Why I'm Not That Worried About Latest Increase in U.S. COVID Cases (At Least Not Yet)

As we have lifted the lockdowns, we move closer and closer to whatever the "new normal" may be. Those have been clamoring for extended lockdowns worry that the lockdowns have been lifted prematurely. The New York Times dismally lamented about a gloomy picture. The Los Angeles Times analogizes the spread of COVID-19 to a forest fire. One of the main metrics that lockdown proponents have used to try to justify either returning to lockdowns or prolonging reopening provisions is that of increased confirmed cases. As we see below (figures extracted from Johns Hopkins website on July 5), the number of confirmed cases has reached 52,391. The percent of positive cases has also increased to 7.6% from its 4.4% trough. Dr. Anthony Fauci called last week "a very disturbing week" in terms of this increase. While it might make some intuitive sense to use number of confirmed cases as basis for whether the pandemic is getting worse in the United States (especially relative to other countries), I have some reasons to doubt that assertion.

• Confirmed cases are not indicative of total amount of infected individuals. One of the best features in favor of confirmed cases as a metric is that it is one of the earlier indicators within the infection timeline. Hospitalizations lag infections, and deaths lag hospitalizations. As nice as it might seem, it does not tell us as much as we would like. As a matter of fact, well-renown statistician Nate Silver wrote a piece in April about how coronavirus tests are actually meaningless. Especially at the beginning of the pandemic, the testing was prioritized for those showing symptoms. While the testing is still skewed in that direction, increased testing capabilities has allowed for more mildly symptomatic or asymptomatic individuals to get tested. Since the testing is not randomized, it suffers from a selection bias that makes COVID-19 look deadlier than it is. Over time, we see that the crude fatality rate (CFR) end up higher than the infection fatality rate (IFR), the latter of which being the apparent death rate. 
• On June 25, the CDC said that for every confirmed case, there are about ten people who had antibodies. At that moment, there were about 2.3 million confirmed cases, which means there were at least an estimated 23 million actual cases. 
• In case you did not have enough evidence that there are a lot more infected than we think, the Pennsylvania State University released an eye-opening paper late last month. This Penn State study examined influenza-like illnesses (ILI) surveillance data. After looking through the ILI data, they concluded that the initial infection rate was much higher. Rather than the initially estimated 100,000 new cases in the last three weeks of March, they estimated that there were actually 8.7 million cases, which implies an initial infection rate over 80 times higher than initially estimated. They also found that the number of cases also double twice as quickly as initially estimated (Silverman et al., 2020). 
• As Reason Magazine points out, even if we want to use CFR as a metric (although it is a poor one), the CFR has fallen from more than 6 percent on May 16 to less than 5 percent as of June 28 (see Worldometers data here).
• Demographic shift in who is getting infected. At the beginning of the pandemic, what we observed in the United States that it was those 60 and older disproportionately contracting COVID-19. Using Florida as an example, the median age dropped from 65 in March to 35 in June. On the whole, 43 percent of COVID-19 deaths in the U.S. took place in long-term care facilities. As the Heritage Foundation reminds us, the age demographics matter a lot when it comes to a serious illness. Younger adults are not immune from contracting a serious case of COVID-19, but the probability of a severe case or death for this demographic is much less likely. What has happened in recent weeks is that younger adults are accounting for a greater share of those infected. With more young adults contracting COVID-19, it is likely that the incident of severe cases and deaths vis-à-vis the IFR will be lower. While there is concern for younger adults infecting the elderly (which means we can have stricter protocols for long-term care facilities instead of another round of lockdowns), this shift is accompanied by other positive trends. 
• Decline in new hospitalizations. According to The Covid Tracking Project, which provides nationwide and state-level COVID data, there has been an increase in overall hospitalizations. At the same time, we have to be mindful of the number of new hospitalizations. Even when accounting for a two-week lag between infection and symptom onset, the CDC still shows an overall decline in new hospitalizations since mid-April. Looking at the CDC's interpretation of hospitalization forecasting, most of the models show a nationwide plateau of new hospitalizations in the upcoming weeks, although certain states (e.g., Arizona, California) are expected to see an increase. 

• Decrease in COVID-19 Deaths. It is more difficult to draw conclusions from the death data since it can take several weeks between infection and death. At the same time, what CDC data show us is that there has been a decline in all age demographics from the April 18 peak. 

• Our ability to treat COVID-19 has improved. Aside from adequate hospital capacity in most jurisdictions, we preliminarily have two treatments that show at least some promise: remdesivir and dexamethasone. We also have greater knowledge on how to treat COVID-19 in terms of treatment protocol (e.g., how to better use ventilators and their limits, prone positioning). I expect preparedness, palliative care, and treatment to only improve as time passes.

Postscript: Am I here to predict the future? No. I am aware the situation could change in either direction at any given moment. Am I here to say that the COVID-19 situation is under control in all 50 states? Again, no. It should go without saying that the situations are going to be different in each state. The impact of COVID-19 has been uneven geographically speaking. Early on, it was New York that got hammered. Now it is Texas, amongst other states, where hospitalizations are increasing. At the same time, Texas is not over capacity and is nowhere near where New York was at its peak (see press releases from Texas Hospital Association or Memorial Hermann, the largest hospital system in southeast Texas). Florida, California and Arizona are in similar situations, and they should improve protocols to make sure the situation does not get out of hand. 

What I do know is this. Although confirmed cases and hospitalization are increasing in certain states, there is enough reason to conclude that we are not at a stage where panic needs to be the response. There are way many more people infected than the confirmed case count suggests, thereby implying COVID-19 is not as deadly. The current caseload is younger than it was at the onset of the pandemic. We still have not seen the uptick in deaths or new hospitalizations over the past couple of months that lockdown proponents have feared. The fear-mongering never seems to end when it comes to COVID-19, but we should not let it get the better of us. We should use the data we have available to make evidence-based choices. While the number of confirmed cases are increasing, we do not see a negative shift in the indicators that would signal severity or lethality. Given the damage that the lockdowns have already caused, the last thing we should do is allow the media to scare us into another round of lockdowns.