Is it Safe? Update on AZ vaccine

An interesting week ahead

If you find these posts interesting and think others might do so, spread the word.

Share Is It Safe?

A midweek posting because things have moved fast on the AZ Covid-19 vaccine, and tomorrow is a big day at EMA, so I thought I would gather some thoughts and links together before that.

Summary of where we are

Around 13 countries have suspended, or paused, the use of the AZ vaccine at this moment in time. A broad concern about reports of thrombosis were the initial concern, but some countries (Norway and Germany) have reported on a very rare cerebral venous sinus thrombosis (CVST) associated with platelet deficiency.

The concern about a more general safety issue around thrombosis something that needs looking at, but would appear to be no higher a risk than would occur if the same patients had not been vaccinated. The cases of CVST are more interesting, and need more work on. The Paul-Ehrlich-Institut (PEI) have now published an English language document on their analysis. [PDF] That document includes 6 younger to middle aged women (20-50 years of age), and another case of cerebral haemorrhage with platelet deficiency and thrombosis. It occurred 4-16 days after vaccination, and the PEI carried out a rapid observed-versus-expected analysis:

The number of these cases after vaccination with COVID-19 AstraZeneca is statistically significantly higher than the number of cerebral venous thromboses that normally occur in the unvaccinated population. For this purpose, an observed-versus-expected analysis was performed, comparing the number of cases expected without vaccination in a 14-day time window with the number of cases reported after approximately 1.6 million AstraZeneca vaccinations in Germany. About one case would have been expected, and seven cases had been reported.

Three of the 7 cases died.

The PEI notes that there isn’t a causal relationship proven yet, and this is still very rare, and in the document they produced they make comments about the comparison with the oral contraceptive (which carries risks of thrombosis) which has been made. They note that women are meant to be made aware of this risk and it is listed in the Summary of Product Characteristics of oral contraceptives. They then rightly refer this on to the European Medicines Agency to make a judgement.

They also recommend to pause the use of the AZ vaccine in Germany. I’ll park that now and come back to it.

So what are EMEA going to do?

Yesterday EMA held a press conference about AZ vaccine. You can watch the recording here.

From this it seems clear to me that EMA intend to stick to their well thought out plans on how to deal with the safety of the COVID-19 vaccines. Some of the things (not exhaustive) that came out of this conference:

  • More reports of thrombosis have been submitted since this became a media topic (This is known as notoriety bias).

  • EMA are looking at all vaccines with regard to this issue.

  • EMA still think the balance of benefits versus harms means the AZ vaccine should still be licensed.

  • Peter Arlett’s mastery on pharmacovigilance and how this issue should be approached in broad terms was illuminating.

  • The number of cases of thrombosis is lower than might be expected.

  • The rare and usual cases of CVST is something they will look at very carefully, and due to the limited epidemiological data for the condition, clinical case analysis is the main focus.

EMA will be reporting on this on Thursday at some point (presumably later in the day given the complexity of the analysis and debate will have to be had.

What will happen?

Well, let’s see, but I think some points can be made about what is likely to weight on decisions being made. It was pretty clear from the conference yesterday that EMA have a competent plan and will be driven by the standard paradigm about balancing harms and benefits, and use of evidence.


  • The decisions are likely to due to focus on CVST concerns, although the more general concern about thrombosis will need to be addressed.

  • The decisions are not a binary choice between stopping the use of AZ vaccine and continue use of AZ vaccine. Other options exist, such as noting the possible link of a rare adverse effect with the vaccine will exist, even without saying that the link is proven, while at the same time noting that the overall balance of benefit and harms still continues to be favourable.

  • There are options around further pharmacopidemiology studies that can be carried out to look at the risk of CVST, and continued analysis of clinical cases reported to look for aggravating risk factors and/or alternate explanations.

  • There will be decisions made about how to communicate potential risks to the public. This is always the tricky bit.

Back to the pause

My criticism of the vaccination paused isn’t based around some simplistic notion that we should not discuss potential vaccine harms. Having open transparent discussion of such signals is extremely important for public trust. My concern was that the vaccine safety plans around detection, confirmation, decision making, and communication that were in place at EMA seemed to be being supplanted by knee-jerk decisions at a local level.

So, I’m really glad that EMA seem to be back in the driving seat on this issue.

Now they may be technically within their rights to do this, but it does not mean it was the right decision. The pandemic is still ongoing, and cases (including deaths) are going up in many European states. We also have the AZ vaccine being shipped throughout the world via the COVAX scheme, so anything we do should be done in a cautious controlled manner, that is also open and transparent.

My own personal view is that all of these countries should have carried on without pausing the vaccine, referring the matter to EMA as they have done. Given the rarity of the events in 17 million vaccinations, a pause of a few days is likely to cause far more harm from missed vaccinations and damage to public confidence in vaccines than being open about the issues referred to EMA and not pausing. This is the approach, for example, that Belgium had.

Reversing the pause if EMA makes a decision that vaccine’s balance of benefits and harms is still favourable will be difficult. ‘You said there was a problem 4 days ago!’ some people will say, ‘How can you say it is safe now!’

This goes back to my criticism of the use of the precautionary principle by some of the countries involved. I’ve written a quick piece on this at The Conversation.

So rather than avoiding risk, the principle has instead moved countries away from one risk (blood clots) towards another (lower vaccine coverage). The impact of the latter could be much larger.

Even if this weren’t the case, the principle has still, arguably, been misapplied. Plans for COVID-19 vaccine safety monitoring until now have been based around rigorous scientific evaluation of safety signals, careful communications to ensure vaccine hesitancy is not increased, and ensuring that signals are investigated to examine if any risk requires regulatory action. 

Because potential safety signals arise often in vaccine and drug safety, with many being false signals, the precautionary principle doesn’t fit with such plans. It is too sensitive, and in the case of COVID-19 vaccines, doesn’t initiate any safety assessments that aren’t already happening.

As we have seen this week, misapplication of the precautionary principle leads to erratic decision making that fails to do the very thing it intends to: lower risk. The decisions made could potentially have long-term health effects both in the EU and globally. As a result, one might say we need to be more cautious about the application of the precautionary principle.

I’ve also been a few radio stations and TV stations recently, and this is my piece for the BBC where I talk about this.

Other things worth reading:

Stuart Richie Europe’s AstraZeneca vaccine suspension is bad science that will cost lives.

Derek Lowe What is Going on With the AstraZeneca /Oxford vaccine

Tom Chivers How safe is the AstraZeneca jab

Matthew Herper The curious case of AstraZeneca’s Covid-19 vaccine

David Spiegelhalter There's no proof the Oxford vaccine causes blood clots. So why are people worried?


That’s all and stay safe.


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