Is It Safe? 2nd of August 2020

Covid and Evidence, the Infodemic, Vaccines, and Non-deceptive Placebos

Welcome to the first edition of ‘Is It Safe?’, a newsletter about medicines and their harms. Mostly, this will be links related to drug safety that I think are interesting. If there is a foray into other areas, they will have some cross-over relevance.

For the moment it will be twice a month, and should be a short enjoyable read. If you find it useful and think others might do so, spread the word. If you aren’t a subscriber, sign up here.

A short note. While drug safety issues in the news or on regulator sites will turn up here, it is not a replacement for being signed up to their feeds, and I would encourage you to do so if you are a healthcare professional. If you are a UK reader, sign up to the MHRA’s Drug Safety.

So that said, let us begin the journey….and as you expect it features Covid-19.

Covid-19 and evidence

If there is one drug that will be remembered from 2020, it will be hydroxychloroquine. Along with masks, it has been politicised so much that rational discussion seem impossible. Seeing US politicians debating with scientists whether it works or not, rather than a regulator making decisions with expert committees is kind of mind blowing.

Initial, highly confounded, observational studies of hydroxychloroquine showed benefits, but efficacy has not been shown in well conducted trials.

The arena of politics is not the best place to find the effectiveness of a drug. As an aside, Raoult (one of the prime early advocates for hydroxychloroquine) is now trying to use the courts against other medics and using his h-index as evidence of his authority (A dubious activity). Resorting to politics, rather than experiment, reason, and open debate might suggest a generally bad approach to finding the truth.

As the evidence mounted against hydroxychloroquine, there was always a post hoc reason why it didn’t work in the trials put forward, while credulity reigned when looking at observational studies. It wasn’t used soon enough. The dose was wrong. It needed to be used with zinc. It needed to be used on a full moon, after dancing round a sacred stone. It needed to be used with azithromycin. A recent review of the latter has been published in Drug Safety.

‘There is no clear evidence that azithromycin may exert beneficial effects in COVID-19 beyond antibacterial activity in bacterial superinfection.’

As well as this, the potential of azithromycin to cause arrhythmias through QT elongation in older patients, taking concurrent hydroxychloroquine, and other medicines, means that another of the drugs widely touted in the early days of this crisis has proved largely ineffectual.

To find out if a drug works, clinical trials still reign supreme.

ACE inhibitors, ARBs, and Covid-19

Covid-19 also had an impact on the potential prescribing of existing drugs. Two notable concerns arising early on in the pandemic were NSAIDs and ACE inhibitors. In both cases, there was concern, based on mechanistic considerations, that they could worsen Covid-19 outcomes. Both of these appear unfounded now. A large cohort study carried out by Hippisley-Cox et al, carried out in the UK in 8.3 million patients, looked at the use of ARBs and ACEs, the risk of contracting severe Covid19 disease, and the risk of admission to intensive care.

In fact, ACE and ARBs are associated with a reduced risk of Covid-19, and are not associated with admission to intensive care. The (open access) paper is well worth reading, lots of interesting details around risk of Covid-19 and ethnicity, body mass index, deprivation, sex, and a counter-intuitive protective association with smoking status (a true effect or confounding?). The protective effect of ARB/ACE inhibitors did vary between ethnicities, although as the authors note the numbers were small with wide confidence intervals. In short, the fears of harm were unfounded, and patients and prescribers can be re-assured ACE/ARBs are fine to take during the pandemic.

It’s worth noting the asymmetry in evidence here. Clinical trials to prove efficacy. Observational studies to find or disprove harms. This is reflective of the relative strengths and characteristics of each tool, and when each can be used. During the past 3 months, hammers have been used with screws, and nails with screwdrivers.

