Discussion Board

Expand all | Collapse all

False Negative: COVID-19 Testing's Catch-22 | MedPage Today

  • 1.  False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 12:26
    Maybe your negative result isn't good news. A low sensitivity means 30% of patients tested negative may be spreading the disease.
    Helene 

    False Negative: COVID-19 Testing's Catch-22

    - And the consequences of being wrong

    Saurabh Jha, MD
    A blue gloved hand holds a test tube of blood labeled COVID-19 with both positive and negative boxes checked

    In a physician WhatsApp group, a doctor posted he had a fever of 101 degrees Fahrenheit and muscle ache, gently confessing that it felt like his typical "man flu" which heals with rest and scotch. He worried that he had coronavirus. When the reverse transcription-polymerase chain reaction (RT-PCR) for the virus on his nasal swab came back negative, he jubilantly announced his relief.

    Like Twitter, in WhatsApp, emotions quickly outstrip facts. After he received a flurry of cheerful emojis, I ruined the party, advising that despite the negative test, he assumes he's infected and quarantine for two weeks, with a bottle of scotch.

    It's believed that the secret sauce to fighting the pandemic is testing for the virus. The depth of the response will be different if 25% of the population is infected than 1%. Testing is the third way, rejecting the choice between death and economic depression. Without testing, strategy is faith-based. But what'll you do differently if the test is negative?

    That depends on the test's performance and the consequences of being wrong. Though coronavirus damages the lungs with reckless abandon, it's oddly a shy virus. The Chinese ophthalmologist who originally sounded the alarm about coronavirus, Li Wenliang, had several negative tests. He died from the infection.

    In one study, RT-PCR's sensitivity – that's the percentage of infected testing positive – was 70%. Of 1,000 with coronavirus, 700 test positive but 300 test negative.

    Good enough?

    Three hundred "false negatives" may believe they're not contagious and could infect others, undoing the hard work of containment.

    Surely, better an imperfect test than no test. Isn't flying with partially accurate weather information safer than no information? Here, aviation analogies aren't helpful. Better to think of a forest fire.

    Imagine only 80% of a burning forest is doused because it's mistakenly believed that 20% of the forest isn't burning. It must be extinguished before it relights the whole forest, but to douse it, you must know it's burning – a Catch-22. That "20% of the forest" is a false negative – it's burning, but you think it's not burning.

    Testing may enable precision quarantining, enabling us to think globally but act locally. But it's the asymptomatics who drive the epidemic. To emphasize – asymptomatics haven't yet developed symptoms such as cough and fever. Theoretically, if we mass test, we can find asymptomatics. If only those who test positive are quarantined, the rest can have some breathing space. Will this approach work?

    RT-PCR's sensitivity, which is low in early illness, is even lower in asymptomatics, likely because of lower viral load, which means even more false negatives. The virus's incubation time of five days is enough time for false negatives – remember they resemble the uninfected – to visit Disney World and infect another four.

    Whether false negatives behave like tinder or a controllable fire will determine the testing strategy's success. The net contagiousness of false negatives depends on how many there are, which depends on how many are infected. To know how many are infected, we need to test. Or, to know whether to believe a negative test in any person, we must test widely – another Catch-22.

    Chest CT is an alternative. It's rapid – takes less than an hour, whereas RT-PCR can take over a day to report. In one study, CT had a sensitivity of 97% in symptomatic patients and was often positive before RT-PCR. But there are caveats.

    The real sensitivity of CT is likely much lower than 97% because the study has biases that inflate performance. CT, like RT-PCR, has a low sensitivity in early illness and even lower sensitivity in asymptomatic carriers for the same reason – lower viral load. Furthermore, CT has to be disinfected to prevent spread, which limits its access for other patients.

    Coronavirus's signature on CT – white patches in lungs, known as ground-glass opacities – doesn't have the uniqueness of the Mark of Zorro, and looks like lung injury from other rogue actors, which means we can mistake other serious conditions for coronavirus. Imagine hyenas in wolf's clothing.

    No test is perfect. We still use imaging despite its imperfections. But, let's ask: what'd you do differently if the test is negative, and you have mild symptoms of cough and fever? Should you not self-isolate? What if you're falsely negative and still contagious? If the advice dispensed whether the test is positive or negative is the same – i.e., quarantine for 2 weeks – what's the test's value?

