The reporting of coronavirus statistics needs a complete overhaul
‘If policy keeps intervening at the first sight of rising cases then we will never learn if any given intervention has worked. Policy should be less reactive’
On Oct 12, a new set of three-tier restrictions for England were announced, with the Liverpool city region being the only area put into the most severe “very high” risk tier.
Manchester has already been subject to a number of interventions, as many areas have, since the relaxation of lockdown rules in late May.
Yet, less than 48 hours after the initial introduction of such measures, the Government announced further changes. These included placing Greater Manchester into the “very high” tier of restrictions.
The approach was described as a new significant simplification of restrictions. Mayor Andy Burnham, however, came out fighting, defying the directive that forces Manchester further into lockdown.
The question on everyone’s mind is whether Covid is out of control in Manchester, or has Burnham got it right, and the Government should hold off? First, let’s look at cases, which don’t seem to be out of control, and are, if anything, declining. They peaked on Sept 30 with 596 cases and a seven-day average of 461. As of Oct 9, the seven-day average has fallen from the peak by nearly 20 per cent to an average of 374.
Nine out of the 10 Greater Manchester’s boroughs report recent increases in their Covid rates whereas Manchester’s are on the wane. But this further underpins the need to understand the context of what is actually happening.
The only critical cases are those with serious symptoms and those who are contagious.
Suppose the increase is due to healthy student-age people. In that case the increase is unlikely to be reflected in either admissions or deaths, especially in the student age group, which plays no part in national mortality statistics. Contagious youngsters are unlikely to pass the pathogen on after a week from symptoms onset and if they continue to test positive with few or no symptoms, it’s likely because they are shedding pieces of dead viruses.
Manchester University’s figures can help as the university keeps data on known positive cases among its staff and students. It includes those on and off campus but does not indicate where someone might have contracted the virus. These data also do not support further measures: cases peaked on Oct 2 – a spike driven by the return of more than 70,000 students – and have fallen significantly since.
They are 75 per cent less than what they were at the peak.
Hospital admissions may shed some light on the need for lockdown. Data from NHS England show the current seven-day average for admissions from the community is 12 a day to NHS Manchester University NHS Foundation Trust, up from an average of eight a day in the previous week.
But more telling is the data from the weekly- and monthly-mandated Secondary Uses Service repository for healthcare data in England. These show a dramatic reduction for respiratory condition admissions compared with what we would normally expect at this time of year.
Looking at all respiratory episodes at the trust we see that until week 15 the 2020 admissions matched closely with the four-year average for the corresponding week: average monthly activity was 96 per cent of the four-year average. However, this began to diverge significantly after week 16, when the monthly activity became only 57 per cent of the four-year average. Looking at emergency episodes alone shows a similar pattern: from week 17 they were at 60 per cent of the average.
The selective use of data is hindering policymaking. The data is all too often presented in the worst light, which informs no one. The Prime Minister said the situation in Manchester is worsening with every passing day. However, cases have decreased for nine days in a row. And yet Manchester is heading into its third month of restrictions.
If policy keeps intervening at the first sight of rising cases then we will never learn if any given intervention has worked. The time has come to follow the data, watch the trends and slow down the thinking. In doing so policy should be less reactive when faced with the daily diet of covid figures.
More vital now is placing the Covid figures in the context of other acute respiratory infections and what constitutes a benchmark for admissions at this time of year
The current reporting of data is confusing, incomplete and missing vital components. The better and more comprehensive the data is the more informed the policy will be.
A complete overhaul of the current reporting of Covid data is urgently needed. Meanwhile, let’s leave Mr Burnham alone.