Official data on Covid-19 cases in HE in England

Since the start of term providers in England have been returning data, via the Office for Students, to DfE on cases and testing among students and staff. Now we get to see it.

David Kernohan is an Associate Editor of Wonkhe

Tags

There’s been around 64,254 confirmed cases of Covid-19 among students, and just under 8,000 confirmed cases among staff, reported to higher education providers this academic year. Of these, 55,239 student cases, and 4,376 staff cases were reported in the autumn term.

[Full screen]

But there are serious questions to be answered about the quality of this data.

Where the data comes from

Unless you are intimately concerned with making these data returns, you may well have forgotten that this was going on. We’ve certainly never seen any of the data, outside of the peculiar confidence with which Michelle Donelan has made statements about the number of cases among students.

On 20 October providers were asked to record confirmed cases reported to them by staff or students – confirmed meaning via a PCR test before 25 January and via an LFD test or PCR tests afterward. There was also space to return numbers of self-isolating students, the impact on case numbers on teaching (via those higher education tiers of restriction), and other contextual information).

Surprisingly, this data return was not compulsory – and there has been no attempt to control for this issue. Response rates were generally between 70 and 80 percent, dipping to below half for the first week of this year. The first weekly response in 2021 also includes all the cases notified over the winter break. The frequency of collections has also changed – after an initial (“baseline”) submission for cases between 1 August and their first submission providers returned data on every weekday, which shifted to twice weekly (Monday and Thursday) on 1 February.

What’s clear to me is why this data has never been published – there are so many caveats and discontinuities that its very difficult to say for certain what we can learn from it.

The shape of the pandemic on campus

If you’ve been tracking the Wonkhe Covid-19 dashboard – in particular the time-series plot, you’ll know the approximate shape of Covid-19 through this year. There was a peak in cases in areas where many students live in early October, linked to a large number of outbreaks reported in halls of residences, but in other ways these areas have behaved in a similar way to other areas.

This new data only shows cases from early November – so we do not see this singular point of interest regarding student cases.  DfE tells us that:

The first reporting week is 29 October to 4 November 2020. We do not present weekly totals for data collected between 20 October and 28 October due to data quality issues in the first few days of the collection. These were resolved with an update to the OfS COVID-19 reporting tool that was made available to HEPs from 27 October. Confirmed cases reported between 20 and 28 October are included in the grand total.

Collection issues

So, for those keeping score – there was a problem with the first two weeks of data, then a discontinuity, then another (the mass availability of LF tests, leading to more active use of PCRs to confirm), then another (the increased demand before the “Student Travel Window”), then it was Christmas, then a smaller but uncertain proportion of students returned in January, then there were difficulties about the earlier weeks of the January returns, then the rules about which tests count changed, and finally the collection frequency was changed. All the while the returns were optional and there was no attempt to control for this.

DfE would have been better off using the Wonkhe dashboard. Or, frankly, chicken entrails.

There were always going to be issues about incomplete coverage, but the Welsh government appear to have made a decent fist of it. It particularly alarms me as I seem to have spent a good few months arguing with people who believe – for whatever twisted reason – that the return of students to campus somehow “caused” the second wave of the pandemic. It would be good to have the government on record putting this conspiracy theory to bed.

The data

If you were hoping for data disaggregated by provider I think at this stage you’re probably grateful this isn’t provided. Instead we get data overall for England for cases (and on the tab) for isolation.

[Full screen]

And the tiers! Remember the higher education tiers? We get data on which tier providers were on for each week through to 12 December, beyond which the tiers stopped being a thing.

[Full screen]

By 12 December (after the student travel window had closed), 16 providers were at Tier 4 (wherein provision was supposed to be online only) and 25 were in Tier 3 (broadly the same restrictions as currently exist). The data shows gradual increases in providers reporting being in higher tiers.

One response to “Official data on Covid-19 cases in HE in England

  1. Good analysis as always. I think the key point here – which will be lost (deliberately withheld??) from the public – is that these are cases reported by university students and staff, not people who caught the virus on campus or in the pursuit of their studies/duties. I can tell you that at my institution an infinitesimally small number of our ‘cases’ are from people who had been anywhere near the campus during their infectious period. And an even smaller percentage still reported any possibility that they could have caught the virus on campus.

    In other words this is just the number of people who are ‘associated’ with Universities who have tested positive. From a certain point of view it’s about as valuable as knowing the number of people who support in any given city.

Leave a Reply