here in Idaho, I only become more and more convinced that I am right. We will know soon enough.