The epidemiologist Prof Neil Ferguson has said changes in the rate of hospital admissions suggests the spread of Covid-19 in the UK might be slowing down.
But can we be sure that is the case? And are hospital admissions the best metric to look at? To unpick this we spoke to experts about what the numbers tell us about the spread of Covid-19.
Number of new cases
Every day public health bodies in the UK report on the number of confirmed cases of Covid-19. So could this tell us about how the disease is spreading?
One difficulty is that, in recent weeks, only patients in hospital with suspected Covid-19 have been routinely tested for the disease, although testing is now being rolled out to frontline NHS staff.
“The only data we really have to go on at the minute is confirmed cases – there will be obviously a vast number of increased cases in the community that we simply don’t know about,” said Dr Michael Head, senior research fellow in global health at the University of Southampton.
That, he added, has a number of implications, besides not knowing the true number of cases. “We don’t really quite know the dynamics [for] transmission between minor cases, and what become more serious cases that end up being hospitalised,” he said.
Current UK data shows that over the past three days the number of new confirmed cases has dropped from 2,900 on 27 March to 2,400 new cases on 29 March.
While Head said that could reflect the start of a slowdown in the spread of Covid-19, he urged caution. “It is too early to really say if that is an actual trend or if it is just a natural variation,” he said. “As we go through the pandemic, cases will not increase at the same rate every day, they will fluctuate a bit anyway. So a steeper incline than the day before or a more shallow incline than the day before means little unless there is a sustained change in that rate of increase.”
Indeed, Head said it could be a couple of weeks before the new social distancing measures are clearly seen to have an impact on new confirmed cases because of the lag in the development of symptoms.
Changing the number of people tested, or who is being offered tests, will also affect the number of reported cases.
Hospital admissions
Talking to BBC Radio 4’s Today programme, Ferguson
said that the rate of increase in admissions had slowed.
While Covid-19 admissionsare not among the data released each day by public health bodies in the UK, it may be expected to track the number of confirmed cases since, as mentioned above, in recent weeks testing has been confined to patients in hospital.
Experts say hospitalisation data gives an idea of how many serious cases there are. But, even so, such figures may also have biases, said Wood, noting people may be less enthusiastic about going to hospital as the epidemic develops.
Prof John Edmunds, of the London School of Hygiene and Tropical Medicine, added the lag between changes in social distancing policies and new cases or hospitalisations means that any slight slowdown in the rate of the latter would be unlikely to reflect the impact of the lockdown yet, but more probably would reflect earlier interventions. “It takes that long to feed into the system,” he said.
Deaths
While death figures are updated daily, there can be a delay between a death occurring and it entering official figures.
And it could be several weeks before any impact of the recent lockdown measures are reflected in changes in death rates.
In addition, for deaths to be attributed to Covid-19, they need to have a confirmed infection. “The figures we release are for patients who have died in hospital with a positive test result for coronavirus,” a spokesperson for NHS England told the Guardian.
That, experts have said, means some deaths may not be recorded. “It is quite possible that other people are dying both in the community and in hospital – if they haven’t presented typically, maybe they haven’t had the test,” said Wood.
However, Wood noted that the biases in the death rate in hospitals are more likely to be consistent over time, potentially offering a more useful way to investigate how the epidemic is playing out in the UK.
From Tuesday, the Office of National Statistics will be publishing weekly data that includes deaths in the community.
Adherence to lockdown
Since the UK’s lockdown measures came into force on 23 March, a number of measures have recorded a decrease in public movement, from air pollution figures to statistics from apps such as Citymapper.
The latter reveals that, on 2 March 100% of London was moving as normal. By 23 March this figure was down to 25%, and by 28 March just 8% of normal journeys in London were being taken. “We can see where the major cities have much reduced movement, which is a measure of adherence to the social distancing measures that have been put in place,” said Head, adding that air pollution data may shed light on changes in the burden of other diseases.
But could such figures offer a different way to measure the spread of Covid-19? One problem is that they data is only available for certain cities, and does not cover rural areas. Another is that the data may not be fine-grained enough to give a detailed picture of how people are changing their behaviour.
Edmunds said his team have begun deploying other approaches to investigating people’s behaviour, by quizzing a panel of participant about their social interactions. “We are measuring people’s contacts,” he said.
The idea is that by doing so, the team can track changes in how many new cases an infected person generates on average – and get an idea of how quickly coronavirus is spreading far sooner.
