Bill Anderson on coronavirus mistakes and the power of vitamin D
What does the loss of NASA’s Mars Polar Lander, as it attempted to touch down on the red planet at the turn of the millennium, have to do with the current rise in cases of Covid-19? Is the connection some kind of punishment for the hubris of our Daedalian interplanetary ambition, meted out by a distant God of War we disturb at our peril?
According to Dr Ronald B Brown it’s more calculus than Kubrick. In an article, recently published online by Cambridge University Press (open access), he describes how the Mars Polar Lander was a victim of a schoolboy data error – British Imperial measurements weren’t converted to metric, with catastrophic consequences for the mission when the Lander was calculating its descent to the surface of the planet.
Brown asserts that a similar global error has been made in our calculations of the risk of Covid-19 – overestimating its lethality, potentially by a factor of 10. The basic mistake we have made this time is to confuse the case fatality rate of Covid-19 with the infection fatality rate.
Brown points the finger at an editorial in the highly respected New England Journal of Medicine on 28 February 2020 which stated “…the overall clinical consequences of Covid-19 may ultimately be more akin to those of a severe seasonal influenza (which has a case fatality rate of approximately 0.1%)”. It’s the bit in parenthesis they got wrong: severe seasonal influenza has an infection fatality rate of 0.1%, not a case fatality rate. When American scientists compared the case fatality rate in the early stages of Covid-19 with this mistaken 0.1% equivalent for severe seasonal influenza, they came to the conclusion that Covid-19 would be about 10 times worse than normal winter flu and they shared this prediction with US Congress and the world on 11 March.
Why is what might appear to be a trivial distinction between “case” and “infection” so important? Because in epidemiological terms a “case” of any disease requires symptoms of the disease, whereas an “infection” may present no symptoms at all. A swab that tests positive for Covid-19 is only a “case” if it comes from an individual with symptoms – whatever the media call it, a positive test, by itself, only indicates an “infection”.
At the start of the outbreak in the UK, most of the people being tested were in hospital: they had symptoms, often severe, sometimes fatal. But now we’re testing many more people in the community and many more of them who test positive have very mild or no symptoms at all, and many fewer of them die. This is why case fatality rate is higher than infection fatality rate:
As we increase the numbers we test still further, as with other coronaviruses, even more people will probably test positive without symptoms, which will tend to push the infection fatality rate down even further. So even though increased testing tells us more people are becoming infected, it doesn’t necessarily follow that the NHS will be overwhelmed or deaths will spiral. And while Covid-19 can be a devastating disease to those who do become a symptomatic case, lockdown can be devastating to those who don’t, so the accuracy of how we perceive the predictive data is essential for policymakers to be able to determine the greatest benefit-to-cost ratio.
But how many asymptomatic, untested university students will unwittingly spread the virus and turn into “granny killers” if we don’t lock down? We can’t know. And we also can’t know how many of the population have already been asymptomatically infected and were spreading the virus since it came to our shores at the beginning of the year. In the UK we were told that the death toll could be 500,000 if we took no action. So we enacted a mandatory lockdown and currently Covid-19 has killed less than 50,000 of us. In Sweden, where lockdown was voluntary, their per capita death rate is comparable to that of countries who locked down assiduously, not 10 times higher, and Swedes are currently not experiencing the beginnings of a second wave. They have been testing for T-cell immunity as well as the antibodies we are testing for and they have found that asymptomatic Swedes who were in close contact with confirmed Covid-19 cases often don’t have antibodies but do test positive for T-cell immunity – so even though they were infected, without symptoms, an antibody test on its own would wrongly indicate they never had the virus or developed any kind of immunity.
Predicting the path of a novel pandemic is fraught with such complexities and hindsight is a great thing: at the outbreak of swine flu (H1N1) in 2009 Professor Neil Ferguson and his team at Imperial College London estimated the case fatality rate would be 0.4% which translated into a prediction of 65,000 deaths in the UK. In the end swine flu killed 457 people and the UK government spent £437 million stockpiling Tamiflu – a drug revealed (after hidden drug test data was unearthed by assiduous freedom of information requests) to only reduced the symptoms of flu by one day, with an efficacy equivalent to paracetamol.
