Is Covid-19 over… no it’s not!

In this blog I try to keep to the science I know, and If I don’t have enough expertise, I reach out to those who do. I declare therefore, I am wandering into fresh territory in this post and, at least in part, it’s my opinion – although I believe backed up with the facts. I know others might disagree with the opinions – and that’s fine, but I hope the facts stand for themselves.

I’ve seen comments on social media declaring the Covid-19 lockdown is over and we can get back to “normal”. I’m witnessing pubs opening and people gathering in my local area and a distinct relaxation of social distancing. I find this disturbing because we are not out of the Covid-19 crisis yet and the scientific experts are warning of the likelihood of a second wave.

The claim from government, up to now, has been they are following the scientific advice. If this was true, then coming out of lockdown now is certainly not following scientific advice. The independent SAGE report on re-opening schools explicitly says,

“…. The most recent estimates for the UK are that R is between 0.7-1, meaning that all scenarios modelled by SAGE are at risk of pushing R above 1. The school reopening scenario chosen by the government is not one of those modelled by SAGE making the potential impact of reopening even more uncertain. Robust testing systems are not in place everywhere. Additionally, public adherence to social distancing is influenced by trust in the government and its messaging. This trust is increasingly strained. We therefore believe that by going ahead with a general school reopening from 1st June, the government is not following the advice of its SAGE group and is risking a new surge in cases of COVID19 in some communities.”

There are those who point out the economic damage of a prolonged lockdown and I understand that. The problem of course is how do governments and their advisors balance the risk of Covid-19 fatalities with economic damage. That’s a hard question and not one that I would want to tackle. The UK however, has an appalling record of Covid-19 fatalities. The current statistics show the UK is one of the worst hit countries, certainly in Europe and, depending how you cut the statistics, also in the world.

If this situation is because the country was late to enter lockdown, then we will not compensate by coming out too early. And what I fear most is that given the polarised politics of today, the government will try to shift blame towards scientists. If this happens, then just remember the words of the SAGE report,

“… the government is not following the advice of its SAGE group and is risking a new surge in cases of COVID19 in some communities.”

Covid-19 vaccine pessimism

Another blog on Covid-19 for the non-expert

76c741_25e433c9b3314bde82eee474c60c51a0~mv2.jpg

I’ve seen claims on social media and in the press saying a vaccine for Covid-19 may never arrive. The Telegraph, and The Guardian are two examples.

The three most commonly quoted reasons I’ve seen for vaccine pessimism are: (1) vaccines take many years to make, the mumps vaccine took four years. (2) We still don’t have an effective vaccine against HIV and (3) we still don’t have any vaccines against SARS and MERS.

All three points are correct but only superficially, and so let’s look under the surface and put the claims into more detailed context.

Scientists isolated the mumps virus in 1945 and the first vaccine appeared in 1948, although its effectiveness was short term. The Jeryl Lynn vaccine was launched in 1967 in the United States and entered routine use in 1977. The mumps vaccine therefore took between three years or 22-years to develop, depending upon how you look at it. The problem with comparing Covid-19 to mumps however, is the year the first mumps vaccine appeared, 1967, was 53-years ago. At that time there was only one way to make a vaccine and that was from the virus itself. It was necessary to isolate the virus, then inactivate it so it remained potent to the immune system but non-pathogenic (known as an attenuated vaccine). Things have come a very long way since 1967. Nowadays we no longer need the virus itself, but only its genetic sequence (RNA in the case of Covid-19).

Covid-19 was first reported in December 2019, and by February 2020 the 26,000 – 32,000 RNA code sequence of SARS-COV-2 (the Covid-19 virus) went round the world via the internet. Genetic sequencing and the internet were both science fiction in 1967; in fact the structure of DNA was only elucidated 14-years before. Vaccine development today is another world, and comparing a Covid-19 vaccine to one for mumps is like saying electric cars are impossible because there weren’t any in the 1960s.

HIV was first identified as the causative virus for AIDS in 1984 and despite some trials, no effective vaccine has emerged. HIV however, is an entirely different virus and if SARS-CoV-2 was anything like HIV, few scientists would be optimistic of ever getting a vaccine. HIV attacks the immune system, the very thing a vaccine stimulates in the fight against infection. HIV also mutates rapidly, and it’s ability to mutate appears to be built into its very biochemistry because a key enzyme in its reproduction (reverse transcriptase) cannot translate its genetic code accurately. This has led to over 60 strains of HIV virus world-wide. SARS-CoV-2 is far more stable and all evidence suggests it does not attack the immune system. Comparing HIV and SARS-CoV-2 regarding a vaccine is like saying I can’t rid my lawn of dandelions because I can’t eliminate Japanese knotweed.

There was an outbreak of SARs in 2003 and MERS in 2012. The reason why no vaccine has emerged for these viruses is less technical than for HIV – no one has really tried. There was an initial effort and there are vaccine candidates, but it soon became clear that neither of these diseases turned out to be the deadly pandemics they were first thought to be. They were successfully contained and haven’t been seen for several years.

By far the biggest difference between attempts at a Covid-19 vaccine and any other are the enormous international collaborative efforts currently taking place. There are an estimated 100 Covid-19 vaccines somewhere in development as well as other possible treatments such as a range of drugs, both established and new, including antibody-type treatments which proved successful against Ebola. (Remdesivir was aimed at Ebola but was succeeded by antibody-type drugs. Remdesivir is now being tested against Covid-19). At the leading edge of fresh approaches is research into RNAi. (Richard Jorgensen discovered RNAi, after he became curious about patches of colour on petunia petals. His investigations led to the discovery of a whole new branch of genetics and won him the Nobel Prize in 2006).

