Taking a covid-19 test at home: the fragile base on which track-and-trace is built

Like millions of others every day I have “done my bit” by entering into an app my symptoms and whether I’ve had a test for covid-19, and day after day I had no symptoms and had not been tested. Then I developed epigastric discomfort and dutifully entered my symptoms into the app. Days later I was sent an email offering a test and thought I should accept as one point of the research is to identify all the symptoms associated with covid-19. So began a miserable process.

Now that I’ve written the account it’s become absurdly long—because there are many steps in testing and many places at which it can go wrong. As you may not have the patience to read all the way through, I’ll give you a summary now: I suspect that many results from tests conducted at home are wrong and that a fair few people must give up.

As you read this through (if you can be bothered) imagine an 85-year-old with limited education, little experience of computers, and slow access to the internet following the same process—or somebody with limited grasp of English.

I accessed the website to get a test and was told that I could go to a test centre or ask for a home test. I started by thinking that I could go to a test centre and so tried to book a slot. There were two slots available, both in the next hour on the North side of Hyde Park. It would take me about 30 minutes to cycle there, but I couldn’t go in the next hour. I tried booking one for subsequent days, but was told nothing was available for the next four days. It seems highly unlikely that tests were available in the next hour but not for the next four days. Perhaps I should have simply gone to the website later and tried again, but I didn’t. I ordered a test online.

Not one, but two tests arrived several days later delivered by Amazon. I opened one of the boxes and out fell a 12-page instruction booklet, a swab, a vial, a plastic bag, a sheet of absorbent material, a biohazard bag with a silver seal, and a return box. I learnt from the instructions that I had to post my test in a priority postbox an hour before the post was collected, and step 1 of 5, including a 3B, is to find your nearest priority box. To find the priority box I had to go online to the Royal Mail website, and I was confused because the link took me not to priority boxes, but to delivery/post offices. But then I saw where I could link to priority boxes.

There was a priority box close to where I live, but the instructions point out that if you are unable to get to a priority postbox then “you should not take the test at this time.” I received my test on a Saturday and there was no collection until Monday, so I had to wait. By Monday it would be about six days after I requested a test.

On Monday afternoon I sat down to take the test. In the morning I received an email saying that some people had been having difficulty registering their test. Step 2 is to register your home kit. I went to the link given in the instructions and arrived at a site with about 2000 words of information, but I could not find where to register my test. I tried various links, but they took me back to ordering a test. I also used the search function several times, but that never took me to the right place. I remembered the email that I had received that morning, and it provided a link that did take me to a place where I could register my test.

But this didn’t go well. I had to have my Order ID. This ID was included in the email I’d been sent saying I was being sent a test. I searched for the email and found it eventually. After answering a few questions I had to either scan a barcode or enter the numbers under it. I tried scanning, but I’d never done this on my computer and failed after several attempts of waving the barcode in front of my camera. The kit has two barcodes, and I entered the wrong one first. Eventually I submitted the right one. Then I was taken to a screen that worried me: it said that I had one kind of test when I had another. I tried entering the barcode several times, but it never worked.

Now I was reduced to ringing the helpline. I went through several of the usual “press-one-or-two” questions, and after about a wait of a minute or two I spoke to a woman. I explained the problem, and she said there had been a change in the system that day. We went back to the beginning of the process and eventually managed to enter all the bar codes. I asked whether I should take the test now, recognising that it was close to the cut-off time for the post.

“I’d wait until tomorrow, love. I’ll put you down for 9am tomorrow.”

“But the post isn’t collected until 5.30.”

“Ok, I’ll make it the afternoon.”

“Should we register the other test?”

“What other test?”

“I’ve been sent to two.”

“You shouldn’t have been. Just throw it away.”

“That seems a waste. Shouldn’t I do one test now and try to get it in on time and then do the other one tomorrow?”

“No, just throw it away.”

“Oh well thank you for your help.”

“Don’t worry, love, I’ve been putting people right on this all day.”

The next day with at least an hour to go before I had to post the test I prepared to take the test. You have to “gently rub the swab” over both tonsils for 10 seconds each. A diagram shows the site of your tonsils, and you are warned that: “This may be uncomfortable, and you may feel like gagging, but it should not hurt.” You are also told: “Take care not to touch the soft end of the soft end of the swab on anything apart from your tonsils and nose.” Perhaps if you have large, inflamed tonsils this is not so hard, but I had my tonsils removed in 1959. I found it hard, and I fear that I did touch both my tongue and my teeth. Next you must put the swab into a nostril and rotate it for 10-15 seconds. This was easier, but rereading the instructions now I realise that I forgot to blow my nose before inserting the swab as I was supposed to do.

Once I’d taken the test I had to insert the swab into a vial and break off the end. This was easy. The final stage was to insert was to insert the vial into the biohazard bag, pack the bag into the box, and post it. Assembling the box was not easy: the instructions were like those in an IKEA kit where arrows point everywhere and there is no text. I did it well enough, stuck it down with sellotape to be sure, and posted it.

I posted it on the Monday, and at 6am on Thursday I got a text that the test was negative. This was now more than a week after I’d requested a test.

The antigen test has a false negative test of about 30%, and I’m not surprised. The 30% figure includes tests conducted in hospitals and test centres as well as at home, and I suspect that the false positive rate from home tests is higher. To determine whether it is higher would mean letting people do the home test and having them professionally tested at the same time, which is probably not a priority. The much criticised track and trace system obviously depends on tests being done reliably, and I worry that home tests may give unreliable results—and I wonder how many people just give up.

Then why was I sent two tests? The government has set itself targets for the number of tests and been criticised for counting not successfully completed tests, but the number of tests sent out. Is that why I was sent two?

Any dataset is only as reliable as the quality of each datapoint, each test result, and I fear that with home tests the result may be unreliable. The whole track-and-trace system may have a shaky base.


Richard Smith was the editor of The BMJ until 2004.

Competing interests: None declared.