At-home coronavirus tests may miss infections, study suggests

Tens of thousands of Americans who swabbed themselves for coronavirus testing may have falsely been told they were free of infection, a new study suggests

In an effort to expand stunted, lagging testing in the US, the Food and Drug Administration gave the green light for coronavirus testing using samples collected from low down in the nostrils, rather than from uncomfortable deep nasal swabs. 

It later determined that people could even collect the samples themselves, at home – so long as they are supervised by medical professionals via telemedicine – without sacrificing test accuracy. 

But scientists at the University of Utah analysed the accuracy of self-collected nasal swabs, like the ones used at drive-through testing sites and mailed to Americans as part of at-home testing kits, and found a different story. 

Results from lower nostril samples were compared to nasopharyngeal swabs – which are conducted by health care workers and are inserted so deep into the nose they can make people gag, their eyes water or even trigger nosebleeds – and saliva tests.

Scientists found only 70 of 86 out of 354 people who had Covid-19 were picked up by nasal swabs, which are more comfortable because they only require a gentle nostril swab. 

The nasopharyngeal swabs and saliva tests were not perfect either but did detect more cases – 80 and 81, respectively.

Although the researchers admitted it is not a huge amount of ‘wrong’ results and that the results were not statistically significant, experts have repeatedly warned that any missed case can have consequences.

If a person who is positive for Covid-19 receives a false negative result, they will not self-isolate and may spread the virus to other people.   

Nearly 54 million tests have been conducted across the US, and at least a tenth of approved tests can be done at-home, using shallow self-swabs. 

Professionally-done nasopharyngeal tests are known to be uncomfortable because the swab must go so far back into the person’s nose (Pictured: A man getting tested in Surabaya, Indonesia)

Public Health England instructions for taking a DIY home test kit

Public Health England instructions for taking a DIY home test kit

The tests people take themselves in the comfort of their own home are far less invasive than nasopharyngeal swabs – what is considered the gold standard by scientists. 

About 10 of the 100 FDA-authorized tests available in the US are self-swab diagnostics. 

In at-home kits, there are instructions to use a swab around the back of the throat as well as the nostril. 

This can make people feel sick or tickle, but the swab is not pushed deep inside the nostril to reach the nose ‘floor’ — as it would be with nasopharyngeal swabs.


Nasopharyngeal swabs are used to detect respiratory viruses, such as the flu and the new coronavirus.

It is the preferred choice for SARS-CoV-2 testing, according to the Centers for Disease Control and Prevention (CDC).

It involves inserting a long, flexible cotton bud into the nostril and along the nose ‘floor’. This is supposed to be done slowly so that it is comfortable.

The aim is to reach the posterior nasopharynx, a cavity made up of muscle and connective tissue, covered in cells and mucous that are similar to the nose. It continues down into the throat.

The swab is rotated several times in order to get enough cells.

The sample is then sent to a lab, where it will be tested to determine if the patient’s cells are infected with the virus.

The coronavirus is a RNA virus, which means it uses ribonucleic acid as its genetic material. A process called reverse transcription is needed to transcribe the RNA into readable DNA.

A swab sample doesn’t collect much RNA in one go, therefore a polymerase chain reaction (PCR) is used to rapidly make billions of copies so it can be analysed.

The DNA is dyed a fluorescent colour, which glows if the coronavirus is present, confirming a diagnosis.

Anterior nasal swabs, which are read in the laboratory using the same methods described above, can be carried out by the patient and use a shorter cotton swab. 

The swab only needs to enter the nose to a depth of roughly one centimetre before being rubbed around the nostril wall. 

It is less likely to make the patient feel sick or to produce a gagging or coughing reflex. 

Experts say they have less contact with the mucous membranes, which are areas of thin tissue inside the airways where most of the coronaviruses live. 

Experts say anterior nasal swabs have less contact with the mucous membranes, which are areas of thin tissue inside the airways where most of the coronaviruses live.

Instructions given to Britons shipped the postal tests say: ‘No force is needed and you do not have to push far into your nostril.’ 

One benefit is that because they are less invasive, they are unlikely to make a patient cough — meaning that healthcare workers are less likely to be exposed to the virus.

Professor Paul Hunter, an infectious disease expert at the University of East Anglia, said the study findings ‘certainly suggest these DIY tests are not as good and miss a proportion’.

He claimed around 10 per cent of coronavirus tests that would have been spotted by the nasopharyngeal swabs test would be missed by the at-home test. 

But Professor Hunter told MailOnline the study does not prove the tests are useless and called for larger trials of the swabs currently being used in the UK. 

He said: ‘This study is fairly small and it needs to be repeated using a larger number of tests, preferably in the UK, before we draw reliable conclusions.’

Testing in the US, and even studies of testing, have been plagued by uncertainty and flaws.  