We have to be realistic about the possibility and timing of a vaccine, or other treatment, and it may turn out no such vaccine emerges, or if it does, it might be years away. The honest answer is we don’t know but I do believe there are reasons to be optimistic. The Jenner for example are entering Phase II and III trials after promising phase I results. But in the meantime, if a more pessimistic climate arises, then at least let’s make sure it’s for the right reasons and not some spurious comparisons.

Covid-19 – warnings ignored

I’ve heard it said, we have not experienced a pandemic since the Spanish flu of 1918 so how could anyone predict Covid-19? Just about everything in that statement is wrong.

Everyone has heard of Spanish flu, but how many have heard of the 1957 flu pandemic, or the one in 1968? We see an annual upsurge of influenza every year which on average kills about half a million people worldwide. The United States sees about 20,000 deaths a year and Europe(1) about 70,000 due to influenza. Spanish flu was far more deadly, killing an estimated 17 million to 50 million people – and possibly many more. It was by far the worst flu pandemic in over a hundred years, but it wasn’t unique. The 1957 flu pandemic, also known as Asian flu, was first identified in Singapore in February of that year. It rapidly spread around the world and resulted in about one to two million deaths. Here in the UK it killed an estimated 30,000 people, making it comparable with the death rate of Covid-19. The 1968 flu outbreak was first identified in Hong Kong and killed around a million people worldwide.

It’s difficult – and potentially misleading – to classify influenza viral potency by simply comparing death rates because there are many confounding factors, such as modern vaccines for seasonal flu, susceptibility and age of the fatalities, the proportion of survivors left with lasting medical conditions and the measures taken to stem the rate of infection. They took very few measures to prevent the spread of Spanish flu and in 1918, there were no flu vaccines of any kind. In 1957, the UK were very slow to react, eliciting outrage from scientists such as John McDonald at the Public Health Laboratory Service (later to become director of the Epidemiological Research Laboratory and then chair of the Department of Epidemiology and Health at McGill University in Montreal). He wrote to the Royal College of General Practitioners, “Although we have had 30 years to prepare for what should be done in the event of an influenza pandemic, we have all been rushing around trying to improvise. We can only hope that at the end it may be possible to construct an adequate explanation of what happened.”

The 1968 flu pandemic was a little different in that it was more contagious than previous outbreaks, and the morbidity rate in younger people was higher. There was yet another influenza pandemic in 2003 (SARS) and then MERS in 2012, both of which fizzled out before they became major pandemics. In the historical context of flu outbreaks therefore, the idea Covid-19 came “out of the blue” and no one could have predicted it, seems ridiculous. Scientists have in fact been warning for years that a viral pandemic wasn’t just likely but inevitable. Papers have been published in the scientific literature including the most prestigious journal Nature in 1997, there were reports from the World Heath Organisation, the United States Federal Emergency Management Agency report 2019 and the UK’s Exercise Cygnus in 2016 to give just a minimal number of examples.

Despite this, in the main, warnings were ignored and the collective governments of the world were caught with their pants down. The warnings of John McDonald’s letter have been repeated and instead of having a well-prepared plan, the world just tried to improvise. Some countries improvised more than others, and some became the proverbial headless chickens. And realise when Covid-19 is over, the next pandemic is just as inevitable. The scientific outcry of 1957 have been largely ignored, and so let’s hope we don’t make the same mistake again. The German philosopher Friedrich Hegal wrote, “we learn from history that we do not learn from history”. Regarding viral pandemics, so far Hegal has been proven right. Let’s hope in the future he’s proven wrong.

1 – As defined by the WHO European area

What is Remdesivir and how does it work?

Another Covid-19 blog post for the non-expert.

Parts of the media are claiming a drug called remdesivir has arrived to save us

The BBC:

“Remdesivir: Drug has ‘clear-cut’ power to fight coronavirus”

RFI – Radio France Internationale, live news:

“US claims major breakthrough after positive trial of coronavirus drug remdesivir”

The Express:

“Hope of coronavirus breakthrough as Ebola drug shows results in landmark study”

Financial markets rallied on the news

“Gilead surges after WHO says remdesivir may combat coronavirus”

Others were more cautious, with the Financial Times headline, “Gilead’s coronavirus drug: why experts are cautious on its prospects”

Overall, my advice is, be careful about getting your science from the general media because it’s the sensational headline that usually sells.

Enthusiasm comes from a clinical trial of about 1,000 patients where recovery was shortened on average from 15-days to 11-days, but there was no recorded increase in the survival rate. Assuming investigators can repeat this result then it’s only moderately effective, but another study of just over 200 patients showed no statistical difference in effect. The 1,000 patent study is currently unpublished and the 200-patent study suffered problems recruiting participants and was foreshortened. The data so far really don’t support the headlines, but then again the equivocal nature of the trials doesn’t mean remdesivir doesn’t work – the fact is, at this time, we don’t know. The focus of this blog post however, is not on the clinical trials but on how an antiviral drug such as remdesivir works.

