The erosion of our country’s pandemic response resources started a while ago, when President Donald Trump named John Bolton his National Security Adviser in the spring of 2018. The appointment of Bolton, a famed war hawk who less than two months prior penned an op-ed in the Wall Street Journal outlining the case for a preemptive strike against North Korea (The Wall Street Journal, “The Legal Case for Striking North Korea First,” 02.28.2018), had implications that reverberated throughout the National Security Council and the White House as a whole. Upon assuming office, Bolton requested and received the resignation of HSA Tom Bossert (The Washington Post, “White House homeland security adviser Tom Bossert resigns,” 04.10.2018), who had called for a comprehensive pandemic defense strategy. A few weeks later, Timothy Zeimer, Senior Director for Global Health Security and Biothreats, resigned as well. He was the only National Security Council member focused solely on global health security and would have been in charge of a defense strategy such as the one Bossert touted. Bolton then disbanded the global health team Zeimer oversaw (The Washington Post, “Top White House official in charge of pandemic response exits abruptly,” 05.10.2018). How would we fare upon encountering a major pathogenic threat without these crucial officials?
We learned the answer to that question this January. The Center for Disease Control (CDC) confirmed the first case of the novel coronavirus on U.S. soil on Jan 21 (CDC, “First Travel-related Case of 2019 Novel Coronavirus Detected in United States”, 01.21.2020). Trump had replaced Bolton in September and Zeimer’s team was never put back together, yet Trump responded to this very real threat by saying on Jan 22, “We have it totally under control” (The New York Times, “A Complete List of Trump’s Attempts to Play Down Coronavirus,” 03.15.2020). While he did eventually bar U.S. entry to non-citizens who had recently traveled to China, many public health experts who called for expanded testing—including two former Trump administration officials, one of whom served on Zeimer’s team—were ignored (The Wall Street Journal, “Act Now to Prevent an American Epidemic,” 01.28.2020). Testing, or lack thereof, has proven to be our greatest weakness in fighting the virus. How can we know who to isolate if we don’t know who has it? The CDC’s first test kits suffered from a technical flaw (possibly lab contamination) that rendered its results unreliable (Politico, “How testing failures allowed coronavirus to sweep the US,” 03.06.2020). The Trump administration failed to fix the flaw quickly, acquire different tests from the World Health Organization (WHO), or allow private labs to develop their own tests (The New York Times, “A Complete List of Trump’s Attempts to Play Down Coronavirus,” 03.15.2020).
With the federal government’s failure to take action, responsibility for stemming the outbreak has fallen to the states. Referring to medical equipment such as ventilators, Trump told governors to “Try getting it yourselves” (Business Insider, “‘Try getting it yourselves,’” 03.16.2020). On March 16, the Food and Drug Administration (FDA) announced a regulatory change that gave states the option to approve testing methods developed by laboratories within their borders; otherwise, they would be slowed by bureaucracy and continue to face the failures of the federal government (Reuters, “FDA giving states authority to approve own coronavirus tests, official says,” 03.16.2020). Although a state-level response helped overcome the shortcomings of the federal government, this approach gave rise to inequalities in testing capacity and medical response based on the efficiency of each state’s government and each state’s resources.
Shortly before the outbreak worsened, QuoteWizard recruited their health care analysts to sift through reports and rank each state in terms of its preparedness for public health and natural disasters (QuoteWizard, “States Most Prepared for Public Health and Natural Disasters,” 03.05.2020). This was how I chose to measure the inequalities in preparedness among the states. Their rankings were based on six factors, and I analyzed how all the factors, as well as overall rank, predicted testing capacity (Kaiser Family Foundation, “State COVID-19 Testing,” 04.02.2020), recovery rate (Worldometer, “United States Coronavirus,” 03.28.20), and cases per capita in each state (World Population Review, “US States – Ranked by Population,” 03.28.20) in multiple linear regressions. While the rankings were only available for the 50 states, the testing capacity data also included information on Washington D.C. and Puerto Rico (Kaiser Family Foundation, “State COVID-19 Testing,” 04.02.2020).