Currently we yearn for a safe and effective vaccine for Covid-19 – as a species we’re throwing billions at a silver bullet that may never hit the target if the virus mutates faster than our fast-tracked vaccine development programmes. As positive test results increase we are warned to “remember March” and curtail our freedoms to save lives. But the implication that the increase in infections (that are still all described as “cases”) will result in the same devastating death toll we saw in the spring also rests on the assumption that doctors have learned nothing since then, and have no other medical interventions to make a significant difference to our fate until a vaccine arrives. In fact treatment and survival rates for Covid-19 have improved with our increasing understanding of this virus.
In another example of statistics being revered but not fully understood, Jon Moynihan, a foundation fellow of Balliol College and chairman of the technology-focused Ipex Capital, argues that Sage’s scientists could use a few more rough-and-tumble statisticians, and his case in point is vitamin D – a subject not unfamiliar to readers of the Idler blog, Why Lolling in the Sun Could Save Your Life.
Moynihan examines the notion of statistical significance – an essential tool to determine whether a scientific study is valid. Results of a drug trial that have more than a 1 in 20 chance of being random are classified as not significant – they might just be coincidental, accidental, and so they fail this gold standard test, and are dismissed. If there is only a 1 in 10 chance that a good, useful result might be down to random luck, this result would be dismissed even though there might be a 9 in 10 chance that the drug was genuinely, repeatably, universally helpful.
Being on the safe side of “significance” protects us from snake-oil profiteers. But what if not just one study, of vitamin D for example, narrowly fails to meet this gold standard, but like Hamlet’s sorrows, studies showing benefit “come not single spies but in battalions”? How does this combination of multiple silver and bronze standard studies effect the potential usefulness and safety of vitamin D? Moynihan’s position is clear: “The usual application of the much-abused precautionary principle is ‘don’t wait for overwhelming evidence before banning something that might be dangerous’, but the very same principle should tell us ‘don’t wait for overwhelming evidence before trying something affordable that might save lives.’” You can buy a year’s supply of high-strength (4000i.u.) vitamin D for less that 4p per day – think what price a national government’s buying power could secure for its citizens.
There is no money to be made in vitamin D for drug companies – now the biggest funders of notoriously expensive gold standard drug trials. And the correlation between this lack of potential profit and the lack of statistically significant studies on the benefits of vitamin D is becoming, in itself, increasingly significant.
Multiple studies showing the effectiveness of vitamin D in the prevention and treatment of Covid-19 are steadily climbing up the “significance” medals podium: a pilot randomised clinical study from Spain (just too few participants for the gold standard) took a group of patients admitted to hospital with Covid-19 and gave one half the standard medical treatment. The other half were given huge, restorative doses of vitamin D in addition to the same standard medical treatment.*
In the group receiving only standard treatment, 50% had to be admitted to intensive care, where 7.6% died.
In the vitamin D cohort just 2% had to be admitted to intensive care, there were no deaths and all were discharged without complications.
The doses of vitamin D were massive: 100,000i.u. on admission, 50,000i.u. on days 3 and 7 and then weekly. This might seem extreme when a supplement of 4000i.u. is now considered a daily high. But in midsummer if we stand outside in our swimwear at lunchtime on a sunny day we can easily make as much as 20,000i.u. in our skin – in less than 30 minutes.
Was Nature wrong to give us this prodigious capacity to make D? Might it be in anticipation of need in extremis? Today in the UK we are two weeks away from the start of the winter months, when the sun doesn’t rise high enough for any of the vitamin D-making UVB component of its rays to penetrate our atmosphere. Standing outside, stark-naked at solar noon might generate some frost bite but no D. As we brace ourselves for the more familiar winter flu arriving to combine with Covid-19, shouldn’t we at least be ensuring that vitamin D supplements are freely available to NHS staff today? As a bare minimum they are the ones we need to be fit and healthy enough to administer the Spanish-study-sized doses of D we might require if we end up in their care.
As comedians move into politics to become national leaders, has the time come for them to be recognised as great philosophers? Eric Morecambe and Ernie Wise must surely front the queue with their increasingly profound treatise Bring Me Sunshine.
Bill Anderson writes the beekeeping column in the Idler magazine and is the author of The Idle Beekeeper (Abrams)