funded by our supporters
Failing the test: Slow start and flawed decisions in Britain＊s coronavirus testing have cost lives, warn health leaders
Investigation: A fatally slow start. An opaque decision to build a new lab network. A High Court battle over ＆flawed＊ software. And when the Health Secretary declared success in hitting his testing target, lab staff simply felt ＆upset＊ at the blatant boost from home tests. Now this system is being expanded at a cost of billions. Insiders tell Shaun Lintern how failures have already cost lives, and why urgent improvements are needed before a second wave
Lives were lost because of delays to Britain＊s expansion of testing and the decision to create privatised mega-labs to process swabs, NHS leaders and health experts have told The Independent.
An investigation funded by The Independent＊s Supporter Programme has uncovered widespread concern over the ※Lighthouse Lab§ system for processing coronavirus tests, which was not fully functioning until late April 每 weeks after Britain＊s Covid-19 peak.
Three national leaders in pathology warned NHS bosses in a letter seen by The Independent that the strategy would cause problems that would ※inevitably cost lives§ yet were ※wholly avoidable§. They called for changes before a second wave of the virus hits.
They said privatised labs were often taking 72 hours from the time they received tests to determine a result 每 by which point the results were of no use for wider strategy or policy. By contrast, they said local labs could give results in six hours from the point the test is taken. Our investigation also found that:
- Rather than being able to identify and contain outbreaks using test information, local public health directors say they have been ※sidelined§ with access only to minimal aggregate data until the start of this month, by which point the UK death toll stood at more than 42,000;
- A legal challenge claims that more than 1,000 tests a week from the labs are false negatives because of issues with the system＊s accuracy;
- Experts including Sir Paul Nurse, Nobel laureate and director of the Francis Crick Institute in London, have branded the decision to commission the laboratories 每 funnelling millions to private companies 每 a tactical mistake that was made too late, without consultation and remains shrouded in mystery;
- NHS staff have reported waiting for as long as seven days for test results;
- Staff working in labs believe standards varied hugely between sites, with significant proportions of some batches of tests discarded;
- With a feared second wave of Covid-19 possibly months away, patient-level test results are still not being passed to local health officials.
Despite the warnings, the government is understood to be pressing ahead with its Lighthouse Laboratory scheme with plans to create seven more across the UK. Earlier this month, Public Health England issued a contract notice worth ?5bn to ※appoint suppliers§ to provide diagnostics, research, supplies and ※provision of testing capacity§ not just to Public Health England and the Department of Health and Social Care but also to NHS trusts and health authorities in Wales and Scotland.
Sources told The Independent this was part of the second phase of testing expansion which would see the regional labs established across England. The contract notice said the framework would be in place for two years but could be extended 每 adding that it had been designed to support ※all five pillars of the UK testing strategy set out by the secretary of state for health and social care§.
Race to increase testing
The need for a comprehensive testing system to fight coronavirus once it reached Britain was clear to experts advising the government as early as January. Minutes for the government＊s Scientific Advisory Group for Emergencies, Sage, show testing had been highlighted as a problem in Wuhan that month. Discussions over the coming weeks emphasised that scaling up diagnostic testing to manage the epidemic was of ※critical importance§.
But by mid-March, as the World Health Organisation urged countries to ※test, test, test§, Public Health England could carry out only about 4,000 tests per day, with its goal to reach 10,000.
As the country headed into lockdown on 23 March, about 5,000 tests a day were being carried out, amid a global shortage of chemicals vital to the tests, as well as lack of swabs and consumables. Critics pointed to Germany, which was then nearing 100,000 tests a week.
Inside Downing Street, officials lost patience with Public Health England and discussions were held involving No 10 advisers such as Andrew Gilligan, as well as Wellcome Trust director Sir Jeremy Farrar and Sir John Bell, who are both members of the Sage group.
On The Andrew Marr Show on Sunday, Sir Jeremy admitted mistakes had been made in the UK response, saying: ※We got behind it, we were too slow, and as a result the epidemic took off and we weren＊t able to control it.§
One senior government adviser, with knowledge of the decision to set up the Lighthouse Labs, told The Independent there had been ※a lack of strategic thinking§ by government earlier in the process, in regard to the scale of testing that would be needed.
They said there was ※linear growth§ in capacity planned by Public Health England and NHS labs when ※exponential growth was needed§ and a ※call had to be made§.
