Posted 27 April 2020
TAKE A CHANCE ON ME
The title of a song from the Swedish pop group Abba. In a locked-down COVID-strickern world, maybe it should now become the Swedish anthem.
Most countries of the world are now teetering on a brink. Amid declining new COVID case numbers and falling daily death totals, calls to ease lockdown are becoming more than a whisper. Economies that have been caged and deprived of life-giving sustenance are making last ditch gasps to be unleashed before it really is too late.
But of course there is push-back. The strong narrative that coronavirus is bad, really bad, is not going to be shaken-off without a fight. The media love the click-bait headlines such a narrative produces and they are not going to let that go willingly. And by many measures it has indeed been bad. Deaths no matter how caused make us grieve. If those deaths also come with a belief that they could somehow have been avoided, then other emotions get attached – fear, worry and anger. But how real is the evidence that these deaths could have been avoided? How much are the anger and fear byproducts of an illusory belief that Homo sapiens can control everything?
So the key question forming in my mind today, and one that hugely troubles me, is what is the evidence that our interventions have made a difference? Or perhaps more pragmatically, how do we measure what difference they have made as we must surely hope the measures have had some impact. But is the 80:20 rule at work here? Was it the small measures that made the big difference (hand washing, isolating the vulnerable for example). Have the big measures (switching off economies, telling everyone to stay at home) delivered just small, marginal benefits for big cost?
As a re-awakened statistician those questions excite me but at the same time they worry me too. It troubles me how the embedded narrative “this is really bad”, can be used to deliver a trump-card over any rational analysis. The meme “stay at home, save lives” for example, has become so embedded one can be seen as a heretic for just asking the question “how many lives did it actually save?”. To even posit that there is a follow-on question “Is that number less than than future lives lost because of economic downturn” can been seen as treasonous.
Whilst we may not be able to control everything, one thing we can be and should be in these uncertain times is rational and reflective. No question should be out of bounds but any possible answers should be based on evidence and hypotheses that can be tested, rather than unprovable dogma.
To lockdown or not to lockdown, that is the question. Well actually it is one of many questions regarding lockdown that should not be out of bounds to be asked. I will not subscribe to the “of course, we should have locked down” unless there is the analysis and the evidence to prove that. No one is going to shame me into not asking the questions. Inevitably whether we should or should not have gone into lockdown is now a question only history will truly be able to explore as it is the path we have taken. But I believe there is analysis of value now even before history delivers its verdict.
Let me offer some thoughts therefore on lockdown.
Lockdown, if it achieves anything, gives you time. Allow me to compare our modern day lockdown with fortified communities in medieval times that chose to barricade themselves in the face of threat from an invading army.
They would do this for one of two things to happen:
1. If they believed the wait changed the threat of the invading army. Either the army would disperse if they barricaded long enough, or the weather/season will change to make conditions tougher for the army to invade. Maybe wait long enough and the invading army could get attacked by someone else. The long-game here is that when lockdown was eased, the threat to the barricaded community had passed.
2. If the barricaded community could use the time to strengthen itself against the army – make more weapons train more soldiers, build better defences. The end of lockdown here would still mark the start of the attack but by delaying it, the defending community had made themselves stronger to fight it.
To put it more succinctly, reason 1 is waiting for the threat to change; reason 2 is waiting for the threatened community to change.
Where does our lockdown – coronavirus is the army sitting outside our ramparts – fit with this?
One thing I think is clear is that coronavirus is not an army withering on the vine just because we are hiding indoors. It is not going anywhere. It is no less likely to attack now than five weeks ago. The change of season is unlikely to make a difference – respiratory viruses are thought to be less dependent on temperature than influenza viruses – but even if summer does slow its progress, we will in all likelihood be back at square one in the Autumn. No knights in shining armour are going to come galloping-in and destroy the awaiting virus for us and then let us all out when they have finished. I therefore do not believe we are in lockdown for reason 1. The virus is still with us.
So what of reason 2? This is the one that is buying us time but it is only of value if those locked-down are doing something useful with that time. Otherwise it is time wasted.
Some of that time has not been wasted for sure. Many countries, including the UK, in very short order significantly increased their critical care hospital capacity. If extra deaths were going to be caused because demand on healthcare systems from viral infection exceeded capacity, increasing capacity was a key strategy to reducing those deaths.
“Flattening the curve”. This is the other half of the strategy mentioned in the previous point. If the limit in critical healthcare capacity was going to cause extra deaths, another way forward would be to reduce the peak demand on the system. Reduce the spread of the infection, fewer cases, fewer hospitalisations, less demand on critical care beds. Did lockdown do this? This is the big question and one for which I have yet to see evidence. History may provide that evidence and so it is important not to confuse “evidence of no benefit with no evidence of benefit yet” (just because there is currently no evidence available does not mean it will not become available later). But we do have Sweden.
Sweden chose not to impose complete lockdown. Gyms, bars, restaurants and schools remained open. From a statistics perspective this is great news. We have a country that did not adopt lockdown, we have lots of countries that did. The analysis has to be: Is there anything about the data coming from Sweden that suggests difference or similarity with the data coming from other countries? And can we mitigate any observed differences being just because Sweden is different in other regards, not just lockdown?
