The Week in Public Services: 13th October 2020

This week: responding to rising coronavirus cases; the hospital discharge debate resurfaces; and the next exams crisis?

General

It’s been a while…life, work, everything else, got in the way. Cutting through the noise of the last few weeks, here’s what — on reflection — I found most useful:

First off — this blog from Tom Chivers, which looks at whether Covid-19 is less likely to kill you than it was in March. Thanks to advances in medical knowledge, doctors know how to treat the virus better than they did in March. The infection fatality rate in the UK — how many die as a percentage of how many people have the virus — has probably come down a bit since March (partly because younger people are getting the virus, partly because of improved treatment, and possibly because people are now being infected with a smaller ‘dose’ of the virus due to more widespread mask wearing). To be clear — this does not mean that we can all just relax and let the virus spread unmitigated; but it’s good to reflect on the positives and the medical advancements. We all need hope.

This Atlantic longread on the ‘k rate’ — a measure of dispersion, rather than transmission, of a disease — is super interesting. (It’s also long and quite involved, so grab a cup of coffee for this one). The point is basically that Covid-19 is “an overdispersed pathogen [meaning that it tends to spread in clusters]”, and that this explains some of the seeming puzzles that transmission rates have been so different in otherwise similar countries with similar restrictions. One thing I hadn’t realised is that this is one reason why flu plans were so inappropriate — flu typically doesn’t spread in the same clustered way that Covid-19 does. It may potentially mean that contact tracing in reverse — looking for the origin of an infection rather than who a person may have passed the virus on to — could be very valuable.

Moving back to the UK, this longread from Chris Cook explores (£) why Britain responded badly to Covid-19. His argument is that as flu pandemic plans were so detailed and thoroughly planned, it proved hard for ministers and officials to follow a different course. That much is fairly well-established already, but it’s worth reading in full for the insights from civil servants and scientific advisors on why the government made its decision to initially abandon a containment strategy in March, and why local authorities were not more involved with contact-tracing. This longread from Matt D’Ancona on testing specifically, is also a good read, which pins the decision to go for the centralised Deloitte-run ‘Lighthouse labs’ on “an almost religious belief [within the current Cabinet] that big private sector solutions are better than fragmented public services”.

The big debate is of course what to do now that coronavirus hospitalisations are rising in England. My two cents, on why NHS England and Improvement should remove some of the national targets they reimposed during the summer — is here.

But looking at the big questions and stripping out the central/local and internal Tory party politics for a second, there is one really important point to make: there is not really a big scientific division about the public health response. There is more consensus about public health measures (even if the right economic response is still a matter of significant controversy) than some of the media coverage I’ve seen. For sensible commentary, I find the following people really helpful — Tim Harford, Adam Kucharski, and Christina Pagel.

For all the ideologically-tinged New Zealand vs Sweden debates, either approach would be viable (£) — but both require a working public health system including, vitally, a functioning test-and-trace system. On which note — my colleague Cath wrote an excellent blog arguing that the new ‘tier restrictions’ announced on Monday will just be a rebranding exercise unless there is local backing and an effective test and trace regime. Andy Cowper’s BMJ blog is also clear: restrictions aside, the government still needs an adequately functioning Test And Trace system and proper support for self-isolation (there’s a theme here, pay attention at the back).

And in a blog that looks particularly pertinent now, my colleague Sarah has looked at how the government can make better decisions if there is a second wave: set time aside to think about the medium- and long-term, bring businesses, unions, and local government into decision-making, and appreciate scientific advice — without treating them as policymakers.

Looking to the future, Reform have published an interesting collection of essays on how to build a ‘resilient state’ with lots of the great-and-the-good contributing ideas. Worth reading with Tom’s back-and-forth on the essays with Owen.