It was decided that a centralised approach would be more efficient and give greater control, rather than mobilising NHS or academic labs, and funding them to expand operations. The ability to secure the supplies of chemicals and swabs needed to make the system work was described as a major factor in that decision.
Two months after the initial warning from Sage, this meant a race against time to set up the testing programme for the main part of the crisis 每 and to help meet Matt Hancock＊s target of 100,000 tests a day by the end of April.
※Was that the right thing to do? What was achieved was pretty miraculous,§ the adviser said. However, the rapid improvement came after a significantly delayed start.
Damage already done
The Lighthouse Labs were brought in as part of ※pillar two§ of the health secretary＊s plan to ramp up testing. While pillar one 每 involving the NHS and Public Health England 每 meant carrying out 25,000 tests, the Lighthouse Labs were designed to boost mass testing for the population as a whole, with tests at drive-through centres and at homes.
The first Lighthouse Lab in Milton Keynes was launched on 9 April at the offices of the UK Biocentre, a not-for-profit business established in 2011. The second opened on 20 April at Alderley Park in Cheshire, run by Medicines Discovery Catapult Ltd, funded by Innovate UK, an agency of the UK government.
A few days later, a third lab was opened at the University of Glasgow, located at the city＊s Queen Elizabeth University Hospital campus. These were later followed by a private lab in Northern Ireland run by Randox, and a lab in Cambridge run by pharmaceutical giants AstraZeneca and GSK.
Ministers have hailed the speed at which these labs were established, citing an unprecedented effort. But experts told The Independent the damage had already been done by this point. The system was established from a standing start, meaning the labs had been unable to handle tests during the peak of the virus around 8-9 April.
Professor Colin Fink, from Micropathology Ltd, a leading UK laboratory serving more than 200 hospitals for more than 25 years, said: ※Governments of both colours over the years have shot themselves in the foot, because when the last coronavirus crossed species there was a warning then that we were sitting on a time bomb and nobody in this country prepared for this and they should have done. The whole infrastructure was missing.
※It was a failure compounded by them not thinking about it early enough, which they should have done. And not even investigating whether labs like ours could do more testing.
※I think the government was in an impossible situation, the NHS labs have been downgraded terribly over the years. Public Health England should have been able to do much more testing but failed totally and so the responsibility was left really up in the air. The only option the government had was to set up these Lighthouse Labs. But it was a bit like hiring a horse when you could have had a Rolls-Royce. It＊s been a coterie of incompetence.§
He added he ※definitely§ believed the problems had led to lives being lost, pointing to the decision to stop testing in the community and care homes in March as the epidemic began to reach its peak. NHS data shows more than 25,000 patients were discharged to care homes between 17 March and 16 April, before routine testing was available.
Professor Alan McNally, from the Institute of Microbiology and Infection at the University of Birmingham, volunteered at the Milton Keynes Lighthouse Lab for 12 weeks. He said Public Health England and NHS labs alone would have struggled to carry out the amount of testing needed, but he believed academic labs and medical schools could have been mobilised quicker.
He said: ※What was pivotal was how long it took for us to get an effective testing strategy. Why did it take until late February, early March to think we＊re going to have to test more than we currently can? And then why did it take so long to make a decision on how that was going to be done? It doesn＊t matter if it was going to be the academic route or the lighthouse route but the decision had to be taken much earlier.
※If the Lighthouse Labs had been operating four weeks earlier it would have made a huge difference. There＊s no single way of controlling a pandemic. But the most important in my opinion is test, trace and isolate 每 and we didn＊t have any of that in place.§
The decision to create new, privatised labs rather than expand existing systems was made without consulting industry bodies. And the companies handed contracts for the project had little or no experience of this kind of mass diagnostic testing. The Department of Health has pointed to the experience of scientists and volunteers involved, describing ※hundreds of years§ of experience in testing.
Meanwhile, London＊s Francis Crick Institute had developed its own testing process and was supporting the NHS in north London, completing hundreds of tests a day.
Its director, Sir Paul Nurse, told The Independent he believed it had been a mistake for the government not to mobilise local laboratories across the country.
※I reached out to Downing Street early on but it seems that the local route was not even considered,§ he said. ※That was a tactical error in my view, because it was self-evident from the beginning that a locally managed solution would have been effective.
