Less than two weeks til the next big student migration

Various bits and bobs of note today. The Office for National Statistics (ONS) has a couple of articles on students and Covid – the first summarises its pilot research that we actually looked at in detail on the site a few weeks ago when the underlying data emerged.

TL;DR is that students followed guidance in much the same way as the general public BUT were more likely to be confined – students were more likely not to have left their home or accommodation in the seven days prior to being surveyed than the general public. Less sociable than the rest of the population, by all accounts.

29 per of students reported being dissatisfied or very dissatisfied with their academic experience in the autumn term; students reported lower levels of life satisfaction, life worthwhile and happiness, and higher levels of anxiety, compared with the general population; and when asked about how satisfied they were with their social experience, over half reported to be dissatisfied, or very dissatisfied.

The main reasons were limited opportunities for social or recreational activity (86 per cent), limited opportunities to meet other students (84 per cent) and limited access to sports and fitness facilities (52 per cent). Good job we didn’t oversell them that stuff in the summer.

Another ONS article examines trends in transmission with case studies from Exeter and Loughborough universities, and research from other higher education institutions in England. It finds the risk of transmission to be greater in residential settings such as halls and student houses, with minimal evidence found of the virus being spread in face-to-face teaching settings such as classrooms and lecture theatres. The largest outbreaks occurred in halls of residences.

We knew that would happen, of course. It’s just that we didn’t do anything about it. So on the assumption that international students due to start in January actually come (and those that haven’t made it yet also arrive) how’s the plan looking for even more densely packed halls in the new year? You know. In less than a fortnight.

The other day you’ll recall that universities minister Michelle Donelan was telling universities in England that mass testing was moving from a nice-to-have bit of extra pre-Christmas reassurance to something that will be “pivotal” in January, hence the “clear expectation” that “every student” will be tested.

And that was before we (and presumably DfE) knew about the “mutant strain” and the creation of Tier 4.

Now via some participation stats we’re starting to get glimpses of why it was that the department has started to ask universities to make students take part and return data on their compliance. Sadly via some efficacy stats it also looks like we’re also starting to get data on how useless, and potentially dangerous, the whole thing is.

Public Health Scotland (PHS) has published figures on the pre-Christmas testing period, which began on 30th November and finished on the 13th December.

We don’t yet know how many students are in the numbers, or how many students in the numbers got two tests – that will be made available in subsequent publications, once all results have been validated. But we do have numbers.

In the period, 43,925 LFD tests were carried out – of these, 79 (0.2 per cent) had a positive LFD test result, and preliminary analysis of 31/79 of the positive LFD tests has shown that 13 out of 31 subsequently had a confirmed positive PCR test.

If we assume most participants got two tests, as there’s around 300k higher education students in Scotland, we’re on just over 7 per cent participation – and that’s before we account for a handful of staff in the numbers. Yikes.

We might say to ourselves “ah well, at least prevalence looks low”. But that depends on the tests being accurate. We already knew that the LF tests in “real world” use in Liverpool were picking up about half of positive cases, making the chances of catching a positive over two tests a less-than-stellar 75 per cent.

But now Jon Deeks (who’s Professor of Biostatistics at the University of Birmingham’s Institute of Applied Health Research) had posted up a worrying thread of findings from the exercise at the university.

(Some caveats here. We’re not even at preprint stage on this one and so we’ll need more before we can base decisions on this. Alternatively, you can argue that we shouldn’t be rolling out a mass testing programme ahead of having findings we can base decisions on! And so on…)

In the Birmingham exercise, Deeks says that 7,189 students were tested 2-9th Dec. Two tested positive – a rate of 0.03%, in line with the low rates from other universities. Neither were false positives.

Trouble is, they also randomly sampled 10 per cent of the negatives and retested their samples with PCR. That found six false negative cases.

In other words, if that rate of false negatives was right across the testing, the exercise will have missed 60 positives. That would make the true prevalence rate 0.86 per cent.

Put all that together, and you have a test sensitivity rate of just over 3 per cent.

So at the risk of sounding particularly Grinchy – if we extrapolate and read across (and again that’s a big if) it’s possible that test participation amongst students of just over 7 per cent and sensitivity of those tests of over 3 per cent would mean we’ll have been picking up 0.21% of student cases.

And all of that’s before we factor in the risk profile of those more or less likely to have volunteered for a test in the first place.

Before you ask, work is “ongoing” to evaluate the ability of lateral flow devices to detect the new strain. But that’s when you actually deploy them. Given how fast transmission ripped through halls in the autumn, how often are we proposing to test students on arrival?

I’m all for not letting the good be the enemy of the perfect but there comes a point when the vast expense – and potentially false reassurance – starts to look like a problem.

Adding in uncertainty about the new “mutant strain” (was it around in high circulation when the student exercise was carried out? Do a disproportionate number of “away from home” students have family homes in what’s now England Tier 4? What if this finding of 40 per cent of student contacts are with individuals not affiliated with their university is true across the sector? Even on reduced social mixing does the migration event pose an intolerable risk that LF testing can’t spot?) and you end up with really serious questions about this exercise in the new year.

Which begins, as I say, in less than a fortnight.