Health and Social Care

Right, in testing news, the director of the Francis Crick Institute has criticised the government’s centralised approach to testing, arguing that the Department of Health and Social Care has got in the way of increasing testing. Meanwhile, care homes are still finding that most test results aren’t turned around within 48 hours. The longer they have to wait, the harder it is for them to keep residents safe, and the bigger the risk of infection. Gaaaaahhhhhhhh.

The headline that really made me choke on my cheerios, though, was last week’s news that a technical glitch meant that nearly 16,000 cases of coronavirus went unreported — meaning that the last week’s daily figures were underestimates and, more importantly, contact tracers will have been delayed in their efforts to trace contacts of people who tested positive. This is really pretty bad, as the BBC explained. What has that meant in practice? Jen Williams chronicles the problems this late-reported data caused in Manchester, where local officials think they were delayed in tracing the contacts of around 2000 people with coronavirus. “Intergalactic level incompetence”, according to one local official.

Putting that in context, this Tortoise longread (£) is a good factcheck — with great visualisations — of how testing in the UK compares to other countries. While the UK does perform more tests per 100,000 people than the EU average, the way we count tests includes antibody tests (measuring whether someone had the infection the past) and tests posted home (only some of which are returned). It’s not clear how other countries measure their tests so…we might be testing more people than other countries. And the more important question is whether people with symptoms can get a test and a result quickly. Here, Test-and-Trace’s performance has been getting worse (although, to be fair, I have seen no comparisons to other countries), as Adam Briggs has been dutifully documenting.

Billy Palmer also continues to produce great graphics about NHS Test-and-Trace. There are a lot of unused antibody tests which are (sometimes) counted in headline testing capacity numbers, and the number of swab tests processed is (still) not the same as the number of people tested. Great effort to piece together disparate bits of information to come up with a best guess of how much testing is actually happening, where, and for what.

Perhaps the most brazen news though is that Deloitte — already in receipt of significant government funding to deliver testing for Covid-19 — is attempting to sell councils a “local test and trace solution” in a move which has been described as “brassneck” profiteering by senior public health figures.

In better news — the Department for Health and Social Care now think they have a more resilient PPE supply chain. I was heartened to read that “we want to consider stockpile requirements for other types of pandemic threat”, alongside an immediate Covid-19 stockpile. The Department says this review will take advice from scientific advisors on NervTAG. This is sensible — as the IfG said in our report about resilience earlier this year, the government’s focus on planning for pandemic influenza after 2009 was a weakness of pre-coronavirus emergency planning. Excessive focus on any one threat, or being prepared for the last crisis, is unlikely to pay off. Slightly worried, however, to hear that the immediate four-month stockpile in case of a future surge in demand will only be in place for November — there is presumably a risk of further shortages if there is another surge before then…

Also new in the last few weeks: a National Audit Office investigation about how the Department of Health and Social Care and the Cabinet Office bought ventilators in the heat of the first phase of the crisis. The NAO are remarkably sanguine about the whole affair, concluding that “both departments maintained sufficient record of their programmes’ rationale, the key spending decisions they took and the information they had to base those on […] effective programme management, controlled costs where they could and recovered some of their committed spending once it became apparent that fewer ventilators were needed than they had originally believed”.

In other new research, the Institute for Fiscal Studies have analysed how use of NHS care varies by educational status. The headline finding is that “after taking into account individual health needs, the highly educated aged 65 and above use considerably more NHS outpatient (specialist) care than their less educated peers”. This only emerged after 2010, which raises an important question which the authors are unable to answer — why?

(Nerdy sidenote: it’s based an awesome method of linking longitudinal survey data and NHS hospital records — admin data is cool. What’s even more fun is talking about research methodologies, and Harry Evans has written a wonderfully insightful thread about why evaluating integrated care initiatives is so hard when people want to show results quickly, finding control groups is difficult, and it’s often hard just to define the intervention. Still, given the alternative is plausible guesses, it’s gotta be worth trying).

