Care Home Size & Covid Deaths - the unhelpful concept of ‘association’
A suggestion has been made on the back of a Scottish Government report that larger care homes have done worse in Covid than smaller.
The government report that is supposed to suggest this is based on data that even their report admits are flawed, for reasons anyone can understand.
Decisions were made with honourable intent under pressure and with limited knowledge in the emergency.
It is unnecessary to blame others for the outcome of these Health Department decisions even if hindsight suggests they could have done better.
There is no need to drum up additional reasons for doing what they did when they honestly acted on the information available at the time, but the report does this.
Mixing up the ideas of association and causality is a notorious tactic.
Highlighting a poorly evidenced idea that large care homes are more dangerous than other care homes is a bit of a distraction from the question of why so many people needed a sudden transfer during this dangerous time.
This blog is about a Scottish Government report that has just been published about deaths from Covid 19 in care homes in Scotland from the start of the pandemic.
Talking about size of care homes at this point is unhelpful
A false impression given when it says there is an association between numbers of deaths and the size of the care home. ‘Association’ is a word that should be banned because it is often picked up by the listener as meaning ‘cause’. To illustrate, think of the association between Santa Claus and Christmas. Santa is uniquely ‘associated’ with Christmas, but he doesn’t cause it.
The report says that there is a statistically significant association between larger homes and more deaths. ‘Statistical significance’ always sounds like quality control for evidence. The First Minister, who has been magnificent in her communications about Covid19 disappoints in the press release by being said to speculate about a reason for this association. That suggests that she thinks there is a causal relationship. We trust her on many things but this won’t do.
Patchy data in the government report
This government report itself says that their own data is patchy and outlines how it was cobbled together from a lot of sources, which is never a good look.
The report says that hospital data is challenging (page 6) and some of the data is not yet available. Why has the report been issued and conclusions drawn before the data is even properly available? They are now going to work to improve their data and that is good, but it’s a bit rich to draw conclusions that could undermine families’ confidence in care homes, and the damage the morale of people on the front line there, on estimated numbers.
Pressure on managers to discharge and care homes to accept residents at short notice
“There’s no question that thousands of older patients were discharged from hospital to care homes from the start of March in order to clear the hospitals for the feared onslaught.”
The report estimates the number of hospital discharges. There’s no question that thousands of older patients were discharged from hospital to care homes from the start of March in order to clear the hospitals for the feared onslaught. Hospital managers will tell you privately they got the call telling them to do this, and care homes will tell what pressure they were under to accept residents at short notice.
Hindsight is a great thing but it is not always evidence of wrong doing
Hindsight is a great thing, but it would be good to get an admission that this was to an extent avoidable. The incubation period for older people is greater than young people and so people testing negative before leaving hospital is something that you’d have to wait a long time for. And we thought we needed the beds. The pity was they were never cleared earlier. But someone had to make a tough decision and they did. No shame in that. But the report hints at back peddling now.
For example, the report emphasises that everyone who left hospital did so for a good reason. Even I would not question that there were valid clinical reasons for moving many of the people out of hospital (page 5). Many have agued that the majority were overdue for moving out. It might be said that there was no valid reason for many of them to even be in hospital because they were medically fit for transfer already. I think this statement about clinical reasons is made to try to forestall any question about whether the reason for the mass exodus was clearing the hospitals. Why try to deny it at all? It could be justified as a decision made on available evidence about what was needed at the time. It would be astonishing if such a massive, unprecedented, number of old people suddenly got clinically well enough to leave hospital in such a short period of time. Not credible.
Reasons for not testing people are not transparent
The report then gives reasons for not testing people who were discharged to care homes. They could have said the NHS made a tough decision in a hurry in a public health emergency. But they don’t. They say there are patient-related reasons for not testing.
Of course, there are sometimes valid reasons for not testing. But consent, which the report leads with, is not a very good one in these circumstances. If the person could consent to being transferred, they could consent to testing. Is the report suggesting they could be transferred without consent but not tested because they couldn’t agree? For people who lack capacity to agree to anything, there are legal methods of getting permission. It is appropriate to test someone even near the end of life if they are going to put others at risk of infection. It could help you decide if they can have a hug, for example. I am not sure why this rationale is placed so near to the front of the document, as it makes the report sound defensive. If it was meant as a defence, it is not a very good one.
