Coronavirus Update - Broader symptoms; uncertainly about case fatality rate; assessing treatments
Updated: Apr 15
At the beginning of the pandemic, scientists were trying to find symptoms or signs that distinguished COVID-19 from other infections so it could be easily diagnosed. For example, initial case series suggested that the two symptoms, cough and fever, were highly suggestive of COVID-19.
The loss of sense of smell and taste was another symptom rumored to be specific for COVID-19. However, any ENT can tell you the loss of sense of smell and taste is a very common symptom of allergies and just about any respiratory tract infection. In fact, 31% of influenza cases are left with a loss of sense of smell and taste. The good news is that most people regain their abilities. Regardless, loss of a sense of smell and taste IS NOT a specific symptom of COVID-19.
Now we have more data. Putting many case series together, a different picture emerges. COVID-19 can present with many symptoms. Anything from a typical “cold” with runny nose and sore throat, to flu like symptoms with high fever and body aches, to no symptoms at all. The message here is: it is very difficult to distinguish COVID-19 from other illness by symptoms and signs alone (clinical picture). It has a broad presentation. The only way to tell is to get tested.
The limitations of testing in individual patients
As most of you have realized it is very difficult to get tested, unless you are very sick and go to the hospital. But do you need testing? Will it change the way we handle your situation? Currently outpatient tests takes 5 days to come back (if you are lucky). By the time you get the test results, you will likely be better. If you test too early in the course of the disease, you may get a false negative. People who have symptoms of a respiratory tract infection need to isolate themselves at home, regardless of whether they have COVID-19 or not. If a patient gets short of breath they should go to the ER, regardless of whether they have COVID-19 or not. In reality, knowing whether you are positive or negative does not change management.
It is also important to understand that the current molecular test (a DNA test), only tests for the active disease. If you had the disease 2 weeks or 2 months ago, for instance, the molecular test will be negative. A more helpful test for many patients is an antibody test. An antibody test, tests your immune system to see if you were exposed to SARS-CoV-2 at any point in the past. It will also inform you if you are immune to the virus and a “survivor” - ready for the post-COVID world we are all looking forward to. We have been promised an antibody test. It should be available in the near future; the FDA has expedited the approval process. Details are sketchy.
Still uncertainly about case fatality rate
While the death rate is increasing in the US every day, and hospitals in some cities are congested, we still do not know the virulence of SARS-CoV-2. The virulence is how likely an infection will cause harm. The easiest way to approximate this is by calculating the case fatality rate (CFR). It is simply the number of fatalities (numerator) over the number of cases (denominator). The CFR is easily determined with readily available stats. As of yesterday, there were 33,881 fatalities in the world and 718,685 cases. (See below from the Johns Hopkins database). This is a CFR of 5%.
However, you may have noticed that this number varies greatly; compare country to country (Italy to Germany) and from week to week. In actuality, the calculated case fatality rate does not reflect the true virulence of COVID-19. Instead, the data is prone to large collection errors. We misattribute cause of death (the problem of the numerator) and we do not test everyone (the problem of the denominator). Let me explain…
The problem of the numerator - Are all the deaths being attributed to COVID-19 really caused by COVID-19? Determining cause of death is something that has plagued medicine for centuries. Let me give you two extreme scenarios of death attributed to COVID-19:
A 45 year old healthy male develops cough and fever, and then shortness of breath, dies in the hospital with viral pneumonia; he tests positive for SARS-CoV-2 and his cause of death is listed as, “COVID-19”. Sounds reasonable.
A 30 years old male dies in a fatal motor vehicle accident; he happens to tests positive for SARS-CoV-2; should his death be attributed to COVID-19?
Approximately 150,000 people die every day on our planet - on a good day. A significant number of these people will be carrying SARS-CoV-2. Is SARS-CoV-2 the cause of death in these people? To make things even more complicated, a significant number of people who die of respiratory tract infections die with more than one infection - co-infections - which one of these infections is responsible for the death?