Gilead Sciences original developed remdesivir to treat ebola, another viral disease but with a much higher fatality rate than SARS-CoV-2. Remdesivir itself is not active against viral infection, but has to be metabolised in the body to its effective form. Known in the business as a prodrug, it has a phosphorus containing molecule added to its structure as it enters the cell. Why, you might ask? There’s nothing new in prodrugs, the well known and long-standing analgesic codeine for example is similar. It’s metabolised in the body to morphine, which is the active pain-killer. The reason for this somewhat circuitous process is because the active forms of the drug (phosphorylated remdesivir or morphine) find in hard to get through cell membranes to reach their site of action. And so instead, they are given in a form which can cross membranes, where they are metabolically converted to the active drug. Developed in a collaboration between Gilead, the U.S. Centers for Disease Control and Prevention (CDC) and the U.S. Army Medical Research Institute of Infectious Diseases, remdesivir was aimed at treating RNA-type viruses such as ebola, but also subsequently the coronaviruses responsible for SARS and MERS. Remdesivir is directed at a viral, as opposed to a mammalian biochemical system. The target in this case is the replication of viral RNA (mammalian cells replicate DNA and RNA acts as the messenger to make proteins). The structure of remdesivir is similar to base-molecules in RNA (nucleoside analogues) which then interfere in the replication process. Another drug of this type is zidovudine (AZT) successfully used to treat HIV.

Used during the West African Ebola epidemic of 2013-2016, remdesivir proved to be a promising drug but other treatments called monoclonal antibodies proved superior (monoclonal antibodies – or mAbs are outside the remit of this blog post – but perhaps another time). The clinical trials with remdesivir and SARS-CoV-2 weren’t exactly spectacular, but with a rational mechanism of action and data to support its action against RNA-type viruses, we shouldn’t dismiss remdesivir just yet. But of course, as they say, other drugs are available and it’s early days regarding any pharmacologic intervention. Not very scientific I know but let’s keep our fingers crossed remdesivir, or one of the other drugs, comes through.

I just know I'm right

Apart from death and taxes, the next most certain thing in life is how social media will suddenly light up with self-proclaimed experts on whatever topic grasps the public interest at the time. Over the past few years we’ve had upsurges in confident opinion on everything from economics, British constitutional law, international trade, vaccines and autism, climate change, and now infectious disease. It’s a phenomenon many laugh at but there’s something deeper going on. It’s what psychologists call the Dunning-Kruger effect.

The effect first came to light when, on April 19 1995, a bank robber, McArthur Wheeler, held up two Pittsburgh banks. He did so in broad daylight, making no attempt to conceal his face. He expressed great surprise when arrested because, as he explained to the arresting officers, he had covered himself with lemon juice and was therefore invisible. You might think Wheeler was delusional, but psychiatrists found him to be perfectly sane. He reasoned because we use lemon juice for invisible ink, it also made him invisible. His logic, as it was, faltered not on his insanity, but on his complete misunderstanding of invisible ink. To use an old saying, a little knowledge is a dangerous thing.

Two psychologists, David Dunning and Justin Kruger, picked up the case of Wheeler and after several phycological studies, gave their name to the phenomenon – the Dunning-Kruger effect.

In a nutshell, the Dunning-Kruger effect is where someone believes they have great insight or ability where in fact they have minimal knowledge and aptitude. The illusion of confidence leads to erroneous conclusions and resistance to any attempt at correcting them. They are, as it were, ignorant of their own ignorance. Charles Darwin put it well in his book, The Descent of Man, “ignorance more frequently begets confidence than does knowledge.” To put it another way it’s when people know enough to know their right, but don’t know enough to know when they’re wrong. Little knowledge can lead to great confidence but as one’s knowledge grows, so one becomes more aware of one’s own ignorance. When a certain level of expertise is reached, then confidence grows once more. Or you can let David Dunning himself explain it.

There are a wealth of recent examples of the Dunning-Kruger effect but rather than poke those hornets’ nests I thought I’d find some less controversial examples. I came up with the following two.

The TV show “Ramsay’s Kitchen Nightmares” stars Chef Gordon Ramsay as he tries to save failing restaurants. In every episode, the restaurant owner believes they are the greatest cook in the world whereas in reality they are serving barely edible food. The restaurant owner genuinely believe frozen seafood and tinned pasta is just as good as anything served by a Cordon Bleu chef. Before they stand any chance of becoming a competent cook, they have to recognise their own culinary ineptitude (and get past Ramsay’s expletives). Some really struggle with this, even when close friends and family try to help them through it.

In 2012 the UK Taxman discovered the Scottish football club Rangers were secretly overpaying their players and sent them a bill for £50 million. In trying to avoid paying the tax Rangers changed their name to Rangers Football Club but in 2017 the Supreme Court found in favour of the Inland Revenue. Many Rangers fans became experts in tax law overnight as they covered the media in their opinions, focusing on the witch hunt against their club (one example from the Daily Record).

Before any reader gets too smug, as names of those they believe are suffering from the Dunning-Kruger effect come to mind, we can all be guilty of it. It’s a form of confirmation bias, where we reinforce our own opinions by seeking out only that which supports it. Things may seem obviously clear and you wonder why others don’t have such clarity. But then you realise the world is complex and so it’s easy to think you know a lot about something when in truth all you have is a rudimentary understanding and there are depths you never dreamed of. The truth is we all over-rate our own abilities in one area of life or another. How many people do you know who would say they were bad drivers for example? We can’t all be above average drivers, from the very definition of average. It doesn’t mean you are stupid, in fact the Dunning-Kruger effect doesn’t correlate with IQ, you are – well, just being human.