※It was needed until the big labs got going, which was going to take time given the lack of preparation. What we did at The Crick could have been done and activated at university and medical school labs across the country, which were dormant because of lockdown.
※Our local Crick lab can turn around tests in 24 hours, even under 12 hours. In these big labs it＊s been as long as five days 每 that is next to useless. There was a failure to think creatively about how to deal with testing and the decision to set up these labs has been shrouded in mystery, at least to me. Who made the decision? Why was it made? Who advised on it? What did it cost?§
On 5 March, the government made coronavirus a notifiable disease 每 meaning doctors, GPs and, crucially, diagnostic labs were required by law to report cases to Public Health England along with patient details, to help manage the epidemic.
But, under the testing system, which is overseen by accountancy giant Deloitte, swabs are sent to the Lighthouse Labs with a bar code, but no patient data. The result of tests for each bar code is then passed to NHS Digital, which marries this with patient information before sending it to Public Health England, which in turn should then be able to distribute key data to GPs and health directors 每 though experts told The Independent that this fragmented approach had caused issues in passing on information.
Questions have also been raised about whether it was sufficient to comply with health protection regulations relating to notifiable diseases, which suggest the lab should make direct contact with Public Health England.
Jason Coppel QC, a leading public law barrister, said: ※The Lighthouse Laboratories are under a legal duty to notify Public Health England of a positive test for Covid-19, which is a notifiable disease under health protection regulations. Failure to notify Public Health England without reasonable excuse is a criminal offence.§
Mr Coppel, who is also representing the firm taking legal action against the labs over its testing system, added: ※What is troubling from a regulatory standpoint is that it has been thought necessary to impose a legal duty on laboratories which 每 it would appear 每 they are not complying with themselves, leaving Public Health England to rely on notification by other organisations who are not under the same legal duty.§
Hugh Risebrow, consultant and former UK chief executive of Synlab, a leading European laboratory group, said the decision not to expand existing labs, staff and information networks had caused delays and ※inevitably cost lives§.
He added: ※Given that there is a robust structure of NHS labs with capacity to do testing in terms of equipment, staff and the IT connections to get test results back to GPs and public health directors, I am surprised the government didn＊t decide to leverage that capacity more, rather than create completely new labs from scratch, with all of the risks, complexities and delays that this inevitably involves.§
On 15 May, three senior NHS clinicians who lead national efforts to improve pathology services were so concerned about the operation of the Lighthouse Labs that they wrote a letter to health chiefs setting out their fears.
Dr Tom Lewis, Dr Marion Wood and Dr Martin Myers sent their concerns to Professor Tim Briggs, who leads the NHS England Getting It Right First Time programme, and Jo Martin, president of the Royal College of Pathologists. They warned that the lack of data on patient identification meant it was ※impossible§ to match patients to their NHS records, adding: ※The impact of this on local healthcare delivery cannot be overestimated; patient safety may be compromised by the lack of connectivity to the NHS systems and structures.§
The letter said NHS labs could provide a faster, more integrated service, continuing: ※The current turnaround times of the central labs is stated as being 72 hours from arrival in the laboratory. This is too slow to run an effective track and trace operation, even if the positive cases could be accurately identified. There are numerous reports of results never being reported back to the individual, or being received weeks later.
※We believe that the current national testing strategy is leading to problems in delivery of both acute care and the public health response, that will inevitably cost lives, and which are wholly avoidable.§
Asked to comment on the letter, Dr Lewis, an NHS microbiologist and national lead for pathology in NHS England＊s Getting It Right First Time Programme, said: ※I think out of 100,000 swab tests performed each day, about two-thirds have been done outside hospital labs. These results have been too slow to be useful clinically, and they have not helped us direct our efforts.
※If we believe that testing is important in reducing the impact and ultimately saving lives, the fact this testing has been sub-optimal means we have to learn lessons. So now is our chance to take stock, and we＊ve probably got three months to get this right before a second wave.§
In May, the Royal College of Pathologists said using existing NHS labs would ※reduce the risks of fragmentation and overlap that are likely to follow from sending specimens to remote laboratories§.
Earlier this month, the college set out a roadmap for improving the UK＊s testing strategy, warning testing was not something to be ※just done and counted§.