    Perhaps people will more likely comply with voluntary quarantine if they know they're infected. Information can nudge behavior. But the logical corollary is that to comply with social distancing, you need to be tested. People flocking to CT scans to affirm they're not infected could infect those hitherto uninfected.

    Testing is valuable in managing populations. To individuals, the results must be framed wisely, such as by advising those who test positive to quarantine because "you're infected" and those who test negative to keep social distancing because "you could still be infected."

    Even when policy goals are uniform, messaging can be oppositional. "Get yourself tested now" contradicts "you must hunker down now." When messages contradict, one must choose which message to amplify.

    The calculus of testing can change with new tests such as antibodies. The value of testing also depends on what isolation entails. A couple of weeks watching Netflix on your couch isn't a big ask. If quarantine means being detained in an isolation center fenced by barbed wire, the cost of frivolous quarantining is higher, and testing becomes more valuable.

    I knew the doctor with the negative RT-PCR well. He's heroically nonchalant about his wellbeing, an endearing quality that's a liability in a contagion. In no time, he'd be back in the hospital; or helping his elderly parents with groceries. Not all false negatives are equal. False-negative doctors could infect not just their patients but their colleagues, leaving fewer firefighters to fight fires.

    Saurabh Jha, MD, is a radiologist and can be reached on Twitter @RogueRad. This article appeared on KevinMD.com.



  • 2.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 13:01
    Just to be clear, 30% of patients WITH THE DISEASE who test negative could be spreading disease if their isolation/quarantine is lifted mistakenly. It's not 30% of all negative tests in all people. Post-test probability depends on the test result but also the pre-test probability (prevalence of disease in the subpopulation to which you or your patient belongs). If you have classic symptoms, a negative test is not reassuring. If you don't have classic symptoms, a negative test would support the notion that you don't have the disease. See article in the NYT today by Harlan Krumholz for a good discussion.  

    John E. Brush, Jr., M.D., FACC

    Professor of Medicine

    Eastern Virginia Medical School

    Sentara Cardiology Specialists

    Sent from my iPad





  • 3.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 13:25
    Link to Krumholz article in NYT.

    https://www.nytimes.com/2020/04/01/well/live/coronavirus-symptoms-tests-false-negative.html?searchResultPosition=1

    John E. Brush, Jr., M.D., FACC

    Professor of Medicine

    Eastern Virginia Medical School

    Sentara Cardiology Specialists

    Sent from my iPad





  • 4.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-03-2020 10:55
    With a false negative rate of 30 percent, if we assume the true negative rate to be 100 percent, the negative likelihood ratio (LR) of the corona virus test will be 30/100 = 0.3. If we have a patient with a very highly typical presentation in whom the prior probability of corona virus disease is say, 95 percent, a negative test will lead to a post test probability of 85 percent. In a patient with somewhat typical presentation in whom the prior probability is 50 percent, a negative test will lead to a post test probability of 23 percent.
    As this test is currently being done only in symptomatic patients in whom the prior probability is around 50 percent or higher, it leads to an
    unacceptably high post test probability. Therefore a negative test is worthless in ruling out corona virus disease in such patients so that we should stop interpreting it as absence of coronas virus disease in these patients. We should perhaps only interpret a positive test as indicating presence of disease in these patients.
    When we start doing this test routinely in asymptomatic patients or in those with minimal symptoms in whom the prior probability is say 5 percent, a negative test yields a post test probability of 2 percent. As the prior probability is already very low at 5 percent, this negative test does not add much to ruling out disease in these patients.
    My impression is that a negative corona virus test with a likelihood ratio of 0.3 is practically worthless in ruling out corona virus disease in any patient. We need a negative test with LR of 0.1 or lower to help us rule out corona virus disease in a patient.
    It would be of interest to know how we should interpret a positive test in a patient who is asymptomatic and thus has a prior probability of less than 5 percent. A knowledge of the likelihood ratio of a positive test would help, I believe, in this interpretation.