The real test comes when those presented with contrary evidence, say, “OK, I was wrong and I have now changed my mind”. Those who already overrate their own expertise however, often just redouble their position and cannot adjust their beliefs no matter what the evidence. Unfortunately, evidence is more likely to contradict those who hold the strongest opinions than those less certain.

It may sound like I’m being holier-than-thou, and saying no one is allowed any opinion unless they are an expert. I’m not saying that at all, of course people have their own political opinions, or musical taste or favourite restaurant (those were the days) without being a professor of political science, or a master musician or a food critic. What I’m saying is (1) we should all be aware of the Dunning-Kruger effect and question our own beliefs accordingly and (2) be willing to adjust those views as we gain new knowledge. And we should all seek new knowledge driven by our curiosity, not our dogma. There’s no shame in not knowing, in fact in the modern complex world, we are all far more ignorant of things than we could possibly be knowledgeable of. The correct answer 90% of the time is, “I don’t know” but that doesn’t sit comfortably with the human psyche. And for that reason, the Dunning-Kruger effect will, I’m afraid persist. And for full disclosure, I’m not a phycologists and so I have no idea what I’m talking about.

(Thanks to Glyn Horner for his very much non-Dunning-Kruger input)

Covid-19 and blood pressure

As well as a vaccine, studies are being conducted into drugs which might stop, or at least inhibit, the effects of the SARS-CoV-2 virus. Modern drugs are designed to target specific sites, called receptors, in the body and the right choice of receptor is often the difference between a successful drug and a dud. I tell the story of one possible receptor and its link with blood pressure – an interesting tale and one that shows how so much of what happens in the human body is all linked together.

I must clarify one very important point first. There have been reports that those people taking medication for high blood pressure are more susceptible to Covid-19. The claim originated from a paper posted on the MedRxiv website, but this was not peer reviewed and so in question. The current scientific consensus is that blood pressure medication has no effect on susceptibility to the virus. This might change as we gather more data.

Before we get to the SARS-CoV-2 virus, some background on proteins. I’ve discussed the complexity of proteins on this blog before, but with the risk of being repetitive, they are more than just a juicy steak. Proteins are complex molecules made from strings of up to 20 different distinct types of amino acids. Some proteins contain hundreds or even thousands of individual amino acids; 30,000 of them go to make the muscular protein titin, for example. The strings of amino acids coil together in very specific ways, leaving pockets and grooves where complex biochemistry takes place. Proteins digest food, process energy, build cells, and control all the functions of life.

Some proteins, known as receptors, sit on the surface of cells where other molecules bind like a spacecraft docks to a space station. We call one such protein anglotensin-converting enzyme (or ACE for short, shown in the image on the right). ACE comprises chains of several 76c741_97a363f93936451e83781afb7b3a92d0~mv2.jpghundred amino acids which stick out of the surface of cells in the lungs, heart, kidney, intestines and blood vessels. It binds to chemicals in the blood known as vasodilators, which open up blood vessels allowing more blood to flow. This has the effect of lowering blood pressure. Once bound, those active pockets and grooves in ACE alter the vasodilators to vasoconstrictors, chemicals which cause blood vessels to constrict and hence increase blood pressure. The situation is a little more complicated than my description as it forms part of a cascade, but the upshot is a balance of vasodilators and vasoconstrictors, controlled by ACE, maintains the correct blood pressure – not too high or not too low.

Back in the 1980s, pharmaceutical companies developed drugs to target the ACE receptor, slowing down its action and decreasing the amount of vasoconstrictor in the bloodstream. We know these blood pressure reducing drugs as ACE inhibitors, the best known being ramipril, which has been in use since the early 1990s.

Now let’s turn to the SARS-CoV-2 virus and find out what that has to do with ACE? To explain, I’ll use an analogy. Have you ever been about to enter a room to discover the door handle is missing? Loose door handles are a nuisance because it’s very hard to get into the room when one falls off. This is similar to when viruses meet the cell membrane, they need a door to get into the cell but require a handle to open it. The handle with SARS-CoV-2 is the ACE protein sticking out of the surface of the cell. SARS-CoV-2 crown proteins lock on to the ACE receptor and use it to open the membrane so it can infect the cell. To be more exact, it uses one particular type of ACE, ACE2 as its door handle.

In the same way ramipril targets ACE, other drugs targeting ACE2 might inhibit binding of SARS-CoV-2 and hence offer protection against infection. An alternative might be to introduce decoy ACE2 receptors which bind SARS-CoV-2 away from cells. A genetically engineered ACE2 protein called human recombinant soluble angiotensin-converting enzyme 2 (hrsACE2) is one such candidate. A problem with all drugs however, as soon as you mess with cell receptors, you get other unwanted side effects. Push down one damaged domino and a lot of good dominos can fall in a chain reaction. ACE inhibitors can, for example, adversely affect kidney function. But then again, patients are more tolerant of side effects with life-saving drugs such as those against SARS-CoV-2.

Given the link between SARS-CoV-2 and ACE, doesn’t the connection between susceptibility to the virus and those taking drugs for high blood pressure make sense? The problem with that is ramipril and other ACE inhibitors target ACE, not ACE2 and so the connection is tenuous. This is the problem when you’re dealing with such complex systems as human infection, the devil is always in the detail.