It added that problems included ※poor specimen-taking, poor labelling or poor transcription of details, slow turnaround of results, poor quality control, ineffective communication of the result, inappropriate application of the result, and lack of clinical input or oversight§. The college continued: ※Many of these issues have been seen in recent times, all of which must be urgently addressed.§
It said data must be available locally and must be included in medical records, stressing: ※Testing standards must be upheld.§
Data black hole
Local officials trying to control regional outbreaks say an ongoing problem has been the lack of access to detailed patient data to help them identify potential links between virus carriers.
One said: ※As a public health director, I was absolutely sidelined. The data came in at the start of June and what that showed was many areas had different rates of transmission than many of us had appreciated.
※We were getting frustrated with Public Health England, they were frustrated with the data they were getting, and it was also potentially dangerous, because we were not sighted at all on any likely outbreaks from that data and this issue is still very much a live one.
※We are still not getting the data records. We are getting summary data only, just numbers tested and confirmed and not access to the actual cases.§
They added that key details such as workplace or place of worship had not been routinely recorded by the test and trace phoneline, NHS 119, which was crucial for spotting outbreak clusters.
The local official said: ※One of the failures in the system has been to agree a shared data infrastructure between NHS 111 and 119 where patients enter the system, then between the testing bodies including Public Health England and local hospitals, but particularly the Lighthouse Labs and local authorities.§
Public Health England said it had provided some patient-level data to directors of public health, but acknowledged that a greater level of detail became available only this month.
Insiders at the labs suggested that these data issues were further compounded by the problem of double counting of patients, when a nose and throat swab were counted twice. They also said some patients have also had to be retested because the original swabs did not produce a valid result, often because the first test was not carried out properly.
Public Health England said it had been receiving data since April and that its health protection teams had shared patient information with local public health directors in some cases. It added dashboards had been created so officials could view the number of tests completed and positive cases in their areas.
Allan Wilson, president of the Institute of Biomedical Science, said these issues could have been avoided if the government had used existing labs and invested in them to create extra capacity.
He said: ※That would have maintained IT links so the data would have been robust and the reporting systems of the results would have been robust, and we would have one source of data rather than the complex sources of data we＊ve got at the moment.§
He said the decision to set up the Lighthouse Labs cost lives, saying: ※An unavoidable conclusion to come from this, is we would have reduced transmission of this disease if we＊d have gone into lockdown earlier, mobilised testing earlier, and used it in a more targeted way. Then we would have reduced transmission, and it＊s hard then not to say that would have led to a reduction in deaths.§
Dr Jeanelle de Gruchy, president of the Association of Directors of Public Health (ADPH), said officials had been working with national agencies to ensure public health teams had ※consistent access to the high-quality data they need, including test results, to carry out their responsibilities§.
But she said: ※The ADPH continues to request access to pillar-two data for all directors of public health. Timely, high quality and consistent data flows will be a key tool in implementing the local outbreak plans that directors of public health are responsible for.§
Inside the labs
Insiders at the labs said staff worked hard and delivered incredible results in establishing the labs in such a short period of time. But they also revealed the scale of teething problems and issues that some said could have been avoided if the government had used existing labs and expertise.
One worker said: ※That＊s been a source of some considerable controversy between the Lighthouse Labs. At Milton Keynes there was a real desire to test every test. We would print our own bar code. We really tried hard to test every possible tube, I know at the other Lighthouse centres they would just get put in the bin.
※The void rate at Milton Keynes was 1 or 2 per cent, whereas in the other ones, it was close to 40 per cent at one point, but it was down to 20 per cent at the start of June. There have been very big arguments between the labs on that.§
Medicines Discovery Catapult disputed that the void rate was as high as 40 per cent on some batches, but declined to provide its own figures.
Another member of staff at the Alderley Park site said that issues with barcodes meant the tests were often ※unusable§. They cited various basic problems: barcodes stuck on in the wrong direction, meaning they could not be read in the labs; damaged barcodes; and barcodes attached to the wrappings rather than the tubes themselves. These issues were raised ※time and again with Deloitte§, they said.
Further delays were apparently caused by tests being sent between labs after technical faults. One worker said a lot of tests were sent back to Milton Keynes, adding: ※People talk about it taking five days to get a result 每 that was why. It did put a spanner in the works.§
Another added they were ※flooded§ with some test samples with barcodes ※that were not scannable 每 you couldn＊t link the patients to the barcode, it was a nightmare§.
Ministers admitted 67,000 tests were sent to the US for processing, with more than 30,000 later ※voided§, meaning tests had to be repeated.
