COVID-19 update with Dr. Benjamin Marcum

Myth 16:  If we could all be antibody tested, we could beat this virus now.

When any pathogen invades our bodies, several things happen… usually.  Hopefully, at some point, if the pathogen is not too crafty, your immune system will recognize that there is a foreign invader and go to work trying to rid the body of a potentially harmful threat.  It does this through several processes which are beyond the scope of this article but part of the immune system’s function is to create antibodies to try and identify the threat and alert immune cells to its presence and target it for destruction.  Antibodies are proteins built by specific white blood cells called B lymphocytes.   B lymphocytes can produce antibodies after they have had time to analyze the pathogen, identify potential targets for recognition (surface proteins on the outer shell of a virus or bacteria, etc) and design the antibody to recognize those specific targets.  Another name for the antibody proteins is immunoglobulins, often abbreviated to Ig.  There are different kinds of immunoglobulins designated IgM, IgG, IgE etc.  IgE is an antibody that is active in allergic reactions.  IgM is an antibody that is quickly produced after infection by the B cells in a “quick and dirty” fashion so the immune system can get right to work on eliminating infections.  Elevated levels of IgM are indicative of a current, active infection.  Later, after the acute infection, IgG antibodies are produced to match the pathogen and are indicative of lasting immunity.  Theoretically, if one has a high level if IgG antibodies, when that person is infected again with the same bug, the immune system will quickly identify it through the lasting IgG antibodies and kill it before it gains a foothold.  This is the basis for long term immunity.  Vaccinations are used to boost IgG levels against a particular pathogen and that is how they help protect us from infections without having to actually have the disease.  Sometimes IgG levels wane and so we need vaccine boosters.  Or worse, the pesky virus mutates at an alarming rate (like flu) and we have to have a new vaccine every year that continues to try to hit a moving target.  The virus that causes Covid-19 mutated from another coronavirus so we expect it to mutate in the future.  In late March, the Washington Post wrote that early evidence is that the virus has passed through many people and the genome was relatively stable (without many mutations).  This is good news but we are just becoming acquainted with the virus.  We don’t know how fast or how much the virus will change over the years. 

The majority of Covid-19 testing has been focused on PCR testing (see my previous article on the different types of tests).  PCR testing identifies RNA from the Covid-19 virus.  Notably, it does not tell us if there is active, infectious virus present.  It only tells us if there is RNA from the virus remaining in the sample, usually from the nose or sometimes the throat or lung fluids.  After the virus has died, the RNA may still be present for some time but there is no way to determine if there is active infectious virus present at this point.  Notably, our turn around times for PCR testing are much faster now.  Initially we waited in excess of 2 weeks for result and now we are getting next day or 2 day results.  FYI.

But there is a new kid on the block – antibody testing.  The FDA has approved one antibody test (much faster with less rigorous trials than usual due to our present crisis) and several others are on the way.  An antibody test evaluates a blood sample for presence of the IgM or IgG antibody or both.  This test is becoming increasingly important to identify acute infections, people who might have had the infection before and not known it, and people who had it and are now, perhaps, immune.  We can also repeat this testing on people who have recovered and have high levels of IgG (the lasting immunity antibody) and see how long those levels last and what those levels have on imparting immunity to the virus that causes Covid-19.  It is vitally important to know who has an acute infection so we can treat them appropriately – this is an individual health issue.  From a public health perspective, we need to know how many people have been infected in a population and how long their immunity will last so we can make decisions about future public health issues.  Knowing this information has many implications, one of the most important of which is the infection mortality rate, or IMR.  To date, we do not know the true mortality rate of Covid-19.  We know the mortality rate of those patients who have tested positive, only.  Many more people have been infected and either didn’t have symptoms and so, they didn’t know, or they thought they had allergies or a cold and did not get tested.  Excluding these people from the mortality calculation artificially elevates the mortality rate.  Until we can mount a systematic population wide testing program, numbers such as the actual IMR are anybody’s guess.