    Bimal Jain MD
    Northshore Medical Center
    Lynn MA 01904  





  • 5.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 14:07
    Thanks Dr. Brush, that's an excellent clarification to add. However, with the difficulty in getting these tests, only symptomatic patients and exposed healthcare workers are most likely to be tested. Given the large subset of symptomatic test patients, 30% inaccuracy of negative results is way too high a number. Especially if those symptomatic test subjects are coughing and sneezing. In any case, the point is to let people know that the tests are not highly sensitive and that patients with symptoms who test negative still should stay home.
    Best,
    Helene 

      
       Website 
       Twitter 
     






  • 6.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 14:19

    Agreed Helene:

     

    A further limitation of testing for Covid-19 is in those persons who have an early infection yet have no symptoms and a negative test because the small number of viral particles present in the nasopharynx present to be reliably sampled.  These people will be less prone to self-isolation than those with symptoms and/or a positive test.  It may well be it is this population of asymptomatic people who are the major vector within our society spreading the infection over the next few days as their viral load rises and they begin to shed the virus but before the onset of signs or symptoms.  Therefore, the timing of the test is critical and we should all understand that a single test result provides a single static slice of the infection, not a dynamic chronology of it os we shouldn't conclude that a negative test, even if a true negative regarding the specimen having no virus in it, necessarily means the tested person isn't going to spread the virus to many others.

     

    Mark Gusack, M.D.

    President

    MANX Enterprises, Ltd.

    304 521-1980

    www.manxenterprises.com

     






  • 7.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 14:35


    John E. Brush, Jr., M.D., FACC

    Professor of Medicine

    Eastern Virginia Medical School

    Sentara Cardiology Specialists

    Sent from my iPad





  • 8.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 14:36
    I agree. The shortage of tests is frustrating everyone. 

    John E. Brush, Jr., M.D., FACC

    Professor of Medicine

    Eastern Virginia Medical School

    Sentara Cardiology Specialists

    Sent from my iPad





  • 9.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 14:45
    Agree Helene. 

    It's a simple matter of pretest probability. If a patient has a high pretest probability of disease based on clinical presentation, a negative result of a test with limited sensitivity should be assumed to be a false negative. We treat them as if they have disease. When possible we try to repeat a test in a couple days when viral load is higher. 





  • 10.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 05-01-2020 11:35
    
    Is it possible this incredibly inexpensive, simple test developed at the Weitzmann Institute in Israel can help identify which patients will wake from coma? They claim 100% accuracy over a 4 year period. Astonishing. 
    https://www.nature.com/articles/s41586-020-2245-5

    Best,
    Helene
      
        Website 
        Twitter 
        LinkedIn
     
        Facebook





  • 11.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 05-01-2020 12:49
    Dr Brush

    I agree with ur comment :)

    But To clarify ...
    How does one use the NPV result from a cheap, "limited sensitivity" first test to adjust the pretest probability so that one can calculate the NPV from a second more expensive, more sensitive test???

    Thanks





  • 12.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 13:05
    Thanks Helene,

    Accuracy is what I have been pleading for for months on this list. There are so many things we do when caring for patients that, I believe, are dangerously inaccurate. Is but time we tidied up our problems.





  • 13.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 13:37

    Robert:

     

    In order to begin this process we'll have to add advanced statistical analysis as a prerequisite for applying to medical school so that they have an appropriate foundation.  Then, we'll have to introduce the students to the field of Metrology: Measurement Uncertainty to begin to create a culture within the healthcare field that acknowledges the limitations of all testing modalities.  Otherwise, our future physicians will not be equipped to assess the results of published validation trials of these testing modalities nor will they be able to interpret laboratory results in a safe manner when diagnosing and treating patients.

     

    Everyone stay safe.

     

    Mark Gusack, M.D.

    President

    MANX Enterprises, Ltd.

    304 521-1980

    www.manxenterprises.com

     






  • 14.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 14:22
    Probably worse that 30%.
    25-50% may have few to no symptoms. 
    People may shed for several weeks after recovery.
    Tom Benzoni





  • 15.  RE: False Negative: COVID-19 Testing's Catch-22 | MedPage Today

    Posted 04-02-2020 15:59
    Is the severity of the number of particles of virus transmitted to a patient in any way related to the accuracy of the test and the severity of the subsequent symptoms?

    I heard today that breathing and talking may spread the infection. Is that true? Does that mean 6 feet is not correct?

    Also, is there a benefit to viral infections - do they in some way provide something that is a benefit?