Pausing AstraZeneca vaccinations because of suspected links to deadly blood clots may have allowed COVID-19 to spread unchecked and resulted in hundreds of additional deaths among people vulnerable to the virus, vastly outweighing the risk of death from adverse vaccine side effects, according to a new European study.

In an article published April 27 in Chaos, researchers concluded that the benefits of deploying the AstraZeneca vaccine greatly outweigh its numerous associated risks, such as rare blood clots, with wider relative benefits in situations where the virus' reproduction number, or its ability to spread to unvaccinated people, is larger.

AstraZeneca's COVID-19 vaccine in March was suspected of being linked to a small number of deep vein thrombosis cases that emerged in Europe following the vaccinations of millions of people. On March 15, several European Union countries elected to suspend all AstraZeneca inoculations for three days for an investigation into its causal links to DVT.

In response, scientists in Europe conducted a collaborative study, published April 27 in Chaos, that investigated whether a pause in the administration of the AstraZeneca vaccine, even for a brief period, could result in excess mortality, given the virus' propensity to spread more quickly within a population of susceptible individuals. The researchers also estimated the number of potential deaths resulting from vaccine-induced side effects, such as deep vein thrombosis, to provide an order-of-magnitude comparison between policies of vaccine interruption and continued distribution. New variants and other long-term effects of COVID-19 due to higher infection rates were not included in the model's treatment.

Davide Faranda, a co-author of the study, a researcher at the University of Paris-Saclay and an external fellow at the London Mathematical Laboratory, told The Academic Times that his team hoped to provide an improved framework for vaccine risk-benefit analysis. 

"When a new vaccine or therapy is out, it is clear that it has to undergo long testing phases to verify side effects," Faranda said. "The COVID-19 situation is pretty new in this sense because there is a race between the virus spread and the therapy/vaccine available. In other words, you have to evaluate risk [versus] benefits of vaccinating a population during a pandemic."

Using an epidemiological susceptible-exposed-infected-recovered, or SEIR, model and a statistical analysis of publicly available data to estimate excess deaths resulting from suspending AstraZeneca vaccinations, as well as deaths potentially linked to deep vein thrombosis events, the study found that suspending AstraZeneca vaccinations in Italy and France for just three days — without replacing it with another vaccine — resulted in approximately 260 and 130 additional deaths, respectively. 

The SEIR model used in the study was previously used by Faranda in other research to successfully predict both the magnitude and timing of COVID-19's second wave in France and Italy.

"This model assumes that the population can be divided into four compartments: those who can catch the virus (susceptibles), those who have the virus but do not yet display symptoms (exposed), those who have symptoms (infected) and those who recovered from the virus," Faranda explained. "The recovered compartment [counts] not only those healed, but also the deaths from COVID-19 and the vaccinated ones. You can move from one compartment to another by some rate parameter, which depends on the spread of the virus, the time it takes to display symptoms and to heal from the virus."

Though the actual AstraZeneca vaccination pause lasted less than a week, the researchers explored what could have happened if it had gone on longer. Assuming that the interruptions in distribution are temporary, the model suggests that a 14-day delay in AstraZeneca vaccine delivery in Italy would have resulted in approximately 1,700 additional deaths from COVID-19; in France, that number for the same period would have been 430. A seven-day pause in distribution would have resulted in 790 COVID-19 casualties in Italy and 160 in France. 

In contrast, the model's worst-case prediction of deaths from deep vein thrombosis, a 100% mortality rate, in both countries amounts to fewer than 300 deaths. A standard 5% mortality-rate prediction of DVT-related deaths amounts to 13 deaths in Italy and 15 deaths in France.

The predicted difference between the two countries' number of COVID-19 deaths, according to the study, is due to their unique epidemiological situations and, in particular, to the higher basic reproduction number measured in Italy with respect to France, as of March 15.

The reproduction number quantifies the average number of unvaccinated people previously free of infection who will contract a contagious disease from one infected person. In Italy, the study's SEIR model is integrated for 500 days, with a reproduction number of 1.16 for Italy; France's reproduction number over the same period in the model is 1.02.

In other words, those who have not yet been vaccinated when a pause in distribution occurs can easily contaminate other individuals before vaccination resumes. Even in the event that countries promptly resume AstraZeneca vaccinations, the researchers demonstrate that the effect of prior interruption will be difficult to counterbalance and will require that governments double down on vaccine deployment in the days ahead. 

Increased distribution could, however, prove complicated. Some Europeans are refusing the AstraZeneca vaccine because of its link to rare blood clots, and AstraZeneca is currently facing legal action from the European Union, which cites delays in vaccine delivery by the pharmaceutical company and a shortage of available doses compared with numbers contractually outlined. However, the United States also said April 27 that it would share more than 60 million doses of the AstraZeneca vaccine with other countries around the world following a successful federal safety review.

In the meantime, the researchers hope their AstraZeneca model will inspire future studies that investigate similar scenarios for other vaccine formulations in order to better inform scientists' and governments' distributive decisions.

"Our goal was to provide a fast but reliable estimate of the risk [versus] benefit of interrupting a vaccination campaign in a pandemic context," Faranda said. "Our aim is now that other more careful and detailed studies take over to have a precise framework for this kind of epidemiological [problem]."

For example, he said, researchers can change the parameters of the model to reproduce the characteristics of any other COVID-19 vaccine, such as the similarly-beleaguered Johnson & Johnson vaccine, in any other country or situation in order to see the impact of temporarily stopping vaccinations.

The study, "Interrupting vaccination policies can greatly spread SARS-CoV-2 and enhance mortality from COVID-19 disease: The AstraZeneca case for France and Italy," published April 27 in Chaos, was authored by Davide Faranda, Laboratoire des Sciences du Climat et de l'Environnement, Université Paris-Saclay & IPSL, London Mathematical Laboratory, and Ecole Normale Superieure, PSL Research University; Tommaso Alberti, National Institute for Astrophysics-Institute for Space Astrophysics and Planetology (INAF-IAPS); Maxence Arutkin, UMR CNRS and ESPCI Paris; Valerio Limbo, National Research Council of Italy, Institute of Atmospheric Sciences and Climate (CNR-ISAC); and Valerio Lucarini, University of Reading.