Great, so lets antibody test everyone!  Not so fast… there are problems with that approach.  To date there is exactly one antibody test approved by the FDA – developed by Cellex, Inc.  And it isn’t perfect.  No test is.  This one is about as close as you can get in medical testing with a sensitivity of about 94% and a specificity of 96%.  For those of you who are not medical statisticians, that means that the test correctly identifies 94% of the true positive cases and 6% of the time patients who are actually infected will be told they are not (false negative). A specificity of 96% means that 96% of the time when the test result is negative, it is correct.  And 4% of the time, a patient who is actually not infected will be told they are positive (false positive).  So, 6% false negative rate and 4% false positive rate.  If we test 100 people, 6 will be told they aren’t infected when they actually are and 4 will be told they are when they actually aren’t.  This happens for a number of reasons.  Maybe the infection is early and the antibody level isn’t high enough to be detected or one person’s immune system made antibodies to a portion of the virus that is atypical and we aren’t testing for.  Perhaps you have antibodies to a previous infection with a different coronavirus and those antibodies resemble an antibody for the current coronavirus and so you have a false positive.  Regardless of the reason, without 100 % specificity and sensitivity, there are some statistical implications of testing an entire population before the virus is widespread.  

When you employ an antibody test that is not perfect (and remember, no test is) you have 4 possibilities – False +, True +, False -, True -.  What if we tested 10,000 people the day you tested and 1% were actually definitely positive?  That means 100 are truly positive and 9,900 are truly negative.  Out of the 100 true positive cases, 94 of them would be correctly told they were positive.  6 of the people who were truly positive would be told they have never been exposed to the Coronavirus, even though they had.  They may be actively infected, shedding virus, or they may have been exposed previously and immune, but think they weren’t.   Scale this math up to the whole population and 6 becomes 60 and then 600 etc and that poses a significant problem.  If we tested everyone in Erath County on the same day and 1% of us actually were infected at one time, 25 people would be told they were negative for the antibody when they actually weren’t – and may be actively shedding virus.  If we had a 60% infection rate, 42000 residents x 60% = 25,200 actual infections.  With a 6% false negative rate 1,512 of those people who have actually been infected will be missed.  They may have active infection and may be spreading it, or may be immune and not know it.   

On the other side of the coin, the false positive rate has even more serious implications.  If not many people have been infected yet and you are told your antibody test is positive, great!  You feel fine! You must have been one of the lucky people who were infected and didn’t know it and since you were quarantining anyway.  Like a good citizen you didn’t spread it and now you can’t because you are immune and can’t get it again!  Bring on the hugs and handshakes! But if we tested 10,000 people and we assume an actual 1% infection rate, 100 people would be actually infected and so there are 9,900 people who are actually not infected. Based on a specificity of 96%, out of 9,900 actual negatives, 9,504 would be correctly identified as negative.  The other 396 patients would be told they were positive for previous infection and subsequent immunity when they actually weren’t. 

Wait, what?  Almost 400 people would be told they were immune to the virus causing Covid-19 and they actually were vulnerable?  That’s 400 people who might think they can go shopping right now and not worry about contracting the infection or, worse, might feel like they are safe to go visit their grandchildren – since they were lucky enough to avoid serious complication from the infection they must have had and not known they had.  If you thought you had already had the virus and were likely immune, how would you feel?  How would you act?  What if the test were wrong and your false confidence caused you to get sick or pass the disease to someone more vulnerable?  You can see why it is problematic.  If we ramp that math up to the total population of Erath County, 42,000 residents – assume a 1% actual infection rate.  420 people would be actually infected and 41,580 people would not be infected.  96% of 41,580 is 39,917 people who would have been correctly identified as not being infected.  1663 people would be told they are immune and protected when they are not.  This is a public health disaster waiting to happen.  But as the infection rate increases, the numbers start to look a little better.  Let’s evaluate a 60% infection rate in Erath County. (Again, we can’t have 60% of us infected at once due to limitations on health care resources – but herd immunity happens in the 70-80% range and over time we will likely approach that threshold).  42000 x 0.6 = 25,200 people who are actually infected.  16,800 are not.  96% x 16,800 = 16,128 people would be correctly identified as not having been infected but the remaining 672 would be told they are immune when they are not.  
Intuitively, we don’t think a test is more or less beneficial based on the number of people who have actually been infected with a pathogen, but as you can see, it is.  The closer we can get to 100% specificity and sensitivity, the better, but we are not going to get there.  Antibody testing will be a big part of overcoming this virus and getting back to work but it is not a silver bullet.  We currently don’t have antibody testing at our clinic.  For the right case I could probably get a test done through a lab in Fort Worth but at this point, antibody testing is best used on a population basis and we don’t have enough tests to initiate testing on that level, yet.   We will – and it will help us get back to whatever our new normal will be.  But probably not soon.

Thanks for muddling through the math with me.
BAM