What is herd immunity?

We’ve all heard a lot recently about herd immunity (pun intended), but it’s a term which seems to cause a lot of confusion. First off, let me tell you what it’s not. It’s not about a population becoming exposed to the Covid-19 virus and those with some resistance or those being “fitter” surviving while others “less fit” die. This strategy would no doubt leave the survivors with some immunity to Covid-19, but that isn’t what’s meant by herd immunity. I’d also add at this point, although it’s very likely those who have had Covid-19 and survived will gain immunity to the disease, at this early stage we don’t know the extent, how long the immunity will last, or whether it’s possible to suffer the effects of the virus a second time.

To use an analogy, imagine a tinder dry forest and a small fire which starts somewhere near the centre. The fire rapidly ignites one tree and because it’s so close to its neighbour, then that tree also catches fire. The fire spreads from one tree to another until the whole forest is ablaze. Now imagine a forest where some trees are resistant to catching fire. Perhaps miraculous rain clouds soaked some trees and not others – use your own imagination, this is only an analogy in way of explanation. The fire spreads from tree to tree but more slowly, because it can’t get round the soaked trees. If there are a very small number of wet trees, the fire will still rage but it’ll spread at a slower rate. As the number of wet trees increases, there comes a point where the fire burns a few trees but can’t spread any further and burns out. The wet trees protect the dry trees from spread and that’s herd immunity.

Although not exclusively so, the term herd immunity applies to vaccinated populations, where the “wet trees” are those with vaccine immunity, and the dry trees are those unvaccinated. Stopping the spread of a disease does not require 100% of the population to be immune but enough so the metaphorical fire can’t burn round them. Take smallpox as an example. The aim was to vaccinate 80% of the population in infected areas to achieve a herd effect. We irradiated the disease in 1977, although the vaccine uptake was in fact higher than the target 80%.

Epidemiologists talk about the R0 (R-nought), or reproduction ratio, which is the number of people likely to be infected by a single virus in a population with no prior immunity. The R0 for measles, which is highly infectious, is somewhere between 12 to 18 and seasonal flu is around 2. The R0 for Covid-19 is around 1.5 to 3.5.

But here’s the rub. For herd immunity to kick in for Covid-19 you need somewhere around 60% of the population to have immunity. Since we have no vaccine that means 60% of the population need to be infected and recover to gain immunity, which is 36 million people which will lead to tens, if not hundreds of thousands of deaths.

I have no answers I’m afraid because I’m well out my depth with exit strategies. I’m just trying to clarify what’s meant by herd immunity to clear some confusion. In the meantime I think we should trust in the world experts who understand this much better than I do, rather than those claiming to understand herd immunity as a way of getting the economy back on track.

What are the chances?

Another Covid-19 blog post for the non-expert, this time on what’s meant by test accuracy.

Before we get started, a question to stimulate your curiosity. The percentage of the population infected with Covid-19 is uncertain but let’s assume it’s very low at 1 in a 1,000 people (yes I know it’s likely higher than that but for the sake of argument). Let’s also assume a test for Covid-19 has an accuracy of 99%. If someone is tested and found positive, what’s the probability they have the disease? Obviously it’s 99% – right? but you’d be wrong, it’s actually just 9%. The reason why, gets us into statistics, but I have tried to keep the post as non-mathematical as possible. For the mathphiles however, I’ve included the equation at the end of the post. If you’re a mathphobe, please feel free to ignore it. Either way, some of the outcomes might surprise you.

There’s been a lot of discussion recently about new widespread testing of the population for Covid-19 using the so-called antibody test. Current data on Covid-19 infection comes from a test to determine if you have the virus. The antibody test, which we hope is introduced soon, looks to see if it has infected you in the past (the ideal time for the antibody test is around 28 days after infection). These antibody tests are currently being assessed for accuracy, and we’ve heard echoes of Brexit, “no test is better than a bad test”. So what constitutes a good or a bad test? To answer that question, let’s get back to the opening paragraph. Assume Covid-19 affects 1 in a 1,000 people and a test is 99% accurate. If someone is tested and found positive, what’s the probability they’ve got the disease? If you said 99%, then you fell into what’s known as the Bayesian trap. It’s named after the 18th century statistician Thomas Bayes, who at the time believed his finding were nothing special. When you’ve read this post you might disagree.

Although not intuitively obvious, the probability of having Covid-19 following a positive test depends upon your prior chances of having the disease. In our example 1 in a 1,000 people in the population have Covid-19 and so the prior chance is 1 in a 1,000. Assume you are one person in a random sample of 1,000 people taken from the population and you test positive. If the test is 99% accurate, then one in a hundred will give a false positive. This means, if the other 999 people, who do not have Covid-19, were tested, 1% would give a false positive, that’s 9.99 (call it 10) people. You are therefore one in a group of 11 people, all testing positive. You are the only true positive, the others are false positives and so 1 in 11 is 9%. I’ve shown the equation below and if you’re really bored in lockdown, you can experiment with the calculations yourself. Those more trusting can take my word for it.

I must emphasise I’m talking about testing randomly across the population. This is not being done in the UK with Covid-19 currently, it’s only those with symptoms in hospital and health workers who are being tested. This is not random and so different statistical calculations apply. If antibody random testing across the population is introduced however, watch out for the Bayesian trap.

