New day. New press conference. In the U.S., we are being told that more tests are coming. Self-quarantine, social distancing, and hand-washing have been recommended. It seems that not everyone has been following the guidelines, scoffing at the suggestions, and going about their day as if tomorrow is promised. Friday night in Nashville, tourists flocked the downtown honky-tonks after the SEC basketball championship tournament was canceled the day before. They were ready to “have a good time” at the bars, which were “full and crowds filled Lower Broadway throughout the night,” according to the Tennessean.
Monday’s press conference with the WHO, followed by another from President Trump, reiterated the importance of social distancing and hand-washing with a stern warning that it wasn’t enough. More testing is needed according to Dr. Tedros Adhanom Ghebreyesus, Director-General of the World Health Organization. Testing and isolation are key to combating the spread of the coronavirus.
Are we testing enough to measure the true exposure?
The assumption has been made that there are not enough tests in the US to properly assess the true number of cases and deaths that are being reported. Predictive modeling of the pandemic is showing that it is possible the US and other countries do not have a true handle on the volume we are potentially dealing with yet.
In order to assess the situation within each country, we must look at the logarithmic curves formed by the data. The progression of a pandemic should be revealed with this type of analysis, giving a better understanding of its predictability. However, if there are any hiccups in data reporting, such as faulty tests or a change in reporting protocols, the statistical curves will reveal the dissonance.
Italians warn the rest of the world to pay attention
The situation in Italy is dire with high stress on their resources. Non-essential stores have been closed, and their movement is restricted in order to contain the spread of the virus. Their death rate is nearly 8%, which is higher than any other country we have evaluated. The people are trying to remain positive in spite of this, but they are imploring the rest of the world to pay attention to what is happening. Just 10 days ago, they were milling around as if they didn’t have a care in the world, similar to many in the US.
South Korea has a prolific testing program in place
Knowing that South Korea has been proactive in testing many of its people at drive-up stations all across the country, their data appears to be the most reliable to compare to. In this chart, we are looking at the logarithmic curve, which measures the progression of virus detection. The shape of this curve shows that it took a few days before mass testing was put into place. The lift in the curve between February 19 and 25 demonstrates that the virus has started its upward trajectory. Around March 7, the curve begins to stabilize and then move towards the right into a flatter line. The number of days between February 19 (the start of the lift) and March 7 (the stabilization point) is 18 days.
The next chart shows the number of new cases reported daily. These figures are lessening each day after the peak on March 3, which makes sense compared to the logarithmic curve in the prior chart where the trajectory began to flatten out on March 7, meaning that fewer and fewer new cases were being reported.
The next chart on South Korea to review is the logarithmic curve for the progression of deaths from COVID-19. Notice how the beginning is unstable, which is not uncommon when dealing with smaller numbers that give a wider variability in results for this type of analysis. We do not see an end to the deaths quite yet, but based on the number of new cases reported daily and the wide availability of testing in South Korea, we expect this curve to flatten and stabilize, meaning that the number of deaths reported each day will start to diminish.
What is happening in the United States with COVID-19 reporting?
Now that we’ve established what these statistical curves could look like, we can compare to what has been happening in the United States. This first graph of the US logarithmic scale reveals the testing and reporting hiccups that occurred in the first 10 days. The concerning piece, however, is that the trajectory is still on an upward continuum without any flattening. From February 24 to March 15, the logarithmic curve is still not showing the lift that we would expect to see once testing blips have been surpassed. This appears to indicate that we are looking at a much longer time until stabilization, or we have been grossly underreporting cases.
Next, we can see the daily new cases in the United States, which has not reached a peak yet, like the South Koreans have.
Finally, we are looking at the logarithmic curve of the number of deaths in the US. The first reported death is plotted on February 29. Knowing that our first cases were reported at least 10 days prior to this, it brings up the question, “Are we misreporting the number of deaths?” This could very well be possible. With the testing hurdles, we have been through to detect infection and the irregular reporting of the cases, the suspicion is that our death count is also understated.
How do the countries compare to each other?
In the following chart, we have data on the countries with the most cases being reported worldwide. The key figures to note in this chart are the cases per one million of population and the percentages of death. Notice how China has an infection rate of 56 people per one million in population. Italy’s is 37.5, and South Korea’s is 160.6. The United States shows only 13.8. If we believe that the United States’ cases and deaths are underreported, we can model the figures we know from other countries and apply that to the US to estimate.
In the next chart, we are adjusting the USA figures as if we had the same ratios of cases per one million population and the same death rates as the other countries.
For example: In order to estimate USA cases as if they were like China, we do the following calculations:
China’s cases per one million: 56.2
USA’s cases per one million: 13.8
Ratio: 56.2 / 13.8 = 4.072
Multiply the ratio (4.072) by the USA reported cases (4,570) to get the estimated cases if the USA cases were modeled to be like China.
Rounded estimate: 4.072 x 4,570 = 18,600
The estimated cases and deaths in the next chart vary widely and are all larger than the figures that the USA has reported thus far. Pandemics should follow a similar growth pattern in reported cases from country to country if the testing protocols were consistent. Death rates may vary based on age distribution, comorbidity distribution, climate, and other factors. We have not made adjustments for these types of differences in our calculations.
Selecting based on professional judgment
Now it is time to make a judgmental selection based on professional experience. We first took the average of the countries that began reporting cases around the same timeframe as the US, according to the Worldometer website. We also compared to what the US numbers would be like if we followed South Korea’s patterns, and then we averaged all the countries in our study excluding China. Ultimately, we made a judgmental selection of 28,000 cases and 700 deaths for the US. These estimates are made with point-in-time data and should be updated as more empirical data becomes available.
We will never know if our current estimates are close to reality. What’s done is done. However, we can take this information and use it to make sound choices in our actions going forward. Hopefully, the measures that are being taken will be enough to stem the tide.
Join our writing team and develop your writing skills, as you see your articles featured on Apple News, Google News, and all around the world. Subscribe to our newsletter, What Just Happened, where we dive deep into the hottest topics from the week!