Why aren't we seeing more deaths with the recent COVID-19 surge?

The past few weeks have seen an alarming surge of incidence of SARS-CoV-2 in Arizona, Texas, Florida, and South Carolina, with upticks threatening surges in other southern and western locations. However, these surges in incidence have not produced a corresponding increase in COVID-19 related deaths. This has led some to suggest that we should not be so concerned about this surge, or that the virus has somehow become less virulent and less dangerous. Let's investigate the data and think about what factors might be contributing to this effect. I was interviewed by Bloomberg Health last week for an article they just published yesterday -- some of these ideas are in that article, but I elaborate on some of my points here.

We can see that while cases sharply surged upwards in June for all four of these locations, there is little evidence of any surge in deaths, outside of an uptick in Texas starting 2nd week of June and in Arizona over the past week.

I can see several factors that could explain this apparent discrepancy:

1. Time Lag Effect

2. Demographic Effect

3. Treatment Effect

4. Virulence Effect

5. Evolutionary Effect

Let's consider them each one-at-a-time and then assess what the most likely driving factors are:

Time Lag Effect

The simplest explanation for this is that we just haven't waited long enough to see the deaths occur. We know that case counts lag 7-10 days behind day of infection, hospitalizations lag maybe another week later, and deaths lag perhaps 4-5 weeks behind infections. Since this surge has just happened in the past 3-4 weeks, it stands to reason that it has not yet produced an uptick in deaths. It is possible that the case death rate of about 5% observed around the USA will also be observed during this surge, and will become evident over the coming weeks.

If we look at the case death rate, the proportion of confirmed cases resulting in deaths, we see that this value peaked in mid-tune around 5.5%, and has subsequently decreased just below 5% throughout the month of June. This suggests that we should expect a case death rate around 5% for this surge, as well, but it might be decreasing over time.

Demographic Effect

Another simple explanation for this effect is that recent infections have a higher proportion of younger people, in their 20's and 30's, who have far less risk of severe disease and death relative to the older people comprising a higher proportion of April and May infections. The increasing proportion of infections in younger age groups is well documented, and may indicate an age group relaxing social distancing more aggressively and perhaps not wearing masks as much as older individuals. The previous case death rates also may have been inflated by the large number of deaths from infected individuals in nursing homes, estimated to comprise perhaps over 40% of the COVID-19 deaths in the United States to date. If this is a key factor, it is unlikely to produce a lasting reduction in COVID-19 related deaths, as it is expected that the young infected will tend to spread it to others in the community including others who are older or with pre-existing conditions that put them at higher risk for severe disease or death.

Treatment Effect

Early in the pandemic, there was very little known about the disease course, potential treatments, and what life-threatening complications were posed. This made earlier patients with severe disease perhaps more prone to death. Although no strongly effective antiviral treatment has been found yet, advanced COVID-19 cases are now known to be characterized by overaggressive immune and inflammatory responses that put the individual at risk for various life-threatening complications including acute respiratory syndrome (ARS), embolism and stroke. Learning this, hospitals have become more adept at managing these severe cases and thus have been able to prevent more deaths. It now appears that the early use of ventilators was a mistake, and now alternative methods including lying patient on their stomach and giving steroids like dexamethasone and high dose Vitamin C and Zinc have helped, as well as other anti-inflammatory and immunosuppressive treatements.. This is another factor that might be influencing the decreasing case death rate.

Virulence Effect

A report out of Italy in early June created a stir when it mentioned that viral swabs showed a substantial decrease in viral load experienced by cases in late May vs. earlier in the pandemic. It was speculated that the infected cases might have lower viral loads, which might suggest less severe disease and fewer deaths. This report was not scientific, but based on anecdotal evidence, and so was met with a good deal of skepticism in the scientific community. It is possible, however, that viral load in infected subjects is less than it was earlier in the pandemic. And it has been observed that patients with severe COVID-19 shed virus longer and in higher copy numbers. It is known that the virus spreads less efficiently in heat and humidity, so it is possible that this weather factor reduces the viral load to which an individual is exposed, not enough to prevent infection, but perhaps leading to disease with lower viral load and possibly less severe. Similarly, it could be that the mitigation strategies of social distancing and mask wearing are reducing the viral load to which an individual is exposed and also producing less virulent infections that might not have as high of a risk of death. I have not seen rigorous scientific studies demonstrating that in fact we are seeing less virulent disease, and if so, whether it is due to these factors. So the idea that current disease is less virulent remains speculative. If this is the case, this would be mixed news -- if mitigation strategies continue to be diligently practiced, we could see a continuation of this effect, but if driven by the weather, then this effect could be reversed in the fall as the weather gets cooler and drier.

Evolutionary Effect

It is possible that virus has evolved into a less virulent strain that is likely to produce severe disease. We know that all viruses mutate as they spread, which is what makes development of a vaccine somewhat of a moving target, e.g. requiring new influenza vaccine formulations every year. Although SARS-CoV-2 seems to mutate less often than many other viruses, it has been observed to mutate, and some extensive efforts have been done to sequence the RNA for thousands of individual viruses during this pandemic, an effort led by Nextstrain.org. This has led to the realization that most of the USA COVID-19 cases have been caused by a strain of the virus from Northern Italy in Europe, entering to the USA through New York City, rather than the original Wuhan strain. Follow-up studies have found that this European strain, called the "G" strain, has a mutation on the spike protein part of the virus relative to the original Wuhan "D" strain. The spike protein is what allows the virus entry to take over host cells, and the "G" mutation has been shown to enable the spike protein to operate much more efficiently than the "D" strain, leading to a more infectious virus and accelerated spread. It is this faster-spreading strain that has over time completely dominated the original strain in USA SARS-CoV-2 infections.

Here is a figure from an article in the Washington Post that demonstrates the shift towards this strain.

Now IF this increasingly common strain led to less severe disease, then it is possible that the virus is shifting to a less aggressive variant that will not lead as many deaths. However, to date, there is NO evidence that this strain is somehow less severe, and thus this conclusion is unfounded. However, it is worth tracking the evolution of the virus over time, since it is plausible that natural selection will favor less aggressive strains, if such exist, since more aggressive strains that put more of its hosts in the hospital or the grave might have a survival disadvantage over strains that remain mild but still spread asymptomatically. Right now, however, there is no evidence of this.

Conclusions

So, why are we not seeing more deaths during the current surge? The answer is probably a combination of these reasons, but it is likely the first two factors explain most of this effect. In the next few weeks, unfortunately, we are likely to see a certain percentage of cases from this surge produce deaths, and if this proportion is a little lower than we have been seeing the discrepancy is likely to be explained by the increasing proportion of young adults among the cases. However, it is possible that hospitals are getting better at treating the proportion of cases that become severe, thus reducing the death rate, and it is also possible that either environmental or genetic factors have made the virus less virulent, but this is at best speculative at this point.

One thing that is CLEAR is that this is NO REASON to let our guard down or to downplay the seriousness of this current surge. If we allow this virus to spread unconstrained, it is clear that it spreads like wildfire and can easily overwhelm local healthcare systems and produce an even greater proportion of mortality and morbidity than is necessary. So practicing vigilance to reduce the spread of the virus through social distancing and mask-wearing remains key in our management of this virus, and if the public would faithfully follow the guidelines comprised of just a few basic precautionary steps, the spread could be kept under control and such surges prevented.

Regardless, the case death rate over time bears watching, as it affects the severity of disease and may influence mitigation strategies over time. We will have to keep watching the data and emerging scientific studies to clearly understand the factors leading to this apparent effect.