The biggest issue with calculating these Bayesian statistics is knowing the occurrence in a population. In the example above the occurrence was one in a thousand, but in practice that number can be hard to measure. This is the case generally and even more so for Covid-19 because it’s such a new disease, and the infection rate is constantly changing. At the time of writing there are 61,000 positive tests in the UK with a population of 66 million. That’s an occurrence of 0.1% (1 in a 1,000) and if that’s the true occurrence, assuming an antibody test is 99% accurate, a positive test means the probability you’ve truly had the disease is just 9%. That’s worrying low. But the 61,000 positives aren’t randomly spread, these come from hospital cases and so perhaps the percentage tested versus those found to be positive is a better estimate. Rounding the numbers, the 61,000 positives were in a total number of 230,000 tests, so that’s 26% – let’s call it 30% for the sake of argument. With that infection rate, if we conducted random testing using a 99% accurate test and you were positive, then the probability that’s a correct result is 97%, which is very good. The problem is, like the 0.1% figure, the 30% infection rate is unreliable because it comes from hospital cases, which is not a random sample. The best estimates we have currently for the world-wide rate of infection is between 1.88 and 11.43%. Taking the higher value (11.43%), and assuming a 99% accurate test, the probability of you having Covid-19 if you test positive is 92.7%. Taking the lower value (1.88) the probability is 65.5%.

The antibody tests are being assessed for accuracy currently, but what effect does the test accuracy have on the probability of any one test being correct? If the test was 90%, instead of 99%, and the infection rate was 0.1%, then there’s virtually no difference; a positive test still means you have about a 9% probability of having the disease. If the infection rate is 30%, the probability of you having Covid-19 following a positive test is 79%, lower than 97% probability achieved with the 99% accurate test, but still not bad.

So what do all these figures really mean? A good way of looking at this is to plot the data on a graph which is shown in the Figure. The graph plots the proportion of the population infected with Covid-19 versus the probability any one test will give a true positive result. I plotted four lines, showing 99%, 95%, 90% and 80% accuracy in the antibody test in a random population. We can see, as the infection rate in the population falls, so does the probability of any single test giving a correct answer. The shape of the curves is also important. Take the 99% accuracy curve, for example. The curve increases rapidly to an infection rate of about 10% then flattens off. Without knowing the correct infection rate in a population, you need that flat part of the curve to extend across as wide an infection rate as possible. It’s only if the infection rate is lower than about 10%, the probability of any one test giving a correct answer falls off. The curve for the 80% accuracy curve on the other hand, falls away very quickly. A high test accuracy is therefore necessary because the probability of a true test remains acceptable across a wider range of infection rates. This, at least to some extent, offsets the lack of knowledge about the percentage of the population infected with the virus.

Returning to the example of 1 in a 1,000 and a 99% test accuracy, we have already established a positive test equates to a probability 9% of having Covid-19. But what if you did a second confirmatory test, would that double your chances to 18%? Actually no, because with a second test your prior chance of having Covid-19 is not 0.1% but 9% and if you plug that number into the equation below you get a probability of 91%. This assumes (1) the first and second tests are independent of each other and (2) both tests were positive, but all this gets a bit too complicated for a simple blog post.

Epidemiologists are interested in the infection rate across the entire population, and the antibody test is important in this respect. Individuals however, are more concerned whether or not they have had Covid-19 and the problem is, without knowing the epidemiological statistics on the infection rate, you don’t know the probability of any individual test being true. One set of figures feeds into the other and over time, with more data, the statistics get progressively better. In fact this is what Bayesian theory says, the more data you have the more accurate picture of the world you have.

Running down to the chemist, or buying a test on-line, suddenly doesn’t seem so attractive does it? But perhaps the slogan, “test, test, test” now makes sense, because we need to build data across the total population. In this respect Mr Spock was right when he said, “logic clearly dictates that the needs of the many outweigh the needs of the few.” Things are not always as simple as they might seem.

For the mathphiles

Keeping to the example of 1 in a 1,000 people in a population having the disease and a test accuracy of 99% then the calculation of the probability of any test being true is as follows:

PD = 0.09, or 9%

Note that tests often quote one accuracy for false positives and another for false negatives. I have simplified the calculations here by basing them only on false positives. In practice the two values, positives and negatives, are usually close.

Antimalarials for Covid-19

Another post concerning topics on Covid-19 aimed at the non-expert.

Scientists are cautious by nature, some politicians and political journalists are not. While scientists say there’s no convincing evidence antimalarial drugs such as hydroxychloroquine have any effect on Covid-19, President Trump says, “there are some very strong, powerful signs of hydroxychloroquine as a Covid-19 therapy”. Rudy Giuliani Tweeted, “hydroxychloroquine has been shown to have a 100% effective rate in treating Covid-19.” Meanwhile, in my country, the UK, Sarah Vine (Daily Mail columnist and wife of the Chancellor of the Duchy of Lancaster, Michael Gove) Tweeted, she doesn’t trust the World Health Organisation and prefers to believe 6,000 doctors who say hydroxychloroquine works on Covid-19. The claim about 6,000 doctors comes from a Daily Mail poll accompanied by an article.

Putting aside the fact Trump, Giuliani and Vine are not pharmacologists, and the Daily Mail doesn’t exactly have an impressive track record in science reporting, is there any truth in these claims? I thought I’d take a closer look. I must start however, with a disclaimer. Things are moving very fast and this blog post will probably be out of date as soon as it’s posted. There’s a range of drugs being tested to see if they have any efficacy against the SARS-CoV-2 virus and tomorrow, who knows, one maybe found to be a miracle cure. I doubt it however, and so here, I’m just focusing on hydroxychloroquine.

