Eight things we learned from SAGE and DfE papers on students and Covid

The Department for Education (DfE) has started publishing official data on Covid-19 cases in higher education in England, which David Kernohan has been plotting over on Wonk Corner.

DfE has separately published a summary of evidence about Covid and children, young people and education settings – although there’s not much in that on HE.

And the Scientific Advisory Group for Emergencies (SAGE) has published a collection of papers and minutes to accompany the roadmap decision – which includes a major new evidence round-up on Covid-19 and higher education settings, and another on the risk of increased transmission from student migration posed by the return to campus.

The evidence round-up is important, because the paper and the discussion on it in committee is really the first time since August that SAGE has considered higher education specifically in the round. There’s eight big messages for ministers in there and the migration paper – and not all have been heeded.

1. Background and context

As well as HESA stats that we know about, there’s some fascinating background numbers in the paper. According to NUS, 62 per cent of students undertake employment, with 24 per cent in part-time employment, 13 per cent with zero hours contracts, and 12 per cent in full-time employment – and 87 per cent have had to make adjustments to their work since the start of the pandemic, for example via furlough (18 per cent), unpaid leave (14 per cent), a reduction in hours (12 per cent) and redundancy (11 per cent).

We also learn that data supplied to SAGE last Wednesday from 8 universities in the Russell Group in England suggested that between 50 and 85 per cent of students have returned to campus or will not be travelling (especially those who are remaining overseas) – which probably tells us as much about how clueless we really are about students’ locations than it does about anything else.

SAGE is also told that there appears to be variation in the amount of face-to-face learning that took place during October to December 2020, with one university reporting 35 per cent of students being taught face-to-face, and one reporting 85 per cent. We don’t know whether that means “offered” teaching or students attending teaching – perhaps the Russell Group could publish for everyone else to see.

2. Infection was a problem last Autumn amongst students

What is clear is that rates of Covid-19 infection rose among many higher education student populations in October 2020, and rates subsequently reduced in November.

To draw this conclusion SAGE triangulates three lots of data – Office for National Statistics (ONS) surveys (which we’ve looked at on the site a few times), Public Health England (PHE) data (which DK has been keeping on top of on the site), and evidence from COG-UK, which is a consortium that looks at stuff like genome sequencing to provide the UK with insights and data on the virus.

What’s particularly interesting about all of that is the extent to which there are holes in the data. The DfE published numbers that come from provider submissions to OfS are riddled with problems. Weekly ONS data have never included those in “institutional” settings so have always excluded students in halls. PHE testing data has published results by age group – so gives a clear picture on those of school age, and not to clear on those over 18.

The paper says:

Assessments are limited by the lack of representativeness of several data sources, including the exclusion of those living in student residences from ONS COVID Infection survey data, and PHE data by educational age cohort excluding mature students, and including young people (aged 18-22 years) who are not enrolled in HE. Evidence on the amount and approach to face-toface teaching over the period are lacking. Further evidence, including on the differential approaches to mitigation measures applied across the sector or the level of compliance within different institutions would help inform mitigation actions.

Amazingly, SAGE doesn’t even seem to have access to DK’s excellent analysis based on student heavy postcodes, which from here looks more useful than the stuff it has been looking at.

All of that said, the first peak of infection among those attending clearly coincided with the timing of when institutions opened and closed, and there were certainly several case studies of individual outbreaks and/or transmission in higher education settings among students in late September – October 2020.

3. Halls were the problem

As confirmed in other reports (notably one from Scotland), evidence from ONS outbreak investigations, PHE surveillance data, genomic and antibody studies in a number of higher education settings suggest a higher risk of transmission in residential settings, and particularly in some halls of residence.

ONS outbreak analyses from Exeter and Loughborough found greater transmission spikes in halls of residence than private accommodation, and antibody studies from two universities suggest a higher prevalence of antibodies among those living in halls of residence.

Minutes from the 18th February tell us that where transmission is not well controlled, a very high proportion of the population can become infected – something SAGE has seen from university halls of residence and care homes in the UK, as well as some international evidence.

What is described as “limited, anecdotal” evidence from 10 universities presented by ONS suggests that when face-to-face learning was happening, minimal cases of transmission were attributed to face-to-face learning environments. Instances where transmission did occur were associated with guidance not being followed, for example, the removal of a face mask, rather than systemic failures.

And evidence from genomic studies in a limited number of universities suggests that mitigation measures were successful in minimising transmission. We’re talking environmental and behavioural infection control measures, increased testing, use of remote learning, and self-isolation by students all appearing to have been effective during the small amount of face-to-face learning over the Autumn term, and may have contributed to reducing prevalence in student communities.

4. We don’t know if students infected communities (or vice versa)

There’s naturally been lots of debate on whether students and their return to campus “caused” an acceleration of infection rates in communities.

This paper concludes that whilst genomic data indicates significant student-student transmission but only limited evidence of potential onward transmission to the local community, the evidence comes from a limited number of higher education settings, and it is unclear how generalisable the results are across the sector.

Not all universities are Cambridge, basically.

5. Students are at less risk

We’re more sure about the risks to students posed by the disease – although that’s really about using age as a proxy for student.

The paper notes that 11 per cent of students are aged 25-29, and 19 per cent are aged 30 and over, and as a result those groups have underlying conditions that make them vulnerable to Covid-19.

6. There were clear wider impacts

SAGE doesn’t only talk about physical health – and it also reminds us that there was disruption to research and learning, lower wellbeing, and increased mental distress in students.

Data from the ONS Student Insights Survey highlighted here found that a greater proportion of students reported being dissatisfied or very dissatisfied with their academic experience since the start of the autumn term, with limited opportunities for social or recreational activity, meeting others, and access to sports and fitness facilities driving dissatisfaction.

7. Take up of tests

As well as the general evidence round up, there’s a separate paper from the behavioural science sub committee. It’s not especially illuminating – we know that university student interaction in highly-connected environments makes students susceptible to higher rates of transmission, and we know that large scale randomised testing, contact tracing, and quarantine underpin successful infection control strategies for containing campus outbreaks.

It’s what the committee doesn’t know that’s miserable. Where it concludes that additional work is needed to understand the costs, feasibility and acceptability of universal asymptomatic testing in universities, including more diverse student and staff populations, you’re left thinking – why didn’t DfE set up some research around the massive national pilot that we ran in December?

8 Students (still) need support to self-isolate

There’s a whole section in the SPI-B paper on support for self-isolation that says that higher education institutions should put in place strategies to support students who are required to isolate. There’s not much mention of the support that might need to be provided by government, or the support that universities might need to do the things suggested, but in any case most universities are still some distance from meeting the standards suggested here.

Recommendations (for all students, not just those in provider halls) include dedicated accommodation to minimise ongoing transmission in halls of residence or shared housing; financial support (because young people are more likely than adults to work in occupations with high numbers of social contacts, and with less recourse to sick pay); access to mobile data; proactive outreach (inc. access to food) and both emotional and educational support.