Basically, positive tests for COVID-19 should beget more testing!
Image Credit: © Joe Raedle/Getty Images at Live Science
Previously we learned that the United States has conducted the most COVID-19 tests than any other country. But in reality, we haven’t performed nearly as many when you calculate the total tests performed against how many people populate the country, known as measuring our per capita.
Well, predictably, the States within our country also have varying testing rates given their respective populations. We know States have issued varying degrees of social distancing measures and ordered them at different times throughout the outbreak, which could have major implications in the number of cases and deaths.
This time, I wanted to see how the COVID-19 test rates have varied among the states, and if these test rates affected their burden of positive cases per capita. Specifically, are states conducting enough tests to fully understand the extent of the outbreak?
Before I get into this, I want to emphasize that there are many qualified reasons that could explain of lot of these data and statistics, many of which are not quantifiable. As such, some are discussed at the end of this post.
All of these data were generated from Worldometer and COVID-19 Tracking Project. The data are freely available to the public and super cool to play around with. Check them out! Also, I analyzed all 50 states and Washington, D.C. for this post. For simplicity, I am only showing 7.
Okay, let’s do this.
Question 1: What is the test rate per capita for COVID-19 in each state?
First we need to know how many tests the states are performing per capita. We can track their rate over time shown in Figure 1.
Even though New York has conducted the highest absolute number of tests, they still lead the pack by a wide margin when we compare those total test numbers to the state’s respective population — over 150 people out of every 10,000 in New York have been tested for coronavirus. In another state, consider Minnesota, where only 47 people out of 10,000 have been tested (27th/51). On the far end we have Oklahoma, an abysmal 7 people out of 10,000 have been tested for COVID-19 (51st/51).
Generally, it’s thought that more testing = greater understanding of the extent of the outbreak. From this vantage point, we would assume that New York is doing a better job than Minnesota in their testing, and especially compared to Oklahoma.
But what if New York has a really high number of positive cases? (Spoiler: they do). Taking this into consideration, is their testing rate still as impressive compared to other states? Is it necessary to test so many people ( >150 per 10,000)?
Question 2: What are the COVID-19 positive case rates per capita in each state?
To understand if a State’s testing capacity fares better or worse relatively to another’s, we first need to determine the positive case rate in each state. Predictably, as of April 5th, Figure 2 clearly shows New York State has the highest case rate per capita than any other state in the country at nearly 68 people per 10,000 infected with COVID-19. Minnesota? The fewest! Only 1-2 people of 10,000 are infected. Oklahoma? 33rd with just under 4/10,000 people testing positive.
Welp, what does this mean? Basically, positive tests for COVID-19 should beget more testing!
Just how much testing though? We can’t say for sure, but we can compare each state relatively to others. Let’s find out below:
Question 3: Is the test rate in each state giving the whole picture on how many positive cases are actually present? How do states compare to other countries?
Answering this will let us know how the states rank among themselves for their testing capacity. Again, I want to reiterate that this is a relative analysis. There are no standards that define specific COVID-19 testing capacities (although we might be able to determine that when this is done).
In Figure 3, we can see the rate of tests conducted per positive case on each day of the outbreak from March 4th to April 5th. A couple things stand out: First, the rate at which states are performing each day varies significantly. For example, on March 23rd, Hawaii all of sudden was able to start testing 5x more tests than their positive rate. Why? Did tests become available then? Did the criteria for testing change? I discuss some of these ideas below.
Interestingly, when we compare the rate of tests performed per capita compared to rate of positive cases (as of April 5th), we see that New York comes out as the 2nd worst. Minnesota on the other hand? 5th! Even though they only have tested the 27th most people per capita, their positive caseload is so low (the lowest in the country), that they presumably do not need to test nearly as many people as New York does in order to grasp the extent of the outbreak in their state — at least relatively so compared to New York.
Cool. So what does this mean?
Yeah what are the implications of this? This all seems pretty straight forward am I right?
First, this all comes down the numbers. Looking at testing rates alone (Figure 1), yeah New York has tested a lot of people. But after looking at how many people have been infected with COVID-19, it would probably be wise to try to test more to see just how many more people may be infected like other states have. So far though, we can only compare New York to other states. Time will tell how well each state implemented social distancing and performed tests to mitigate the impact of the virus. Eventually a standard will be established.
If we take a look at how well other countries tested for COVID-19 per capita compared to their rates of positive cases, we can get a better sense of just how well states in the U.S. are doing. Figure 4 depicts how other countries compare to the U.S. I chose to compare the U.S. with Italy, still reeling with their outbreak, and South Korea and Hong Kong, two countries that have begun to reopen their country.
Figure 4 shows the states average COVID-19 tests performed compared to their respective positive cases on the left, while the right side of the graph shows the rates for the country data.
**NOTICE: The y-axes are NOT the same, I repeat the y-axes are different scales comparing each group.**
Both South Korea (45 tests per positive case) and Hong Kong (108 tests per positive) imposed aggressive social distancing measures and the number of tests performed against every single positive case per capita. These implementations no doubt helped them control the spread of COVID-19.
The United States? Just 5.3 tests conducted per capita per positive case. Italy? 5.6. Both countries possess much higher case rates and death rates than Hong Kong and South Korea.
Final Thoughts
Again, there are a lot of reasons why some states have not tested more individuals. Why? Let’s look at New York. New York already knows it is dealing with a crisis. It needs to focus on making sure those that are sick are getting the best healthcare they can provide. It doesn’t matter if someone tests positive or not for COVID-19 if they are experiencing similar symptoms and are on a ventilator.
Second, states need to conserve personal protection equipment (PPE). This includes gowns, masks, gloves, etc. Taking nasal swabs to conduct a COVID-19 test takes away some of that valuable equipment that could be used elsewhere like in the ICU.
Third, it’s currently not recommended to go get tested, even if you have symptoms! Not anyone can just go and get tested. That breaks social distancing protocol, so states are bound to only test individuals that get really sick and must come to the hospital, thus skewing the data.
Lastly, states have different availabilities to testing kits and capacities to perform the tests. The tests require laboratory skill, if states are understaffed or do not possess the proper equipment they cannot complete the tests at greater rates than others.
These data were simply presented to demonstrate that we likely do not have any idea how many people are actually infected vs. uninfected, asymptomatic or symptomatic, or where/when states are going to reach that pivotal apex of the curve. That will take a nationwide effort utilizing an antibody test to see which individuals have recovered or not.
Earlier I mentioned the paradox that testing = greater understanding of the extent of the outbreak. That’s half true. In reality, testing = information. It may not be the most helpful tool illuminating just how bad this pandemic has become. But at the end of the day, it’s all we got. Testing gives us more information to make better-informed decisions to beat this virus.
In the meantime, as we sort through these unknowns, let’s keep on abiding by our social distancing; it is absolutely crucial that we do until we can figure all of this out.