Skip to navigationSkip to content
CALCULATED RISK

How to determine the accuracy of a Covid-19 test

People give swab samples during a mass coronavirus disease (COVID-19) testing to allow students home for Christmas, at the Sports Hall of Keele University, in the UK.
REUTERS/Carl Recine/File Photo
Students in the UK get tested, hoping for a negative result and permission to travel home for Christmas.
  • Amanda Shendruk
By Amanda Shendruk

Visual journalist

Published

In the UK, university students are getting tested for coronavirus en masse. The nation-wide plan is to have all students receive a negative Covid-19 diagnosis before traveling home for Christmas holidays, often in other parts of the country.

However, with test results comes potentially false confidence. Because just as with antibody tests, there is always a chance the Covid-19 diagnostic test results will be incorrect.

Let us explain, with the help of some friends.

There are three factors that go into determining the likelihood you received an incorrect test result. They are the sensitivity and specificity of the test, as well as your own pre-test probability, or risk.

✅ Test sensitivity: How good the test is at correctly identifying someone who is infected.

❌ Test specificity: How good the test is at correctly identifying someone who is not infected.

😷 Your pre-test probability: The likelihood of already having Covid-19 prior to your test.

Your pre-test probability is one of the most significant factors in determining whether you are likely to have received an incorrect result. But figuring out this risk is difficult, even for doctors. The calculation ultimately amounts to a GP’s educated guess, and it depends a lot on individual circumstances. Local rates of Covid-19, patient symptoms, likelihood of alternative diagnoses, and history of exposure to coronavirus can all affect pre-test probability.

A rough way to approximate your own risk is to start with the prevalence of Covid-19 in your area. You can often find this datapoint from local governments or media reports. Use it as your baseline. Now, increase your risk if you:

  • 🧒 Have kids that go to school
  • 💼 Work outside your home
  • 👩‍⚕️ Work in a job where you come into contact with many people, or people likely to have Covid-19
  • 🤭 Don’t follow local safety guidelines
  • 🤒 Have developed symptoms of the virus
  • For example, someone who has no symptoms of Covid-19, lives in an area with very low rates of the virus, works from home, and rarely sees anyone outside their immediate bubble would have a very low pre-test risk. If they receive a negative test result, chances are good that it’s accurate.

    Alternatively, a student sharing facilities in a dorm on a campus with high rates of coronavirus who develops a persistent cough and fatigue would have a much higher pre-test risk of infection. She would be wise to regard her negative test result with skepticism. The risks of acting on a false negative could be disastrous, especially as restrictions in many areas are temporarily easing for the holidays, and generations are planning to gather.

    📬 Kick off each morning with coffee and the Daily Brief (BYO coffee).

    By providing your email, you agree to the Quartz Privacy Policy.