The bad news for pro-lockdown positions is that Sweden has not proven to be the apocalypse some predicted. They have had cases and they have had deaths but the shape of their curve is proving little different to anyone else’s. They are on a steep decline now just as other countries are. And because they did not have complete lockdown, they do not have the same fear of surging cases that might follow an easing of lockdown.
We cannot compare total cases in Sweden with elsewhere (Sweden is different in too many other ways) but if our lockdown had flattened our curve then Sweden’s curve would be ‘less flat’. This is not the case.
[An aside here. Early models of COVID suggested deaths in the UK ion around 600,000 unless lockdown be implemented. There is strong evidence that one such reported model swayed Boris from following a Sweden style approach as he had until late March and moved us to lockdown. Already I am seeing this used as evidence that lockdown worked – “we did not get the hundreds of thousands of predicted deaths, so lockdown worked”. No, all that proves is the model was wrong. Sweden have just over 2,000 deaths so far without lockdown and not the 100,000 deaths which would be the proportionate number if the model had been right].
Some now argue that we went into lockdown too late. There are suggestions that Boris was contemplating his navel too much and that has cost lives. This is predicated on the assumption that lockdown flattened the curve and that therefore locking down sooner would have flattened it to a lower level. I struggle with this. Notwithstanding that our curve is no flatter than Sweden’s, a hypothesised earlier, even flatter curve would only deliver benefit if our actual curve had still led to significant deaths because health care demand exceeded capacity. Where is the data on this? If people were dying in their thousands because they required but could not get critical care, wouldn’t that have been headline news? Would that not be a statistic at the top of everyone’s consciousness? Absent data of course does not mean that data is not there, merely that it has not yet been collated and reported but something like this would surely be hard to keep down if it was of significance?
If all an earlier lockdown is going to do is delay cases and not change whether peak demand exceeds capacity, then maybe there is no argument for it. Interestingly a few countries that I see from headlines are “warning the rest of the world do not exit lockdown too early” might actually be those who locked down too soon, and are now experiencing the inevitable cases that lockdown just delayed but could not prevent. Perhaps those who locked down later have experienced more of those cases already. Only later data will tell us this.
And statistically, unless a vaccine comes along or the virus goes away, the area under the curve is the same no matter how much or how little you flatten it. And the area is the total number of cases. Early lockdowns, more enforced lockdowns, even if you subscribe to these actions flattening the curve, will still give the same number of cases in the end. There is no running away – already exposed, be exposed now or be exposed later. They are the only options on the table.
Has lockdown bought us time in any other ways? Well, it has allowed some learning of how better to treat critical cases. There is increasing evidence that mechanical ventilation may not benefit a subset of clinical presentations of the virus. Understanding that may save a few lives and may suggest that ventilator availability are not on the critical pathway after all. That is specifically a benefit of lockdown only if it has delayed cases or flattened curves sufficiently before that knowledge makes a difference.
But the time lockdown has given us has down sides too. Your immune system will be a victim of lockdown. Immunity works and develops by being constantly exposed to pathogens around us. Isolate in your homes and that process slows down; your immunity reduces. When you step out of lockdown, you are more vulnerable to all infections not just COVID. Anxiety, stress, the emotional consequences of lockdown will all impact health too and in particular also the health of your immune system. There will be a surge in illness as we step out and that will get worse the longer we are in lockdown.
Economic downturn costs lives. Unemployment, reduced wealth and uncertainty in people’s lives cause illness and leads to deaths. Every extra day of lockdown and there will be more of these deaths. Worryingly these deaths will be in the young, fit and able as much as anyone else. The average age of a coronavirus victim is currently 80. The curve of lives you might save by being in lockdown is going to be (and may already have been) dwarfed by the ever increasing curve of deaths that will be caused by lockdown itself and these will be in all strata of the population. Staying in lockdown may now be no safer than being exposed to the virus, it is just a different threat.
There is however one thing lockdown could change for the better: your metabolic health. The evidence is mounting that co-morbidities, even in the old, make more difference to your risk of infection and severity of infection. Data shows that 95-99% of patients needing critical COVID care have one or more of these co-morbidities.
And many of those co-morbidities are in your control – dietary and lifestyle choices can make the difference. If you are over weight, obese, have hypertension, cardiovascular risks and other signs of metabolic ill health and you are in the area of the curve yet to happen, then the thing that will make most difference is in your hands. Questions over lockdown and vaccines become a second order of importance. It is how you have used your own lockdown to tackle your markers of ill health that will make the difference.
The mood is changing here. Whereas the first weeks of lockdown were embraced by many in the spirit of “I’ll do my bit”, there is increasing evidence that enough is enough. Frustration leads to anger and anger leads to finger-pointing and blame. I have recently seen it blamed on Boris, I have seen it blamed on Brexit (I haven’t yet seen it blamed on the Boogie but give it time).
Maybe Sweden will be proved right. The Abba song title would then take on more poignancy.