Re: Covid-19: People are not being warned about pitfalls of mass testing
Informing the public about the limitations of covid-19 tests is a worthy aim, but the message can be undermined by errors in statistical reasoning.
It was claimed that ‘false positive results are rare’ with the Innova lateral flow test and that ‘false negatives are much more common’. The test has an overall sensitivity of 76.8% and from this it was wrongly deduced that the false negative rate was 23.2%. If every single person in the test sample had active transmissible disease, then this false negative rate would have been correct. However, when the prevalence of active disease in the population is only 1 in 50, the false negative rate would be calculated as 0.46% (5 missed cases in every 1000 people tested). Note that a different prevalence would have given a different false negative rate.
The lateral flow test has a specificity of 99.68% (BMJ 2020;371:m4469). With a prevalence of 1 in 50, this results in a false positive rate of 0.29% (3 people wrongly identified as positive per 1000 tested). These calculations illustrate that false negatives and false positives are both uncommon using a test with a moderate sensitivity and a high specificity in a population with a relatively low prevalence of active disease.
The sensitivity of the lateral flow test was reported to fall to 57.5% for self-trained staff in a track and trace centre. Despite only identifying just over half of all actual cases, the false negative rate would still be relatively low: 0.85% (9 missed cases in every 1000 tests).
Overlooking cases of active covid-19 infection is always of concern. The public still need to maintain precautions when their test results are negative (whether falsely or truly negative), but in this essential educational message, it is important to avoid accidentally exaggerated claims.
Competing interests: No competing interests