Beware the Infodemic

The wave of information about drugs that we have been subjected to during the pandemic, is referred to by some as an ‘Infodemic’. Tuccori et al take us for a walk through chloroquine/hydroxycholoquine, the ACE/ARB debate, NSAIDs, favipiravir, and umifenovir:

‘The spreading of such an infodemic was likely favored by both the current hyper-interconnection of people (widespread use of smartphones, easy access to the internet, social media, and cross-platform messaging services) and populism. Such terrain was particularly fertile for conspiracy theorists and profiteers. All media are involved in the spreading and amplification of information, and the communication process may have different patterns. Information can be generated in different ways, from the results of small or poorly conceived studies to reports highlighted in newspapers (such as with chloroquine, ibuprofen, or even ACEIs and ARBs), or may start in social media and “go viral” in a few hours (such as with antivirals). Among communication materials, short videos appear to be a particularly effective method of spreading information. Whatever the starting point, inadequate, sensational, or distorted information first affects opinion leaders, often politicians, and those active on social media, which, deliberate or not, lends consistency and reliability to the information. The final effect can be inadequate choices made by individuals everywhere.’

They also highlight the importance of pharmacovigilance in managing safe drug use during emergencies, and it is good to see that they mention the importance of patient stories of adverse outcomes as a way of communicating safety issues.

There’s also a tale to be told here about the reliance on mechanistic explanations for suspected benefits and harms of drugs, rather than data, which has influenced some scientific and lay opinions, as well as being a potential side effect of the popularisation of science more generally. More on this in the next newsletter…

Vaccine Safety

As we await a vaccine, there are some already calling for premature release of vaccine candidates outside of trials, as well as discussions about mandatory vaccines, and some ‘anti-vaxer’ discourse that seems to be repeating the mistakes made in the 2000s. We should be trying to avoid mistakes of the past, and some of these supposed solutions would likely be counterproductive to any good vaccine strategy. I was impressed by Stuart Ritchie’s piece on this, which is far better than some of the ‘vaccines are safe, duh!’ sentiments that abound.

‘We mustn’t underplay any side-effects of the Covid-19 vaccine. We mustn’t provide confident-sounding yet contradictory advice — as we already have during this pandemic in the case of facemasks, first advising that they were useless and maybe even dangerous (which may or may not have been a “noble lie” to preserve mask stocks for healthcare workers), then mandating their use in shops and on public transport. We mustn’t give any credibility to conspiracy theories. And we must above all be honest and transparent about how the vaccine works, who is prioritised to get it and when, and what our targets are for its eventual rollout across the population.’

Stuart has a book out called ‘Science Fictions’, which I’ll be reviewing on my blog this week. A key element of this book is how science has its own internal problems, rather than trying to suggest poor outcomes are based on a public reaction to good science.

One thing we could do now, is to start the process of informing the population about the systems we will use to monitor the safety of vaccines, and it is good to see a dedicated Covid-19 reporting portal at the MHRA in the UK. Communicating uncertainties to the public will increase trust. False promises of total safety are likely to backfire, and can be exploited by bad actors rapidly using systems of information flow that regulatory authorities and governments will be unable to counter.

In non-Covid News, there’s a neat little study looking at non-deceptive placebos and how they appear to provide benefit in emotional distress. Granted it is in young university students, but avoiding adverse effects through the honest use of placebos in some conditions sounds intriguing, especially if it stops someone prescribing a drug with known adverse effects and limited efficacy.

From the Blog: Ten Steps to Safety: A brief post on the outcomes of the medicine and medical device safety meeting I was lucky enough to attend.

Finally, why ‘Is It Safe?’ In the thriller Marathon Man (1976) starring Dustin Hoffman and Laurence Olivier, Olivier (Szell) tortures Hoffman (Babe) , while constantly firing the question ‘Is It Safe?’. Babe, knowing nothing, tries to answer the question to the satisfaction of Szell.

Szell : Is it safe?

Babe : Yes, it’s safe, it’s very safe, it’s so safe you wouldn’t believe it.

Szell : Is it safe?

Babe : No. It’s not safe, it’s… very dangerous, be careful.

So, for the purposes of this newsletter it fits perfectly. Drugs are rarely ‘safe’. We have uncertainty about the nature of harms, uncertainty about their prevalence, and even uncertainty about casual links between a drug and alleged harms. Safety is also a concept relative to the benefits of a drug. Grappling with the concept of safe, often with limited information, is the problem we have when assessing the harms of medicines.

That’s all, and don’t forgot to tell any anyone you think might be interested about the newsletter. See you in a couple of weeks.

Stay safe.


Don’t forget to report your suspected adverse effects from medicines. In the UK, this is the Yellow Card Scheme