The problem of the denominator - The number of cases are not the ACTUAL number of cases. This is because we are not testing everyone. We mainly test patients that are sick. And we know for sure there are many mild and asymptomatic cases. Hence, we are only seeing the tip of the iceberg. The number of actual cases is much greater than we are testing. When the number of cases starts rising in a country, this could mean more cases, or more testing, or both? To solve the problem of the denominator we need more surveillance testing. This means: test the entire population, symptoms, or no symptoms. This will get us close to the real denominator. Iceland has begun surveillance testing. Thus far, they tested 5,571 people randomly without symptoms. 48 were positive (0.86%). If we extrapolate this number to the entire population of Iceland (380,000), it is estimated that there are 3,000 cases. At the time of the survey, there was only one death in Iceland. This represents, a case fatality rate of 0.03% - less than the flu. This is 2 orders of magnitude different than the 5% I quoted earlier in the email.
See the problem? How can we make decisions without this very basic knowledge? Without correct data, diseases are misunderstood, health programs do not accomplish their goals, and resources are incorrectly allocated.
Predicting success rates for different treatments
There are 3 broad types of treatments available for COVID-19: public health measures to prevent spread, vaccines to prevent new infection, and anti-virals to treat active disease.
Public health measures are currently our only defense against COVID-19. There is debate as to how this should be implemented and how draconian the measures should be. Pubic health measures can ruin the infra-structure of societies and their supply chains, and they only seem to be a temporary measure - delaying the inevitable. However, when followed, they work, and they work fast. The real question for you as an individual is how to live your life right now. It is true that the virus can remain in the air and remain on certain surfaces for days; suggesting the possibility of indirect spread. This can be frightening and makes you want to live in a plastic bubble. Yet, tracing COVID-19 patients and their contacts in epidemiolgic studies, the majority of cases were spread by close family and friends in crowded and prolonged encounters. Based on this, what should you do? Here is a video from a Cornell ICU doctor, David Price MD, with very pragmatic advice - it is an hour long.
Vaccines are very promising; our track record with them is pretty good. We have had success with dozens of viruses in the past. Moderna, a Cambridge, Massachusetts biotech company, has already started testing the first COVID-19 vaccine. The reason it could be produced so quickly is because the same vaccine was already developed for SARS and MERS and tested in animals. It targets one of the surface proteins common to the coronavirus family. They started phase 1 trails last week - 45 healthy volunteers will get different dosages of the vaccine to test for safety and determine which dosage elicits an immune response. The finished product may take up to one year to be ready. The other area of promising research is synthetic antibodies - replicating the antibodies produced by convalescent COVID-19 patients. Although this approach has promise, it is important to remember that it has yet to be successful with any other infectious disease.
Anti-viral medications to treat active infection is the least promising potential treatment. I have dealt with this topic in prior emails here, here, and here. Although most of these agents are ready for use, we still need large controlled trials, to prove whether they work or not. Anecdotal evidence won’t cut it for a disease that has a 99% recovery rate without treatment. These trials will be performed in the next few weeks and months. To predict success with these agents it is important to explore our track record with other viruses. It is very bad. The only virus we have had success with is HIV; it took 20 years with hundreds of failures, and even then, does not cure the infection - just holds it at bay. Tamiflu barely works for the flu, and it was biochemically designed for influenza virus. The drugs that scientists have scrounged together in the last couple of months to treat COVID-19 (Remdisivir, Kaletra, Hydroxychloroquine, Avigan, etc) were not engineered for SARS-CoV-2, but instead, co-opted from other uses. The likelihood that a medicine, designed with other purposes in mind, miraculously treats COVID-19, and becomes the only medication ever to successfully treat a cold virus, is low to say the least. It’s possible, but unlikely.
If you are interested in the above topics please listen to this BMJ (British Medical Journal) podcast. The BMJ is one of the top 5 medical journals in the world and, as far as I am concerned, a pillar of evidence-based medicine and critical thinking in medicine.