CMA Loop - Physician Community

Discussion Thread View
Expand all | Collapse all

Antibody Tests For Coronavirus Can Miss The Mark

  • 1.  Antibody Tests For Coronavirus Can Miss The Mark

    Posted 04-15-2020 10:27 AM

    Antibody Tests For Coronavirus Can Miss The Mark

    The increased availability of COVID-19 antibody tests is not a panacea. 

    Many people are hoping to use the results of this test to prove that they have had the disease, and thus cannot be infected a 2nd time.

    A Story on NPR's Morning Edition 4/15/20  "Antibody Tests For Coronavirus Can Miss The Mark "shows why just the results of the antibody test aren't enough.

    These antibody tests may have a specificity as high at 95%.

    Just looking at the specificity alone is not enough. You also have to look at the prevalence of the disease in the population that you are measuring.

    Here is an example from the NPR Story. "Here's what would happen if you used a test with 90% specificity in a population in which only 1% of the people have coronavirus (1% prevalence) . Nobody knows for sure, but that could be the situation in many parts of the country. In that instance, more than 90% of the positive results would be false positives, and falsely reassuring.

    I don't want to discount the importance of the antibody test, but the results you do get from this test should not be used in isolation. 



    ------------------------------
    William Cotton, MD
    ------------------------------


  • 2.  RE: Antibody Tests For Coronavirus Can Miss The Mark

    Posted 04-16-2020 08:30 AM
    Thank you, Bill, for the reminder about how prevalence effects interpretation of test results. I am glad NPR took on this topic of antibody testing. They are good at  boiling down complex topics to a more understandable level. But it still would seem a steep mountain of cognitive effort for the general population to digest this. I hope they eventually get the fact that "safety" from  COVID,or any infectious disease, is a relative term.

    ------------------------------
    Terry Olson
    MedOne Hospital Physicians
    ------------------------------



  • 3.  RE: Antibody Tests For Coronavirus Can Miss The Mark

    Posted 04-16-2020 11:21 AM
    Bill, this is a hugely important point. I still remember this lecture in med school (even the doctor that gave it) It is not intuitive. The math is so important otherwise the conclusion that is drawn is far off from reality. Thank you for bringing it to our attention.

    ------------------------------
    Marc Parnes
    Eastwind Women's Health Inc.
    ------------------------------



  • 4.  RE: Antibody Tests For Coronavirus Can Miss The Mark

    Posted 04-16-2020 12:25 PM
    Bill, those are good points. Perhaps before I sit down at the Buckeye's home opener I will ask all around me to show their "immunity passports". But you're correct about the concerns, hopefully more data will clarify.

    ------------------------------
    Kanny Grewal
    OhioHealth Heart & Vascular Physicians
    ------------------------------



  • 5.  RE: Antibody Tests For Coronavirus Can Miss The Mark

    Posted 04-17-2020 04:58 PM
    Bill,
    very timely for you to raise this valuable perspective.
    Here's my take on this: with 90% spec and 1% prevalence, 1 person out of 100 has the disease. Thus, 99 people don't have the disease. So with 10% false positive rate, then 9.9 (10) people will have a false positive test. Thus, saying that 90% of the positive results (9 out of 10) would be false positive is true but may be misleading. This might make people think that the majority of the population will have inaccurate tests.
    I feel that the more informative stats are positive and negative predictive values. Using the same prevalence and specificity, and adding a sensitivity of 90% (not knowing what has been reported), then the PPV is 8% and the NPV is 99%. That tells a patient that they have an 8% chance of having the disease if the test is positive, and a 99% chance of not having it if the test is negative.
    So we are saying the same thing; the population can not fully rely on the test, and you pointing out that a positive test does not necessarily mean that the patient has the disease is laudable.
    Since hardly any laymen understand these stats, and since few docs probably do, I feel the role of physicians is to make their patients aware of the general concepts regarding the accuracy and value of testing.
    Thanks, Bill

    ------------------------------
    Jeffrey Bell

    Hudson Institute ACTP Certified Life and Leadership Coach

    Professional Certified Coach, ICF
    ------------------------------