Figures based on case notes and NOT on laboratory tests
The figures for most of the early discharged patients’ Covid cases in the report are based on case notes, (page 14) and not on laboratory tests. That means that someone wrote something in the patient’s notes that would allow a researcher to assume they had Covid. This makes the data interesting but not something you should base any important policy decisions on. Anyone could excuse the lack of availability of Covid testing kits, or the chaos that meant they were not available to care homes for months. Anyone could excuse doctors who were advised not to go to care homes or care homes that were advised not to take residents to hospitals or GP surgeries for writing in the notes that the person had Covid without a test, but it is a bit rich to count them as Covid infections ‘caught’ in the care home on this basis.
Of the delayed discharge people who went out in the seven or eight weeks from March 1, about 1,000 were not tested. It was impossible to know at that time how many of those carried Covid into their care home, because old people are often “atypical” in their presentation. This means that whatever was wrong with them, it did not look like this new disease we were just learning about, so it was not spotted. The peak of care home infections rose till week five and six and then started to fall again. So it looks like a strong association between the time of no tests, and hospital discharges, and lack of knowledge about symptoms in old people. That is not mentioned in the report. But the association with the size of the care home is picked up and headlined. You can draw your own conclusion about attitudes to care homes and maybe even defensiveness from the NHS. This is so unnecessary unless someone is looking for a scape goat.
Many reasons people died in care homes during the pandemic, not only due to Covid-19
People in care homes have died of other things associated with this difficult time not caused by Covid itself. There are reports of loneliness leading to decline, and reports of people going on Do Not Attempt Resuscitate status. We hear about hospitals deciding to limit admissions or certain treatments for people over a certain age, and delay in elective treatments. Given the absence of testing and the problem of unusual symptoms, it is now impossible to say accurately how many older people in care homes really died of Covid. The government report has added up anyone where there was a ‘mention' of Covid on the death certificate, but it doesn’t show what variation there might be between different doctors’ tendency to include Covid when certifying death. This would depend on the reporting of symptoms by care home staff that gave rise to the doctor’s conclusion so some homes may have been inclined to suggest it to the doctor and some not. The doctors spent much less time with the old people in care homes than usual and they had little testing capacity at that time.
This would not matter. Except that the report figures indicate that 25 homes accounted for a quarter of all deaths. They say ‘just’ 25 which I think is potentially emotive. The question is implied, “What on earth were these 25 doing to cause such a bad out turn?” And we are told that they were mainly the big care homes.
Association vs. Causation
It must be remembered that to count as a Covid death, there was only "any mention of care home on the death certificate". Any mention? On the basis of no testing? Some more questions must be asked before we can accept the implications of this statement in the report. These questions include
Was a smaller home less likely to realise someone had Covid? Did they miss cases that bigger homes managed to identify?
Given that Covid looks much different in older people than the symptoms we were all being told about, such as coughs and temperatures, did smaller homes with fewer staff have any staff with knowledge to identify the unusual Covid symptoms in older people (ie falling down, losing appetite)?
Care homes that are not for elderly people are typically smaller than care homes for older people - was this taken into account when implying that more older people died because of the size of their care homes?
We really need to talk about the term ‘association’ as used by the report (p 22.) It is a weasel word - a word that creates the impression that something significant has been said when in fact it is intentionally ambiguous or misleading. We laugh at the idea Santa is uniquely ‘associated’ with Christmas without causing it. But this ambiguity is not amusing for people who live and work in large care homes and their anxious relatives, especially cruel if it is misleading.
More evidence is needed for a reliable report
You would need more evidence than this report before you can in any appropriate way raise doubts about the safety of larger care homes or suggest they somehow ‘caused’ Covid deaths. And if you deliberately promote that conclusion, the main doubt raised is whether you are being motivated by something other than available facts, which doesn’t help anyone.
If you would like more information, you can buy my book Dementia, the One Stop Guide or Care Homes: When, Why and How to Choose a Care Home. I am available for consultancy for families or organisations. And if you have any further queries or questions, or suggestions for something you’d like to see me write on, please contact me via the Contact Page
See my new course on Dementia the One Stop Guide on Policy Hub here