Scientists developed a range of antimalarial drugs in the 1930’s and 1940s to treat troops fighting wars in tropical climes. Amongst those drugs were the aminoquinolines because related quinine and quinidine had been used for hundreds of years as antimalarials. Quinine gives tonic water its bitter taste and there’s an adage gin and tonic came about as a malaria treatment in colonial India. This is unlikely because the quinine dose in tonic water is far too low to be effective. Chloroquine and hydroxychloroquine are two aminoquinoline-type drugs which were once useful but are now largely ineffective because emergence of resistant malarial parasites.

Pharmacologists don’t just pull drugs out of thin air and hope they might work. Instead they look for potential mechanisms, that is how the drug interacts with the target disease, preferably at a molecular level. The malarial parasite (Plasmodium species) enters red blood cells and hydroxychloroquine is believed to inhibit the way it interacts with haemoglobin, but the drug has other effects within the body. Chloroquine and hydroxychloroquine have anti-inflammatory properties because they decrease acidity inside certain cells altering the rate of protein degradation, and cytokine production (cytokines are protein-type molecules released by cells in the immune system that regulate inflammation).

There are several feasible mechanisms that might explain how hydroxychloroquine has antiviral properties.

For coronaviruses to enter mammalian cells, they first bind to the cell surface and then transverse the membrane, which is mediated through an enzyme called lysosomal protease. Inhibition of this enzyme by hydroxychloroquine is a potential target but a correlation between lysosomal protease activity and the entry of coronaviruses into the cell has never been established. Cytokines are protein-type molecules released by cells in the immune system that regulate inflammation and hematopoiesis (the process by which the body makes blood cells). The overall picture is further complicated, because the way coronaviruses enter mammalian cells differs from one species to another.

Some say that we have used chloroquine and hydroxychloroquine for many years, which is true, but this is not testament to their safety. The objective when developing modern drugs is to target its action as specifically as possible, so reducing side effects. This type of sophisticated drug design wasn’t a thing 80-years ago when chloroquine and hydroxychloroquine first appeared. These drugs act in many places in the body, which inevitably leads to several unwanted side effects. Although they vary from person to person and in severity depending upon dose, side effects can include anorexia, diarrhoea, nausea and skin pigmentation, as well as liver and kidney toxicity. Some side effects are hard to spot in the early stages, such as changes to the muscles of the heart, but these can later lead to arrythmia and, although rare, cardiomyopathy (where heart muscles find it harder to pump blood). Not everyone experiences these toxicities however, and many take hydroxychloroquine as an effective anti-inflammatory for rheumatoid arthritis and lupus.

Chloroquine and hydroxychloroquine have antiviral properties when tested in test tubes in the laboratory (or in vitro, which translates to in glass) and has shown activity against, Ebola, SARS, MERS and more recently SARS-CoV-2. As any drug developer will tell you however, there is a world of difference between in vitro results and clinical trials. Clinical trials with hydroxychloroquine or chloroquine with HIV, hepatitis-C, dengue, Chikungunya and influenza viruses show either no discernible or very modest efficacy.

Enthusiasm for hydroxychloroquine in treating Covid-19 appears to originate from small clinical studies in France and China. A clinical study in Marseille, France, for example, reported 100% viral clearance in nasopharyngeal swabs in 6 patients after 5 and 6 days (the trial was with a hydroxychloroquine and azithromycin combination). Although I don’t know for sure, this may be where Giuliani’s “100% effective rate” came from. Other studies in China reported some success but these are in contrast with others where no discernible effects were found. One of the principal problems is that Covid-19 exhibits a range of symptoms and severity, with some only getting a very mild disease while others are dying. It’s hard to distinguish true drug-effects from such a noisy background without carefully controlled comprehensive studies. And it’s very dangerous, and wholly unscientific, to choose those studies that give the results you want and ignore those that don’t. Other clinical trials are ongoing and results are awaited. Despite the lack of clinical evidence, the United States Food and Drug Administration (FDA) granted hydroxychloroquine “emergency use authorisation” against Covid-19. This sends entirely the wrong message in my opinion, but having dealt with the FDA in the past, it can be rather politically motivated at times. European regulators are being more cautious and are not authorising hydroxychloroquine until it’s been better tested.

So, in the words of Donald Trump, what have you got to lose? Although hydroxychloroquine is used to treat rheumatoid arthritis and lupus (its anti-inflammatory properties are useful) it is not an easy-going drug and the aforementioned side effects are something you want to avoid in a respiratory-compromised patient. There have been a few cases of people self-medicating who have died from chloroquine poisoning. Perhaps more importantly, without clinical evidence it offers false hope and might even distract from other potentially more effective interventions. Proclamations by senior politicians can also cause over-demand and thereby deprive those in genuine need of hydroxychloroquine, such as lupus patients. Overall, there is quite a lot that can be lost.

