The rhetoric after the latest COVID-19 update has changed, slightly. The discussions are, appropriately, about the burden of COVID-19, but, also, about how we get back to normalcy. I admit I am optimistic about this contagion. It is not polio, small pox, or HIV. People were paralyzed, died, or got AIDS. The frequency of serious outcomes was nearly 100% for each person infected. COVID-19 is not like these, which is reassuring, at least in comparison. COVID-19 is serious but looks more like the 2009 H1N1 flu pandemic than it looks like those mentioned above.
The reports of death and burden of illness are daunting to hear. I understand the public’s concern as I share it. My other concern is, though, that the reports are burdened, themselves, by uncertainty. In my last post I pointed out that plans for testing must be concerned with the imperfections of the test, overestimating, perhaps the burden of illness. I hoped that knowing of those imperfections of testing might be reassuring.
However, I am not a social psychologist, and I am not sure if that knowledge is helpful to you. I remember a poignant example of a movie depicting a nuclear accident (The Day After), shown one week before a national survey about risk perception, being blamed for an overestimated risk of dying of a nuclear accident. The movie was seen by over a million people, influencing, it appeared, the public’s perception.
The point is that our perceptions are susceptible to fear, even if unreal. Knowing the imperfections of testing is a truth, but is it helpful to know that reports of COVID-19 may be inflated by error?
I admit, I don’t believe that knowledge, alone, is the cure for our fear and that knowing more will be the impetus to return to normalcy. We may be so sensitive to reports of those dying we may need to hear of none before we move on. The problem with this is that the reports will continue, unless the contagion recedes, but we can’t continue to act as we are as a society until it does recede.
So, here are some other ideas, suggested by our leaders, that we must ponder as we decide what to do. I will offer my opinion as a starting point for debate. And, then, I will offer what I think we need to do to end our isolation from our social response.
-
Get a serum antibody test on everyone.
The serum antibody test is being touted as an answer to a return to normalcy. Some think that it would be soothing to learn that we have been infected, but survived. Some think that if we have survived and have the antibody, we may be COVID-19 immortal and be given an “immunity passport”. This test, then, may give us hope to return to work, or back to normalcy.
These are hopeful, but not necessarily true suppositions, and the test, even, is not yet available. Hence, these hopeful ideas are, presently, unfounded. First, we do not even develop life-long immunity to flu, a contagion we all are familiar with. We keep getting it seasonally, so we may, also, face the same with COVID-19. Second, just because antibodies are present for COVID-19 does not mean they confer immunity. Consider the fact that antibody testing after the 2009 H1N1 flu epidemic found that only about 50% of people had antibodies six months after being infected (worse in those only vaccinated; about 30% had antibodies at six months).
Third, the antibody test is imperfect, with estimates of 5-10% false positives and 5-10% false negatives. The companies dueling to produce the test report varying true/false rates but none has faced clinical scrutiny. If we want to assure that it is safe to return based on a test result, can we afford an imperfect test? The imperfect test means that some with the disease will test negative potentially continuing to spread the disease, while some will test positive but will have a false sense of security.
Hence, immunity may be fleeting, and when to test becomes an important consideration before wide-spread, unplanned testing. While immunity may, alternatively, be long-lasting, the antibody test will not tell us that. There should be some serious research planning before wide-spread use of a test that may not help us.
-
Better science using random samples with gold standard tests.
An alternative to the present reports is to improve science. This suggestion goes like this. If we could accurately assess the burden in the well and ill, we could plan with confidence. An astronaut recently commented that the way he learned to be calm in space was to be fully informed and knowledgable. He, too, equates calm with knowing.
This suggestion to improve the science of epidemics warms. It may surprise you to know that we don’t even have a scientific plan for assessing when flu is here, or how big is the burden. Testing for flu, even, is haphazard. If we were to improve our ability to know the true burden of any malady, people would be able to plan fearlessly.
The way to scientifically test the burden in the well is to perform gold standard tests (positive only in diseased people) on a random sample, not all people. We don’t have enough high quality tests to test all, so we have to test fewer. Choosing who, then, must be done randomly to circumvent the problems of using inferior tests in everyone.
The place to do random testing would be, for one, the work force. If the prevalence is low, workers might be assured to return to work while following universal infection control practices. The potential problem with this idea, however, is that the prevalence may be high. However, this might be equally reassuring if many are positive, but well.