Back to other new research:

• The Nuffield Trust have analysed why there is a shortfall of new recruits for mental health nurses
• Fantastic Health Foundation research into how general practices in deprived and wealthy parts of England compare (they find, amongst other things, that “practices serving more deprived populations receive around 7% less funding per need adjusted registered patient”, there are fewer GPs and more practice nurses in deprived areas, but regardless of the level of deprivation in their area, patients have a similar probability of having an appointment at their GP practice)
• Useful summary of the latest NHS performance data from QualityWatch
• A good blog from John on whether the NHS can expect a ‘V-shaped recovery’ when it comes to providing non-covid services (TL;DR — “with the continuing need for infection control measures in hospitals, the consequent effect on productivity, and recent signs of rising covid cases and hospital admissions, whether it will be possible to return to average levels of pre-covid activity over the next few months must remain in doubt”; or even shorter — probably not, especially if there is a second wave)
• Mark Dayan reminds us why Brexit still poses big challenges for the NHS when it comes to regulating medicines and devices, finding staff, and getting cash out of the Treasury

Former Matt Hancock SpAd Richard Sloggett wrote an interesting, David Rumsfeld-inspired, blog about the five ‘known unknowns’ of managing the next phase of the pandemic — how best to deal with Covid patients, non-Covid patients with long-term conditions, patients waiting for planned care, ‘missed conditions’ owing to lower-than-normal screening during the pandemic, and patients indirectly affected by the virus. And last but not least, the October NHS England board papers are out, and include an independent report with a fairly expensive set of recommendations into how to improve diagnostics.

In social care, the new Infection Control Fund guidance is out. David Brindle notes that more of the money will be going to domiciliary care this time round. Taking the long view, Simon Bottery has written a great article on why a national care service — whatever that means — isn’t the slam-dunk solution it might appear. Why, you ask? Put simply, social care reform can be thought of in six distinct buckets: who is eligible for it, who manages it, who commissions it, who provides it, who regulates it, and how it is funded. For some of those questions, a national solution (in my opinion — eligibility, regulation and funding) is a good idea. “Local councils know their communities and markets and are best placed to support a person-centred approach. Yet they don’t always deliver, so again there is a need for national bodies to work with councils — a national programme to spread innovation and good practice would help, along with work by integrated care systems to join up services with health.” The picture, as always, is complicated…

And in the spirit of admitting mistakes about complicated things: I was thoroughly, completely, wrong. For a while, I’ve been saying that we’d never be able to know exactly how many cases of coronavirus in care homes came from faster hospital discharge because there was not a robust testing system at the time — and without having tested all discharged patients, we wouldn’t have known how many patients were transferred to care homes with asymptomatic coronavirus. Fortunately, lots of people are much smarter than me, and a new SAGE note summarises five studies into care home coronavirus outbreaks.

These studies, by looking at the timing and spread of infections and retrospective genomic analysis have, overall, found that “the weight of evidence [for the routes in which viruses entered care homes] is stronger in some areas than others […] evidence of staff to staff transmission has emerged in the genomic analysis (high confidence) [and] weak evidence on hospital discharge and modelling the impact of visitors does not suggest a dominant causal link to outbreaks from these sources.”

NHS Providers CEO Chris Hopson thinks this vindicates the NHS Providers view that hospital discharges were not the main cause of excess mortality in care homes. Hmmmm — one big caveat: absence of evidence is not the same of evidence of absence. Just because it’s hard to observe cases originating from hospitals in the data, it doesn’t mean that fast hospital discharge wasn’t responsible for at least some cases of coronavirus in care homes. As the researchers themselves note, there is “evidence for multiple routes of virus ingress to care homes, [and] are not systematic enough to quantify the relative frequency of different routes of ingress […] these studies do not definitively rule out any mode of ingress”. Still, it’s more evidence of the harm that lack of political attention to social care during the start of the pandemic caused.