When health secretary Matt Hancock triumphantly announced at the end of April that the UK was carrying out 100,000 tests, staff in the labs said they were ※hurt and upset§ at the obvious use of postal home-test kits to increase the numbers.
Professor Alan McNally, who led shifts at the Milton Keynes centre, told The Independent that, despite the challenges, the Lighthouse Labs were a success but ※could have happened much quicker§.
※All of the hurdles the Lighthouse Labs had to jump through, there＊s no reason why academic labs couldn＊t jump through those loopholes as well,§ he said.
※It was a huge challenge. Within a couple of weeks we had 250 volunteers and we were trying to run thousands of tests a day. It was the most stressful thing I＊ve ever done in my life. There were a couple of times that I thought about driving away from it. But the volunteers were what made it happen, they were incredible.
※I have no idea why [the testing] was based there, it was not an obvious choice, they had no expertise in infectious disease diagnostics so it took longer to scale up, but the speed [at which] we scaled up was phenomenal.§
Test errors and legal battles
The Lighthouse Labs spent millions of pounds sourcing equipment, machines and software to carry out tests and analyse the results. One such sourcing decision is now at the centre of a High Court battle and judicial review.
The Independent can reveal that more than 1,000 test results a week at the Lighthouse Labs could be wrong because of a system chosen to analyse test results.
The two companies behind the Lighthouse Labs in England 每 Medicines Discovery Catapult Ltd and UK Biocentre Ltd 每 are facing legal action alongside a judicial review brought against health secretary Matt Hancock, over the decision to award a contract in April.
High Court documents show Diagnostics.ai Ltd is suing over the decision by the labs to appoint rival firm UgenTec in April, despite an evaluation report by UK Biocentre showing the UgenTec system failed to spot some positive tests while the Diagnostics.ai system made no mistakes.
According to the court claim, the UgenTec system failed to identify three positive Covid-19 tests in a sample of 2,000 carried out by UK Biocentre. It said: ※All of the relevant samples were correctly identified as positive by the claimant＊s solution. If 100,000 Covid-19 outcome tests were performed each day, then based on the ad-hoc testing results it would be reasonable to expect that 150 positive Covid-19 cases per day would be falsely diagnosed by the UgenTec solution as negative.
※This would equate to in excess of 1,000 Covid-19 test results each week being falsely diagnosed and reported as negative.§
It added the procurement and evaluation ※was flawed and not fit for purpose, such that it poses a significant threat to the success of the UK Covid-19 national testing programme and NHS patient safety§.
The court claim also suggests Diagnostics.ai Ltd system has been validated by published clinical research and meets standards set by the National Institute for Health and Care Excellence (Nice).
According to the court documents, the UgenTec system has not been validated by published research and does not meet the Nice criteria.
Dr Simon Clarke, associate professor in cellular microbiology at Reading University, who has no links to the case, told The Independent: ※It is reasonable to think this would mean 1,000 people could be missed if these numbers are correct. That is of concern as it would mean 1,000 people not being subject to contact tracing and able to circulate in the community and spread the virus.§
In a response to the claim, UK Biocentre, which carried out the evaluation, said the allegations did not constitute breaches of the law because in its ※scientific opinion§ the three positive samples that the UgenTec solution failed to diagnose as negative were ※borderline§ cases. UK Biocentre said Diagnostics.ai had used clinical input to refine the accuracy of its system, so the testing results were not capable of ※fair or reasonable comparison§ and had UgenTec had similar input, it ※may§ have spotted the tests.
Lawyers for Diganostics.ai said these arguments were not part of the evaluation report and were inadmissible justifications made after the fact. The case is likely to go to court in September.
UgenTec chief executive Steven Verhoeven said the suggestion its software had made errors was ※incorrect§.
The Department of Health refused to comment on the legal action but said the UgenTec software had been used for several months and was subject to quality assurance processes, though it did not give any further details.
When approached for comment on our investigation, a government spokesperson said: ※The Lighthouse Labs form part of the biggest network of diagnostic laboratories in British history and has been set up in a matter of months.
※The UK has one of the largest testing programmes in the world and it draws on the expertise and resources of a number of public and private sector partners which has undoubtedly curbed the spread of the virus and saved lives.§
Read the rest of our Failing the Test investigation:
This article was funded by our supporters. If you would like to help fund more investigations, please make a contribution