New research gives a better understanding of the excess population risks after immunisation
In their paper, Pottegård and colleagues report a study of the incidence of thrombotic events in a cohort of Danish and Norwegian people who received the Oxford-AstraZeneca covid-19 vaccine and compared this to the expected incidence in the general population.  The authors reviewed national healthcare registers, looking for specific ICD-10 codes. They found no increase in the incidence of arterial events such as acute myocardial and cerebral infarction, but did find an increase in venous thromboembolic events of 11 (95% confidence interval 5.6 to 17) excess events per 100 000 vaccinations. Among the venous thromboembolic events, the most notable was cerebral venous thrombosis, which was about 20 times more common after the Oxford-AstraZeneca vaccine than expected with an excess risk of 2.5 (95% confidence interval 0.9 to 5.2) per 100 000 vaccinations. Venous thromboembolic events were more common in women than in men and were also more common in people aged 18-44 than in those aged 45-65. Thrombocytopenia and bleeding were mildly increased after vaccination, but not significantly.
This study has been able to give a better understanding of the excess population risks after immunisation. The risk of thrombosis at 11 excess events per 100 000 vaccinations are higher than estimated by the European Medicines Agency. The EMA estimates the risk of thrombosis to be between 1 and 2 per 100 000 people, depending on age.  But what this study and what the EMA are measuring are not the same. Pottegård and colleagues used only ICD-10 codes and so have not been able to identify confirmed cases of thrombosis with thrombocytopenia, which is the syndrome specifically linked to vaccination. Similarly, Pottegård and colleagues would likely not have been able to identify thrombosis in people who had not had a vaccine and who were primarily admitted for covid-19 and might not have been investigated for thrombotic events, or had those events recorded under a primary diagnosis of covid-19. Nevertheless, Pottegård and colleagues’ study does suggest that the risk of thromboembolic events after the covid-19 vaccine might be higher than suggested by pharmacovigilance reporting in either the UK or European safety monitoring schemes.
On the other hand, Pottegård and colleagues found 15 deaths from all causes after vaccination with the Oxford-AstraZeneca vaccine compared withan expected 44 deaths. They argue that some of this reduced death rate after vaccination could be due to a healthy worker effect. [3,4] A large proportion of the people in Denmark and Norway who received the Oxford-AstraZeneca vaccine were healthcare workers. Apparently, this vaccine was not generally rolled out to vulnerable adults younger than 65. However, the magnitude of this difference—a standardised ratio of just 0.34 (95% confidence interval 0.19 to 0.57)—is greater than would be expected from previous studies of the healthy worker effect. [3,4] Furthermore, of those who received the Oxford-AstraZeneca vaccine, more would have received it because they were at an increased occupational risk of covid-19 infection and death.  Indeed, nurses and care workers generally have a relatively higher chance of dying from covid-19 and all causes than many other workers, irrespective of covid-19. 
Pottegård and colleagues’ study improves our understanding of the population risk of thromboembolic events after the Oxford-AstraZeneca vaccine, although given that the authors were unable to identify the specific cases of thrombosis with thrombocytopenia this must still remain uncertain.
Nevertheless, the study does suggest that such thromboembolic adverse events after receiving the vaccine are likely to be more common than estimates based on pharmacovigilance reporting systems. But it does not change the conclusions of both the EMA and the UK’s Medicines and Healthcare products Regulatory Agency (MHRA), that the benefits of the Oxford-AstraZeneca vaccine far outweigh its risks for most age groups. It remains the case that for most age groups, the probability of surviving the year is much greater for people who accept any vaccine when offered than if they decline it.
To decline a vaccine today because it is the Oxford-AstraZeneca or Johnson & Johnson vaccine in the hope of being able to get another vaccine sometime later carries a real risk of dying from covid-19 before being able to get a preferred vaccine. Those countries that delayed their own vaccination programmes at a time of high transmission rates by declining to use available Oxford-AstraZeneca vaccines should know that their decision will have contributed to an increase in the number of avoidable deaths from covid-19.
Paul R Hunter, Professor of Medicine, NIHR Health Protection Research Unit, Norwich Medical School, University of East Anglia, Norwich
Competing interests: PRH is funded by the National Institute for Health Research Health Protection Research Unit (NIHR HPRU) in Emergency Preparedness and Response at King’s College London in partnership with Public Health England (PHE) and collaboration with the University of East Anglia. The views expressed are those of the author and not necessarily those of the NHS, the NIHR, UEA, the Department of Health, or PHE
Linked research: Arterial events, venous thromboembolism, thrombocytopenia, and bleeding after vaccination with Oxford-AstraZeneca ChAdOx1 in Denmark and Norway: population based cohort study by Pottegård et al.
Linked editorial: Thromboembolism and the Oxford-AstraZeneca vaccine
- Pottegård A, Lund LC, Karlstad Ø, et al. Arterial events, venous thromboembolism, thrombocytopenia, and bleeding after vaccination with Oxford-AstraZeneca ChAdOx1 in Denmark and Norway: population based cohort study. BMJ 2021;373:n1114.
- European Medicines Agency. Annex to Vaxzevria Art.5.3 – Visual risk contextualisation 23rd April. https://www.ema.europa.eu/en/documents/chmp-annex/annex-vaxzevria-art53-visual-risk-contextualisation_en.pdf
- Thygesen LC, Hvidtfeldt UA, Mikkelsen S, Brønnum-Hansen H. Quantification of the healthy worker effect: a nationwide cohort study among electricians in Denmark. BMC Public Health 2011;11:571. doi:10.1186/1471-2458-11-571
- Picciotto S, Brown DM, Chevrier J, Eisen EA. Healthy worker survivor bias: implications of truncating follow-up at employment termination. Occupational and environmental medicine. 2013 Oct 1;70(10):736-42.
- ONS. Coronavirus (COVID-19) related deaths by occupation, England and Wales: deaths registered between 9 March and 28 December 2020 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/causesofdeath/bulletins/coronaviruscovid19relateddeathsbyoccupationenglandandwales/deathsregisteredbetween9marchand28december2020#overview-of-coronavirus-related-deaths-by-occupation
- Hunter PR. Thrombosis after covid-19 vaccination. BMJ 2021;373:n958.