The study, not yet published in a medical journal or peer-reviewed by fellow experts, was carried out across hospitals in Utah.

At a drive-thru test center, 354 people took all three Covid-19 tests; they were instructed to swab both nostrils, spit into a tube, and then had a nasopharyngeal swab taken by a medic. 

Some 268 people received a negative result across the board. Eighty-six people got a positive result either from one, two, or three or the tests. 

No singular test produced 86 results, proving that no test is able to always detect the coronavirus. 

Sixty-six patients had SARS-CoV-2 detected in all three tests, meaning they undoubtedly had the virus.

But 13 were detected in only two tests, and seven patients got a positive result from one test. 

It means that, had any of these 20 patients had a singular test, their result may have returned as negative. 

The seven who tested positively from one test were made up of two nasopharyngeal swabs and five saliva tests — meaning the nasal swabs did not pick up the virus on their own at any point.  

The results broke down how many results were given by each testing method, to find that anterior nasal swabs gave a positive result the least amount of times.   

The differences in results ‘did not reach statistical significance’ because the numbers are so small, the paper on medRxiv said.

However, the researchers wrote: ‘Relying on anterior nasal swabs alone could have missed infection in 10 to 11 patients compared with nasopharyngeal swabs or saliva, respectively. 

‘Missed Covid-19 cases have major clinical implications affecting isolation decisions for symptomatic 111 patients and are a lost opportunity for contact tracing.’

How nasopharyngeal swabs are conducted: Paediatric nurse practitioner Jessica Peck shared a diagram of how the coronavirus test is carried out on Twitter and said: 'This is how far back we have to put the swab to test you for #COVID19z. You might want to follow medical recommendations and #StayAtHome'

How nasopharyngeal swabs are conducted: Paediatric nurse practitioner Jessica Peck shared a diagram of how the coronavirus test is carried out on Twitter and said: ‘This is how far back we have to put the swab to test you for #COVID19z. You might want to follow medical recommendations and #StayAtHome’

How coronavirus self-tests are conducted: They require the patient to put the swab into the back of their throat and then a short way into their nostril to try and pick up the viruses, which live in the airways

How coronavirus self-tests are conducted: They require the patient to put the swab into the back of their throat and then a short way into their nostril to try and pick up the viruses, which live in the airways

When the researchers cross-referenced the tests with each other, they found ‘excellent agreement’. This means that, in most cases, the tests produced the same result – either negative or positive. 

Results showed if a nasopharyngeal swab was taken alongside an anterior swab then the results were the same 86.3 per cent of the time.

If a a nasopharyngeal swab was taken alongside a straight saliva test, there was a 93.8 per cent chance the results were accurate. 

The researchers said that the highest number of cases would be detected when combining nasopharyngeal swabs with a saliva test. 


Despite the concerns, scientific studies have suggested that self-swab testing actually is accurate enough to diagnose viruses.

A study by researchers at the University of California Los Angeles said self-tests showed ‘comparable sensitivity’ when trying to diagnose Covid-19.

This shows that both professionally done swabs and ones that people did themselves picked up about the same proportion of positive cases.

That study, published online on April 15, was done on 45 people, of whom 29 were definitely infected with the coronavirus and had already been diagnosed.

It found that the supervised self-collection tests detected 26 of those 29 patients, while unsupervised self-collection only found 19 out 29 (66 per cent).

Meanwhile the tests done by medical professionals successfully found 23 out of 29 positive cases.

In conclusion the scientists said: ‘Supervised self-collected oral fluid and nasal swab specimens performed similarly to, if not better than clinician-collected nasopharyngeal swab specimens for the detection of SARS-CoV-2 infection.’

Another study of generic infections in the airways found nasopharyngeal swabs were better performing than just a nasal swab.

Researchers at the University of Turku in Finland tested the two types on 230 children with chest infections.

They found that using both methods together – the deep nose and the nostril wipe – successfully detected 73 per cent of cases.

The nasopharyngeal swab on its own found 19 per cent of cases, while just a nasal swab found only seven per cent.

Their research was published in the Journal of Clinical Microbiology.

The researchers noted: ‘Nasopharyngeal swabs have historically been considered the reference method for respiratory virus detection.

‘In addition, anterior nasal swabs are used routinely for influenza nucleic acid amplification testing (NAAT).

‘Recurrent shortages of swabs and personal protective equipment (PPE), however, have prompted evaluation of alternatives to NPS including the use of patient self-collected ANS and saliva.’  

The Department of Health and Social Care refuted the findings, saying the evidence shows self-tests are ‘just as effective’. 

A spokesperson told MailOnline: ‘Instructions on how to perform these types of tests are included wherever individuals are asked to undertake self-swabbing.’  