To examine each states’ testing capacity, I examined the variable of tests administered per 1000 people. The only factor that proved relatively significant (p < 0.05) in estimating testing capacity of each state was their water security; having a higher percentage of communities with unsafe water seemed to correspond with a higher number of tests taken. This didn’t make sense, until I realized that New York state, surprisingly, had the highest proportion (45 percent) of communities with unsafe water. Besides Rhode Island (38 percent), no other state had a proportion over 16 percent, making New York and Rhode Island serious outliers. After removing the two from the dataset, water security was no longer significant.
For recovery rate, which I calculated as the number of recoveries divided by the number of closed cases, the results were more interesting, but still confusing initially. When I first examined the data on March 27, whether New York was included or not, higher flu vaccination rates corresponded with significantly higher recovery rates, which makes sense—vaccinations are indicative of superior public health mobilization. Strangely, a higher percentage of hospitals given an “A” grade by Trust for America’s Health corresponded with significantly lower recovery rates. I attributed this to small sample size; only eight states had 20 or more recoveries on record. The multiple regression model mistakenly saw this result as significant, whereas it probably would not have if there were more data. As I predicted, upon reexamination a week later (April 2) with more data available (18 states with 20 or more recoveries on record), the only factor that remained significant was flu vaccination rates, which still corresponded with higher recovery rates.
A higher percentage of hospitals with an “A” grade also initially (March 27) corresponded with more cases per capita. While this result still remained relatively significant upon reexamination (p < 0.10), its significance had decreased, and I expect it to only decrease more as more data becomes available.
For the recovery rate data, 18 states are still a minority, and the initial confusing results show the limitations of small sample sizes. What can we know in the face of such limitations? Well, one of the reasons that we have them is because of the dismal testing capacity in this country. Many people have probably had the virus and been asymptomatic; others have thought it was a passing cold. Both of these types of people would have counted as recoveries and in the cases per capita calculation, but they have instead gone untested. Yet, some states are administering significantly more tests than others. Is there anything contributing to this discrepancy?
One of the most significant relationships I found was the correlation between per capita cases and tests per 1000 people; they were highly correlated (p < 0.005, R2 = 0.378). I also found an extremely strong correlation between total cases and total tests administered (p < 0.005, R2 = 0.851).
This is not surprising; the more people are tested, the more cases we should expect to see. However, this result also made me wonder: could it be that it works the other way around? Maybe, the more people are symptomatic, the more tests will be administered. This might suggest that states are ramping up testing too late, once the virus has already spread within their borders.
To test this theory, I measured the correlation between the proportion of tests that came back positive with total tests administered. I excluded the three U.S. territories (Guam, Northern Mariana Islands, and the U.S. Virgin Islands) which tested fewer than 1000. All other territories and states tested at least 1900. I was left with another strong correlation (p < 0.005, R2 = 0.157), indicating that states which test more also have a higher proportion of those tests come back positive. If this is indeed the case, it could provide more evidence for my theory that states are only beginning to test more once the virus has become rampant within their borders. Thus, more tests confirm more positives.
However, this correlation also lends evidence for another position: maybe more tests mean more positives because a larger percentage of the population is infected than we think. Many people may be asymptomatic or have mild cases often mistaken for common colds—experts have said that these people may account for up to 80 percent of the infected population (The Washington Post, “Trump extends social distancing guidance until end of April,” 03.29.20). States, with their limited resources, have largely reserved testing for the most severely ill. Once we expand testing beyond this group, even more positives may show up. If we have only identified around 20 percent of current cases, it would mean the actual number infected is more than half a million.
This is a scary thought. However, it is one that we can and should address. State governments must take it upon themselves to ramp up testing, even in states where few people have complained of symptoms. Increasing the scope of testing may reduce the correlation between the proportion of positives and the number of tests administered. Most importantly, more tests will help identify cases, provide more accurate fatality rates and in general grant us more information about the novel coronavirus. In the meantime, we should continue to support one another; even from afar, the warmth of the Vassar community has taught me to never underestimate the healing power of even the smallest of gestures.
Unless otherwise specified, all data is as of April 2.