Early on in the crisis President Trump told the American people: “Anybody that wants a test [for the coronavirus] can get a test.” This statement was wrong on multiple levels. Most concretely, there were not nearly enough tests available nor laboratory facilities available to process the tests. In fact, five weeks after his March 6th statement, access to testing remains problematic for many. But more importantly, the statement reveals a lack of understanding of the purpose of testing in the first place. Promising the availability of coronavirus testing is not like a campaign slogan such as promising a “chicken in every pot”. Its not a measure of economic freedom, such as having a competitive market for cell phone service. Testing is the single most important tool for identifying the scope and spread of the disease. It is also the most powerful tool for informing public health programs for controlling and containing an infectious disease outbreak.
With over 2.5 million coronavirus tests administered, we still have a very poor understanding of the scope and spread or the virus, nor have we mounted any serious concerted efforts to leverage testing to control and contain the outbreak. Instead we have a hodgepodge of testing being performed by a broad range of organizations, using a variety of criteria for testing, and driven mostly by patient demand, rather than by public health strategy.
How Individual’s Decide to be tested:
For each of us, the decision to seek testing is fraught with confusion and conflict. There is the huge logical paradox: If we go for testing, but are not infected, the very act of seeking testing represents a relatively high exposure risk. In these days where we limit our trips to the grocery store and increasingly don a mask to do so, going to a medical facility where coronavirus testing is being performed may actually be a risk factor for infection. For those out of work, furloughed, or working from home, knowing your infection status may not be particularly meaningful. We are already instructed to stay home and limit our social interactions. Though faster tests are reportedly on the way, most testing still takes several days to get a result. During that time its likely any symptoms that provoked testing in the first place will have resolved, or perhaps escalated to the point of a serious illness requiring a trip to the hospital. In either case, the test result is no longer relevant.
For those who are still going into a work place where they interact with others or those who are care takers of high risk family members, infection status becomes more important. Though if you are symptomatic, you should behave as if you are infected and not go to work. So again the value of a test result to an individual is of limited use.
There is also a darker side to be considered. Those who are economically vulnerable may see 14 days of isolation as a serious threat to their financial stability. Combine this with the vagary of symptoms makes it easy to allow your fragile financial situation to interfere with the thought process regarding mild symptoms. Is that scratchy throat just a normal seasonal allergy, or a coronavirus symptom?
One of the places testing should have been used, was to monitor the nation for community based transmission. Historically the CDC has excelled at this kind of surveillance. This entails getting testing protocols established in “sentinel” clinical sites around the nation very early on. The CDC would work with 50 or 100 clinical sites around the nation to consistently test for coronavirus based on a protocol. A protocol could be as simple as the first 10 patients visiting the clinic each day. Or it could have been more specific and targeted, for example, testing all patients presenting with flu like symptoms who test negative for this year’s flu. The flu test only takes a few minutes and results are available while the patient is still available.
Not only did the federal government and CDC fail to establish surveillance, they actively worked to prevent surveillance. In Washington State a flu research study had collected samples suitable for coronavirus testing over several months. When the lead researcher sought permission from the CDC to have the samples tested for coronavirus to see if and when coronavirus had started in the region, the request was repeatedly denied. Eventually, the researcher, disregarded the CDC directive, and performed their own coronavirus test on the samples, and found multiple positive tests. The researcher sought permission from the institution’s Data Safety and Monitoring board which agreed with the researcher that failing to test that samples would in fact be un-ethical given the current health emergency. This kind of surveillance in the Washington area could have had a very concrete impact on lives. Had people known about the circulation of the virus earlier, rather than waiting until deaths started at the Life Care Center its quite possible that many lives at the Life Care Center could have been saved.
In another example of local authorities making good use of testing for surveillance purposes, the State of Florida will take over testing sites being cut by federal authorities, to test 800 non-symptomatic state residents each day, as part of an effort to get a better understanding of the scope of the infection rate in Florida.
Systematic testing of samples of the population who are not showing any symptoms is one of the few ways of getting insights into how wide spread the infection is. Currently most data available for the surge projections are based on the non-random, non-systematic results of all tests performed in a state. This kind of data is in no way representative of the underlying infection rate in a community. In most cases it will over represent infection because testing tends to focus on those at high risk, health care workers for example, those with symptoms, or with known exposures. It is correct to test those at high risk, but we need some samples of more representative people, to get insights into the scope of the infection. Tracking and reporting of results should probably be reported separately rather than state by state aggregated totals. There have been criticisms of the various models being used to project the coming surge. These have pointed out that the wide changes in estimated deaths indicate the models are meaningless. It is more likely that the models represent a valiant attempt to make estimates based on our non-systematic testing policy.
Another important purpose for testing is to support traditional epidemiological containment strategies. This labor intensive process involves tracking each infected person, supporting their need for isolation, while contacting all people who may have been in contact with the infected person and having them quarantine themselves and once the appropriate incubation period has passed they should be tested for the virus. Because of economic disparities, execution of containment needs to consider what resources isolated and quarantined individuals will need. Will they need a hotel room to stay in? Will they need food deliveries? Will someone else need to check in on their elderly relatives while they are in isolation? Will missed work cause financial crisis? Should we consider paying infected and exposed individuals for their days of isolation as we provide jury duty pay? For containment to work, it needs to be so much more than a mechanical process, it needs to be done with empathy and compassion.
Containment is generally a strategy utilized in the very early stages of an outbreak. Several countries including Singapore, Taiwan and South Korea have used technology, extensive testing, and robust containment systems to prevent their nations from facing exponential growth in outbreaks. The Taiwanese economy was never shutdown in the way that Europe and the United states have done, yet their infection is well under control.
In contrast to general social distancing and stay in place orders, the beauty of Containment is that only a few people must be quarantined and isolated, while the vast majority can continue life in relative normalcy.
Though containment is usually implemented at the start of an outbreak, it would not be impossible to launch an aggressive containment policy, now that we have greater testing. We could use some of the Trillions in funding to hire and train contact tracers, and people to deliver food to the quarantined and isolated. CDC has started to talk about the need for renewed focus on containment. The State of Massachusetts has announced plans for an aggressive containment program. There is some hope on the horizon.
Value of Testing:
The ability to save lives, and to save the economy, is not measured by the number of coronavirus tests your nation can conduct. It is more how and when your nation uses its testing capacity. The United States had perhaps the longest time to prepare for the outbreak, but by failing to plan, and by failing to quickly deploy testing capability we missed the option pursue a containment policy like Taiwan‘s. By failing to adopt a more centrally managed science driven approach to testing we remain relatively ignorant of the true status of the infection, making our surge estimates unreliable. Having failed to deliver, our government has fallen back on the blunt instrument of social distancing which minimizes government responsibilities (and blame), and maximizes the responsibilities (and blame) that fall on individuals.
US Policy has resulted in a situation where who gets tested and when, is driven by the choice of individuals, and disparate medical institutions rather than using testing to drive a national public health policy.