Scientists first warned about self-tests being less accurate in June after Britain began started to ramp up its swabbing capacity at the end of April.  

Dr Andrew Preston, an infectious lung disease expert at the University of Bath, told MailOnline that shallower swabs in the nose and mouth were not as good.

He said: ‘It’s clear the deeper into the nasopharynx, the better it is picking up the virus.’

Dr Preston added: ‘I work a lot with whooping cough, and we tilt the person’s head right back. 

‘We consider it an unsuccessful swab unless the eyes water. We see real, real issues with the sensitivity of the swab if swabbing in nose.’

In a clear warning about how self-tests could miss the infection, he said: ‘The further back you go, the more chance you’ve got of getting the virus.’

Even nasopharyngeal swab tests carried out by professionals are estimated to be inaccurate in up to 30 per cent of cases due to variations in how medics carry them out.

This is based on scientists’ estimations, because the Department of Health will not release data about the false negative rates of its tests.

It is not clear how inaccurate self-swabs are, even though they are being carried out more than 60,000 times a day in the UK. 

Professor Jon Deeks, a biostatistics expert at the University of Birmingham, said: ‘A single negative test result doesn’t exclude the disease.’ 

It is believed that nasopharyngeal swabs are more reliable because they are better at picking up traces of a virus.

They go much deeper into the nasopharynx, a cavity in the airway which connects to the throat. 

Medics get a sample from along the ‘floor’ of the nose near the mucous membranes — areas of thin tissue inside the airways where most of the coronaviruses live.

Claire Cox, an intensive care outreach nurse working in Brighton, said in May that using the correct technique was crucial.

In a blog on Patient Safety Learning, she wrote: ‘Simply swabbing the inside of the nasal passage is not deep enough to verify that the virus is present.’ 

These swabs have become notorious for being uncomfortable because they require an extra-long cotton bud to be forced into the back of the nostril and rotated.

They can make people gag, make their eyes water or even trigger nosebleeds – but they are considered the most accurate way of diagnosing Covid-19.  


‘Is it in far enough?’

What sounds like a one-liner from a bawdy comedy, Carry On Covid perhaps, must have been heard up and down the country these past few weeks as people suffering coronavirus symptoms attempted to test themselves, either at a drive-through centre or at home.

It was what I said as my wife and I sat in our car this week, each of us feeling significantly under the weather and on a short fuse, as we tried to use the government’s swab test.

For those who have not yet had the displeasure, you have to put a swab like a giant cotton bud into your throat, slide it across your tonsils for about 10 seconds, then put it about an inch up your nose or until you feel ‘resistance’ and twiddle it around like an old-fashioned radio knob.

The instructions are in diagram form and seem straight forward even if the front seat of a car is hardly the sterile environment you’d want for such tests. It even helpfully suggests using the car mirror to see as you guide the swab in.

The throat part seemed pretty easy, if unpleasant and with a little gagging. But then came the tricky part – how do you tell if it’s gone far enough into your nose?

I teared up as I took the plunge and slid it up my nose before letting out one of the largest sneezes I’ve ever done. It was uncomfortable but no big deal.

My wife spun hers around her nostrils a few times, and said ‘That’s as far as it goes. My nose can’t be as big as yours.’

I was pretty confident I’d done it right but she had got it wrong. I was not thanked for pointing this out.

No one watched us do it to make sure we carried it out correctly. After all, an incorrect negative diagnosis could leave us walking the streets infecting people.

After about 36 hours, our results came back – both negative.

This seemed strange as we have most of the classic symptoms – loss of smell, fever, aching muscles, sore throat, nausea, breathlessness. Everything except the cough.

So we arranged through a private clinic for a doctor to carry out a similar swab test.

And what a difference. Who knew you could touch the back of your head through your nostrils?

My wife was wrong to think it would go no further. It went a lot further, although it took the doctor at least four attempts because of her ‘slim nasal passage’. Perhaps that’s the new thigh gap during lockdown.

He told her that he hoped he’d got a big enough sample, but was still unconvinced it had worked completely and she should rely on his clinical diagnosis of positive instead.

I was wrong to push the swab up my nose. It goes in horizontally. Not to the top of your nose where you feel blocked up when you’ve got a cold, but along another passage right to the back of the throat.

At least that’s where I felt it was. After all, I’m not a doctor.

It is not exactly painful, but astonishingly uncomfortable and goes against everything you think possible. At every second I was thinking ‘stop, stop, you’re going to burst something.’

When he wiggled it around I thought I would choke.

It scratched my insides, left me spluttering and coughing, and wondering how it had even been possible to achieve.

I defy anyone to do that to themselves successfully. Some people I spoke to after I had the drive-through test said ‘it’s simple, just took a few seconds.’ My answer to that is, you didn’t do it properly.