The United Kingdom is flying blind on Covid-19

The British government's strategy relies on imperfect and patchy information, warns Anthony Costello. 

Sign Up

Get the New Statesman's Morning Call email.

The United Kingdom is flying blind into the worst pandemic for a century. In order to model with confidence an epidemic like Covid-19, we must know how many people are already infected with the virus and where the clusters are. Furthermore, we must calculate the transmissibility of the virus, called the reproductive rate of infection, R0, which for Covid-19 is probably between two or three times that of seasonal flu, with one infection on average causing somewhere between 2.5 and 3.8 others.

Crucially, one must also know the numbers of people in the population who have been infected without symptoms and become immune. We had a test for detecting the virus in nasal swabs back in early January. The Chinese scientists had already sequenced the genetic code of the virus in December 2019 and shared the data with the World Health Organisation (WHO). But no country has a reliable blood test for immunity yet, although candidates are being evaluated right now. 

The WHO recommendation for epidemic control, based on decades of experience, was simple. Test, test, and test again in communities to find any cases. Isolate them and all their contacts through 14 days of quarantine, and bring in appropriate social distancing measures. It's not rocket science requiring complex mathematical models. Just old-fashioned epidemiological detective work by public health teams. Don’t be like Wuhan where they had let the epidemic spread without tests and prevention measures for 4-6 weeks. Speed is of the essence.

Marc Lipsitch and colleagues, from Harvard and Nanjing universities, report that between 10 January and 29 February Wuhan hospitals faced on average 637 intensive care unit patients and 3,454 serious inpatients every day. Instead, be like Guangzhou, the largest province in China, where strict disease control measures were implemented within one week of case importation. Between January 24 and February 29, Covid-19 accounted for an average of just 9 ICU patients and 20 inpatients on each day. 

The British government did not listen. We had two months to get prepared for large-scale tests in the population, contact tracing and quarantine, to be like Guangzhou rather than Wuhan. Yet the advisory committee, comprising clinical knights of the realm, virologists, mathematical modellers, and behavioural scientists led by a former director of David Cameron's Behavioural Nudge Unit, decided British science knew best.

We don’t know if they ever sought advice from China or Korea, where population testing had been ramped up to over 20,000 per day. Or from epidemic control experts at the WHO like Bruce Aylward, Maria van Kerkhove or Mike Ryan, who continually emphasised the importance of speed and data collection in the community. Instead they relied on mathematical models with incomplete datasets.

On 12 March, at  Boris Johnson's press conference, Patrick Vallance, the chief scientific adviser, announced that they had moved from a “contain” to a “delay”. We were four weeks behind Italy’s epidemic, he said, so the modelling told us we could phase in social distancing measures gradually to flatten the epidemic. People with symptoms should stay at home for seven days. Population testing would be stopped (it's not clear if it had ever started) and 60 per cent of our population would become infected over the next few months building up “herd immunity”. Chris Whitty, the chief medical officer, said he didn’t think the mortality rate would be more than 1 per cent. The mainstream media swooned with their effortless, calm and professional authority. 

Even a Sun journalist could do the maths. 400,000 people might die. Many of us in the public health community were horrified and questioned the science.  Luckily, the whole policy unravelled over the weekend. Neil Ferguson’s team at Imperial College, one of their key groups, re-entered data from Italy into their models, which showed a higher proportion of hospital admissions required intensive care. The NHS would face meltdown, they concluded. And Adam Kucharski’s modelling group from the London School of Hygiene and Tropical Medicine, other key advisers, reported that by the time a single death occurs, hundreds to thousands of cases are likely to be present in that population. 

At a second PM press conference on 16 March, the whole strategy went into reverse: from the gradual mitigation and herd immunity plan back to intensive suppression of the epidemic. Social distancing had to start immediately. We still had time, said Vallance, and they hoped for no more than 20,000 deaths now. How many cases were there? They didn’t know. Around 1100 reported from tests, but it could be ten thousand, said Sir Patrick. Two days later, he suggested 55,000 possible cases, figures plucked out of the air without mathematical justification. And how were we doing with tests and case detection. Er…4000 per day, but we hope to get to 25,000. When? In a month. 

When several of us raised questions about the failure to plan for testing, and the apparent lack of epidemic control expertise, or WHO inputs, we were reassured the strategy was based on the “very best British science”.

Right now, the epidemic appears to be surging through London, Hampshire and the Midlands. Without testing we have few data to inform us. No hospital admission data has been presented. A junior ITU doctor wrote to me saying they had three patients on ventilators on 15 March, which rose to 20, including a health worker,  just two days later. Numerous reports criticised the lack of protective equipment in hospital. Quarantine rules told every health worker to stay at home for 14 days if they or anyone in their household had symptoms. No hotel accommodation had been requisitioned for those who wanted to avoid family contact, and no health workers were being tested. On this basis, the work force would quickly collapse.

We won’t know for another two weeks the severity of the casualties and our NHS capacity to cope. But we’re on the steep part of the curve. Will London be like Wuhan, Lombardy or Guangzhou? The main hope is that when reliable serology tests are rolled out they will reveal a large proportion of asymptomatic cases with an effective immune response. If so, this will change the death rate calculations and allow herd immunity to take effect. If not, we’re in trouble, facing a long crisis, and a new seasonal virus, nastier than influenza. If immunity is like flu, we'll need a new vaccine every time a new strain of Covid-19 emerges.

The post-mortem to dissect this strategic failure must wait. Was it advisory team group-think, led by clinician scientists with no experience of large scale epidemic control? Was it British exceptionalism, dismissive of Chinese and Korean science and capability, and condescension towards WHO “bureaucrats”.

China and South Korea can certainly teach us. A new study from Jiantao and York, Ontario universities modelled the impact of travel and border restrictions, and of enhanced epidemiological surveys and surveillance, contact tracing, school closure and other interventions that reduce social contacts. Travel and border controls were ineffective.  But “a coherent and integrated approach with stringest public health interventions” worked.  

Outside Wuhan, Chinese provinces had also set up 24 hour TV channels to keep the population informed. The government reassured the people that their utility bills, rentals and lost income would be covered so they could work from home without worry. Prevention messages were constantly repeated. Only supermarkets and pharmacies were open, carefully policed to prevent panic buying.  

At home, Public Health England was moved under local authority control in April 2013, and has been ravaged by local government cuts. We shall find out soon how our austerity-lite NHS will cope with the surge in cases. The OECD reports that the UK has just 6.6 ICU beds per 100,000 people, compared with Germany's 29.2, Italy's 12.5 and France's 9.7. Our hospital bed to population ratio in England is 2.3 per 1000 compared with 12 in Korea, 4.3 in China, 4.2 in Scotland, 3.6 in Italy and 2.8 in the USA. 

Now is the time for economic reassurance and support to the population, support and protection for our health workers who face great risks, stricter social distancing and lockdown in hotspots such as London and the Midlands. With great urgency we must re-establish community testing and contact tracing to quarantine household clusters in less affected areas. In the UK, we have one of the best primary care systems in the world, and together with local support, voluntary if necessary, they can identify cases and do contact tracing.

Above all, we must mobilise and communicate with our population through TV, the internet and through local governance structures and civil society. Daily press conferences from Johnson and advisers telling us what to do are simply not the way to change public behaviour. 

Anthony Costello is a paediatric scientist and Professor of Global Health and Sustainable Development at UCL. From 2015-2018 he was director of maternal, child and adolescent health at the World Health Organisation. @globalhlthtwit

Free trial CSS