It’s my view that some politicians are struggling currently, because they so often get their opinions confused with facts. They may have such strong views on the economy or societal issues their opinions became certainty and they then search for “evidence” to support them. Such cherry-picking is the antithesis of the scientific method. And remember, viruses don’t give a flying tinker’s cuss for anyone’s opinion, they aren’t even alive to care about anything. Believe me, no one wants hydroxychloroquine to be effective against Covid-19 more than me and I’d be delighted if clinical trials confirm its efficacy. In the meantime however, the Covid-19 outbreak follows the laws of nature, not humankind but as the number of deaths increase, politicians become increasingly desperate for quick and easy answers. The trouble is there aren’t any, and that’s not an opinion it is, I’m afraid a fact.

(My thanks to Dr C. Edwin Garner, for his comments on this post in a private capacity).

Coronavirus testing

I’m getting questions about Covid-19 testing and so another blog post aimed at the non-expert in way of explanation.

Confusion appears to be coming from the fact there are two tests, one already being used and another about to be released. The one that is available at the time of writing tests for the virus, whereas the one about to be released, tests for the antibody to the virus in the bloodstream. The viral test says whether the patient is currently infected, but it will not say if they’ve had Covid-19 once they have recovered because the virus will have gone by then. An antibody test looks for antibodies to the virus in the bloodstream, and these might not have had time to form in the early stages of infection (when the viral test would be positive). But the antibody test will tell if the patient has had Covid-19 in the recent past. Since some people get mild symptoms, it’s possible they may not know they have had the disease, even though they would have been a carrier. A combination of viral and antibody testing will tell us where the virus is and where it’s been and that gives us the best chance to predict where it’s going.

How do these tests work? The two tests work in very different ways but they are both a marvel of modern molecular engineering. The virus test uses a technique called a polymerase chain reaction (PCR) and it looks for the genetic material of SARS-COV-2. All living things have DNA, which contains the genetic code of life. DNA replicates itself, passing from one generation to the next. So if you look at your grandchild and say, she’s got her father’s eyes – it’s all because of DNA. In what’s called the central dogma, DNA makes RNA which makes protein and it’s protein that governs the biochemistry of life. Proteins are hugely complex molecules made from strings of up to 20 distinct types of amino acids. Some proteins contain hundreds or even thousands of individual amino acids; the muscular protein, titin has 30,000 of them. The long strings of amino acids fold like tangled pieces of string but unlike string, the tangles are very precise. Proteins have molecular grooves and pockets where specific biochemical reactions take place. The grooves and pockets are analogous to spanners and wrenches in a biochemical tool kit, each fitting a particular sized nut or bolt in building the machinery of life. The 2015 Nobel Prize winner for physiology or medicine, Yoshinori Ohsumi, summed it up by saying, “Life is an equilibrium between synthesis and degradation of proteins.”

Some viruses have DNA but some have RNA – technically viruses are not actually living because they can only reproduce inside living cells. SARS-COV-2 has RNA as its genetic material and it manufactures proteins from that by highjacking human cells. RNA, like DNA replicates itself, and it’s this replication process that’s used to detect its presence. The biochemistry of replication is essentially put into a test-tube where any SARS-COV-2 RNA multiplies many millions of times, each new stand of RNA being identical to its predecessor. Once enough replicated RNA is available, it’s assayed using a variety of methods. PCR test kits are quick to develop and were first distributed by the World Health Organization last January. PCR is routine but not straightforward, requiring laboratory facilities and 4–6 hours to complete. Add that to the logistics of sample shipment and the turnaround time is typically at least 24-hours. Demand on some components of the PCR test have led to shortages.

The other test doesn’t look for the virus itself, but an antibody to that virus in the blood. When a virus infects the body, we produce antibodies as part of the immune response. Antibodies are Y-shaped proteins (see image) with very specific shapes which latch onto the surface of the virus. Detecting antibodies in the bloodstream in amongst a plethora of other proteins is challenging but in 1971 two Swedish scientists, Eva Engvall and Peter Perlman, solved the problem with the invention of a technique called ELISA (Enzyme-Linked Immunosorbent Assay). This assay uses more antibodies – in fact antibodies to antibodies, and so it gets a little complicated. Let’s start with the antibody to SARS-COV-2 in the blood, we’ll call this viral-antibody. In the laboratory molecular engineering methods are used to make another antibody (we’ll call this antibody-2) which binds onto the viral-antibody. Then another laboratory-made antibody (we’ll call this antibody-3) binds onto antibody-2. Antibody-3 is different however, because it’s fitted with an enzyme, yet another protein, but one that medicates some chemical reaction, typically one which causes a colour change.

By combining antibodies-2 and 3 in a test-kit a spot of blood is taken by finger prick and mixed with antibody-2. If the viral-antibody is present in the blood, antibody-2 will bind to it. Then antibody-3 binds to antibody-2 and its enzyme mediated reaction causes a colour change. The assay is a sort of protein domino effect, one reaction triggering another. All this takes place on a single device and so is conducted in situ outside of a laboratory. It takes perhaps 10-minutes to give an answer – not that much different to a pregnancy test.

It requires some effort to develop an ELISA assay, which is why there’s been a delay in their arrival. Kits are just being issued and after a period of evaluation, they should be widely available. Don’t rush out to buy one just yet, because the evaluation and then scale-up is likely to take a little while.

The pandemic still rages and both infection and mortality rates are increasing, but there is reason for optimism. A combination of tests will pin down where the virus is and where it’s been and with that knowledge we’ll be able to better target our efforts against it.

I’ve outlined the basics in this post but technology is advancing all the time. New techniques for viral and antibody detection are being explored and if one of those comes through, I’ll try and blog on that at a later date.