Both the idea to stop all testing in the well due to imperfections in the test, and a call for random sampling to accurately assess the prevalence, require fewer tests. The second requires an improved research structure, and, perhaps, this experience with COVID-19 will push us to restructure the NIH and CDC to be better science agencies for epidemics and disease burdens. There is no reason to not have the capability to do random samples of disease burdens in the workforce, or any public domain. I have written essays asking for better science, and suggesting how to do that. Maybe COVID-19 will spur a better scientific response to the next contagion.
-
Test only the ill, and test for every infectious virus.
This idea has merit. Presently, we do not uniformly know how to assign a cause of death, and to assign a cause of death to a specific infection in the middle of an epidemic of that infection will inflate COVID-19 as a cause. Also, right in the middle of all this, the CDC notes an uptick in flu. What infection is causing death? We won’t know this unless we test everyone who dies for all viral agents. The hypothesis is that COVID-19 is not the only dangerous culprit out there, and if the public knows this, they may develop a better understanding of the natural history of COVID-19.
-
Wait until we get a treatment, or a vaccine.
An astronaut recently commented that the way he learned to be calm in space was to be fully informed and knowledgable. He, too, equates calm with knowing.- Situational Awareness! Yes!
1) For any isolation to work – all need to comply.
Whether now or in future. Or it doesn’t work.
2) like 9/11 – there will be noncompliance and a constant infrastructure for testing, policing.
3) HIV did not aerosolize- law of averages and proportions. Apples to oranges. 1% death rate spread to 500k people in 2 months, 20+k deaths —- not HIV. Even if these numbers are wrong, still too rapid too many.
4) +PPV better in dis prevalent populations – so testing frequent and as many times as needed – must become a norm. Like TSA in airports. New reality after 9/11.
5) we need to partition as much as we can this population. Right now we are acting like 50 little nations. With variations —— model of govt won’t work. Must have uniform processes. Protocols. Technology like rapid / remote / home monitoring telemedicine, geomapping dynamically is critical. I don’t care about hipaa —- a relic reg from the past. – will seem irrelevant in the near future. The whereabouts and the testing status, symptoms, verifications, monitoring will need to become the norm. Contact tracing — new reality.
6) the world has and will deal with this again, and it won’t be “wiped”. —- May take years, decades. With new strains.
7) in order to evolve as a species- we need more cooperation and less finger pointing.
First off; I do not blame any person and apologize if my blog suggests that. Fear is hard to manage. But, as a public health person, decision-analyst, I think we must think about how to do better science. I harken back to 9:11 and learning how unorganized and disparate the security institutions. We need to rethink how we reorganize for epidemic science and all disease management science in the future.
Love your comments and obvious passion. Some comments to yours:
Impossible to isolate everyone. Will have to decide if exposure is certain, like flu. If it is certain, delay may be more harmful for immunity?
Don’t buy your estimates of burden, we don’t have good data about how bad COVID-19 is. We can’t even agree on definition of death from COVID-19. Present plans by leaders are based off very poor science.
HIV killing still 6-10000 in US yearly and age of death lots of life expectancy lost. World wide, big problem and immunity does not develop. But, agree, not comparable contagions.
Interesting point about TSA; wonder what would happen if the false positive of TSA was 10%; would be lots of people designated as dangerous?
Not sure what is best, but do think we need to think of all possible avenues for future and forward. That is intent of blogs; spur debate.
Thank you.
Great perspective. Would love to hear you take on current approach on therapeutics for covid-19 (remdesivir specifically). I worry the rush to “do something” could lead to a lot of sloppy science.
Howdy! Great to hear from you. I share your concern. Speed is problematic. I have read reports of “speed to getting a test out” being blamed for poor test characteristics at CDC. Speed in science is never a good idea.
I have not followed the trials, yet, as they are preliminary, phase 1 type stuff, which shouldn’t be reported. But, see some potential problems, like, who are they choosing for study? Random samples? What are inclusion/exclusion criteria. What are the outcomes? Death, or time of illness? Given the co-morbid conditions of patients, how well will randomization work to balance confounding? What is the bias for these studies to be “significant’ (stock market jump on just some good news)? I see lots of potential problems with these studies and share your concerns.
All the best, Kuma.