Children and Young People

Important and worrying evidence from a Department for Education survey of local authorities suggests that the rise in referrals [to children’s social care] that many predicted as lockdown was eased is happening . “The total number of referrals during Wave 8 [the end of August] was 12% higher than the usual number of referrals in the same period over the past three years. An increase was seen by around 60% of local authorities”. Concerningly, 72% of local authority respondents reported that foster care and residential care placements were more expensive due to Covid-19 pressures (compared to just 15% reporting no increase). Ofsted have explained what they’ve been up to during the pandemic here.

In school news, and I don’t mean to alarm you here, we may be sleepwalking into another exams crisis next year. The DfE’s latest guidance on exams for 2021 says that GCSE, AS, and A level exams will go ahead next year, including some changes to assessment. Sounds sensible — better than trying to guess at the results of non-existent exams, right? Well, maybe — but there are a few problems with just running exams too. As the two Chris’* have pointed out: what happens if certain sorts of schools do systematically better/worse than normal in 2021? Is that fair to have the same exams everywhere, given that privately educated kids and kids in wealthier families have had spent more time studying, with higher quality lessons, broadband connections etc.? (God knows how much evidence there is on this point — see past Week in Public Services, pretty much any week since March this year).

In other new research:

• A brilliant report from Luke Sibieta and colleagues at the Education Policy Institute comparing how education policy in each of the four nations during the pandemic to see what they could learn from each other (spoiler alert: England should be looking to Wales when it comes to free school meals and online learning)
• The Centre for Public Impact have published a report on what positive lessons for children’s social care social workers and other practitioners think we can take from coronavirus
• And some nerdy fun: the ONS explain how they tried to measure how much education there was when there was, erm, almost no information about measurable education output

Law and Order

In policing, the government has been crowing about the news that, as of 15th September, 100,000 people have applied to become a police officer. However, as Gavin Hales notes, there remains a worrying uptick in voluntary resignations over the past few years — probable means the government ought to focus on retaining existing police officers as well as recruiting new ones…

In prisons, there’s been an interesting mix of new articles and research published on what prisons were like during the lockdown. First off — this from the BBC suggests that the lockdown made prisons safer. This Prison Reform Trust briefing on how prisoners coped under the restrictive prison regime put in place to limit coronavirus transmission is much less sanguine, and argues that increased safety has come at a huge price: “the day to day misery of confinement to their cells for 23 hours or more, over a period of five months so far”. And this account from an anonymous prison officer in The Guardian suggests a bleaker picture still. I hadn’t realised that prisoners have only had 10 minutes to shower and 30 minutes to exercise every day since March.

Finally, a good Commons briefing has also set out what we know about ethnicity and the criminal justice system, and disproportionality (in this case, comparing the share of the general population that identifies as belonging to BAME groups with the share at specific points in the justice system) in everything from the judicial benches to sentencing.

Local Government

In a point that I think has been missed in the national coverage of coronavirus, an investigation by The Observer looked at what Preston council are doing to try to stop the growth in cases, and found local officials setting up their own testing sites and tracing systems out of frustration with the poorly-working national system. 82 councils have so far set up their own systems. There are about 150 upper-tier local councils in England, so that’s just over half — let that sink in. (If you, like me, were wondering what exactly these local tracing systems were, Adam Briggs has written an excellent explainer).

Ongoing public health debates aside — yes, I know that is a big thing to put aside — there has been less local government news over the last two weeks. The Local Government Association released a report about fragmented central government funding for local government, which found that councils receive — or, in many cases, apply for — a larger number of grants than in the past, even though they are receiving less central government funding. They make a strong case that this is not a sensible way to distribute money.

Grant Thornton and CIPFA have also released a new report about local authority finances, setting out the known unknowns that will affect local authority finances. Not a cheery read, but a very clear analysis — well worth reading.

* Has anyone else has noticed that Chris Giles and Chris Cook have been producing some of the clearest analysis of exams? Is it something about being called Chris?