Coronavirus: Lack of testing data may cause 'avoidable deaths', scientists warn
University of Oxford researchers have said the government is not providing complete information in relation to home test results.
Saturday 1 August 2020 11:42, UK
Scientists have warned that a lack of proper coronavirus testing data means opportunities are being missed to prevent local outbreaks and may cause "avoidable deaths", Sky News can reveal.
Dr Michelle Kendall and Dr David Bonsall, senior researchers at the University of Oxford, say the government is not providing complete information on the time it takes to get results back from home tests, even though the data "must be logged".
Turnaround times for home test kits are measured by asking people to declare what time their test was taken, not from the time the test was requested on the government website, so there is no data on how long tests take to reach people.
Without this "important information", say Dr Kendall and Dr Bonsall, it is hard for epidemiologists to identify local outbreaks quickly and accurately, as it is not possible to be sure exactly when infections happened.
"We have found that accurate estimation of R requires information on the timings of infections," said Dr Kendall, who leads a research project at Oxford's Big Data Institute tracking outbreaks in each of England's local authorities.
"The methods are quite sensitive and you can quite easily get the wrong picture if you estimate the timings incorrectly.
"Without better timing information we don't know the extent to which cases reported today reflect this week's epidemic or last week's."
The epidemiologists warned this "lack of clarity" could lead to "missed opportunities to tackle outbreaks" such as the ones in Oldham and Blackburn with Darwen.
"If we're to prevent a second wave, local authorities need to spot regional outbreaks quickly, but their efforts are currently undermined by delays in home testing and the way in which test results are reported," said Dr Bonsall.
"It's a serious problem that risks a resurgence and many more avoidable deaths."
Boris Johnson guaranteed that all tests would be turned around in 24 hours by the end of June, something the government says it is close to achieving.
But home test kits, which make up around a quarter of all tests, were not included in this promise.
Dr Kendall and Dr Bonsall said data on when tests were requested "must be logged within a central database that is used to return test results back to people as soon as they are available", but was not being released to researchers.
"It seems that the data are available, but they are not shared publicly," the pair said.
Official government guidance says the data is unavailable because "test booking and registration processes for home testing and satellite test centres are currently undertaken on different systems".
Dr Kendall and Dr Bonsall were especially concerned about the timing of home tests, saying that "the longer the delay for mail order tests the greater the risk that an outbreak might go undetected for a while".
The most recent government data suggested that in the week of 16 to 22 July, 76.9% of test results were received within 48 hours of the test being taken, but this figure did not include the delay between the test being requested and its arrival in the mail.
The epidemiological picture was further complicated, the scientists said, by the fact that the government did not provide regional data on how tests were taken, or separate data into pillar 1 and pillar 2, the names given to testing in hospitals and testing in the community.
Instead, the data was broken down by lower-tier and upper-tier local authorities, with combined daily counts from both pillars.
While this method of presenting the data stopped people being counted twice for tests done in different places, Dr Kendall and Dr Bonsall said it was "unhelpful for estimating the reproduction number, R, and for spotting local outbreaks".
They said: "This is because the regional data is currently a mix of people infected as early as three days ago, and as late as 12 days ago, with long delays for home delivery tests, and people presenting late to hospital. It's therefore not clear if today's data reflects this week's epidemic or last week's."
In a recent paper examining the impact of contact tracing on the Isle of Wight, the Big Data Institute group was able to obtain pillar 2 data separated from pillar 1 for the region "via a direct request to Public Health England".
It is not clear why this data is not made public, but a Department of Health and Social Care spokesperson said that it was providing increasingly detailed data to local councils, and that it would increase the amount of data during a regional outbreak.
However, one local public director of public health working to deal with an outbreak told Sky News that she had not seen any data on the timing of tests.
:: Listen to the Daily podcast on Apple Podcasts, Google Podcasts, Spotify, Spreaker
"There's nothing about wait time for tests or anything like that," said Lisa McNally, director of public health for Sandwell, adding that "the only field on this spreadsheet relating to date", aside from date of birth, was "something called specimen date", which she assumed was the date the test was processed at the laboratory.
Asked about these concerns over testing data, Dr Jake Dunning, COVID-19 incident director at PHE told Sky News: "It is entirely wrong to suggest that the correct data are not being shared to help local areas tackle outbreaks. There is a constant dialogue between us, Directors of Public Health and Local Authorities with routine exchanges of information and data so that we can support them throughout the pandemic.
"We will continue sharing appropriate data and intelligence to inform action at a local level."
A PHE spokesperson added that the organisation's priority was to tackle the pandemic, not aid academic research.
The Oxford team said their tool for tracking local coronavirus outbreaks had flagged the Sandwell outbreak several days before the factory at its centre closed, and identified a number of others similarly early, despite the difficulties caused by the data provided by the government.