Why so many mismatches between antigen and PCR tests? They may both be right

A high rate of positive antigen tests subsequently not confirmed by PCR tests in reported data from asymptomatic close contacts suggests anomalies in this part of the HSE testing programme, according to a prominent infectious-disease doctor.

Dr Paddy Mallon, a consultant at St Vincent’s University Hospital in Dublin, says that the main issue with antigen tests has been false-negative rates – which is to say having a negative antigen test followed by a positive PCR test – rather than false-positive rates. “This is why these data suggest an anomaly somewhere in the testing programme,” he says.

Mallon was responding to data on antigen tests presented at a Nphet meeting in early December. The data found just 5,741 – or 54 per cent – of positive antigen-test results from asymptomatic close contacts were confirmed by subsequent positive PCR tests.

So how can so many people test positive for Covid-19 in an antigen test and subsequently test negative in a PCR test? It may at least partly be due to time lags between contact and testing, says Mallon.

Online booking system for boosters opens as record number of jabs given

---

Covid: 6,307 further cases as CMO tells young people he is not ‘finger pointing’

---

Israel set to offer fourth dose of Covid-19 vaccine to people over 60

---

First, it takes time for asymptomatic close contacts of confirmed Covid-19 cases to receive their free antigen tests from the HSE. There may then be a time lag before a close contact takes the antigen test and reports its result. In the event of a positive antigen test there can be further intervals before a person books and then has a PCR test. So both tests can be accurate but show different results.

“If someone does an antigen test one day, reports it the next day and has a PCR test the following day, they might have cleared the virus from their system before they have their PCR test,” Mallon says.

Vaccinated people also shed the Covid-19 virus for shorter periods of time, which may also explain the mismatch between positive antigen-test results and negative PCR-test results in close contacts who are vaccinated.

Mallon is concerned that, at this critical time in the pandemic, people might stop using antigen tests based on data from close contacts, which doesn’t represent the overall picture of the way antigen tests and PCR tests are used together.

Prof Mary Horgan, chair of the Rapid Testing Expert Advisory Group, adds that data from the general population reporting into the HSE portal suggests a much higher tally between positive antigen tests and positive PCR tests. “About 75 per cent of positive antigen tests are confirmed by PCR tests in the general population,” she says.

According to the Amárach Research tracker for the Department of Health, 88 per cent of people with a positive antigen test plus symptoms said they then arranged a PCR test and self-isolated. But only 41 per cent with a negative antigen test plus symptoms arranged a PCR test, and only 43 per cent self-isolated.

Horgan adds that antigen tests are an additional layer and not a substitute for other public-health measures. “The appropriate use of antigen tests is to identify infectious cases to break the chain of infection. They are very good at doing that, and people should do antigen tests before attending Christmas events.” They should not attend if they have a positive result, she says, and should have a follow-up PCR test “to reduce the spread of this highly transmissible disease”.

The HSE adds that of 100,125 positive antigen tests reported by the general public between the beginning of November and December 21st (excluding HSE close contacts or schools’ programmes), 60,816 (60 per cent) were confirmed by positive PCR tests.

In a statement to The Irish Times, the HSE says it does not collect data regarding which day a close contact reports a positive antigen test result.

Commenting on potential confusion around mismatches between antigen and PCR test results, Pete Lunn, head of the behavioural research unit at the ESRI says: “You can’t compare the two results, and there is some concern that if people misunderstand this, it potentially makes the tests look like they don’t work, whereas they do.  A substantial proportion of the population don’t understand that antigen tests are not diagnostic tests.”