Commentary: Making sense of evidence-based medicine for Astrazeneca COVID-19 jab

As of this writing, the Food and Drug Administration has suspended the use of AstraZeneca COVID-19 vaccination for people less than 60, citing the European Medicines Agency (EMA) advisory. The EMA’s decision was based on 18 deaths in a total of 86 cases, almost all after the first dose. The cases were reported in an EU database of 34 million from European countries, including the UK, where around 25 million people have had the vaccine. 

At the same time, the FDA allowed the use of Coronavac made by Sinovac, for the elderly. This would be a good time to discuss evidence-based medicine (EBM) which has now evolved into evidence-informed decision-making and even more recently in the din of COVID-19, science-based decision making just to hopefully make sense of this dizzying developments.

So, what is evidence-based medicine? Put simply, it is a combination of best evidence usually in a peer-reviewed medical journal such Lancet or the New England Journal of Medicine combined with patients’ values and local context.

In the hierarchy of EBM, the highest level of evidence is provided by Randomized Control Trials (RCTs) or better, a mathematical combination of RCTs called a meta-analysis. Lowest levels are case series and expert opinion which represent one opinion or the opinion of one group. Obviously in evidence-based analysis the aggregate of multiple RTCs has a greater validity than the opinion of one individual.

What is the relevance in the present pandemic? Take the latest Pulse Asia survey of 61% refusal to vaccinate oneself. Although this will almost change as cases and deaths rise and more people get vaccinated, there will be a recalcitrant 30-40% who will refuse.

My suspicion is many of the nay-sayers are actually just waiting for the others in the queue and eventually will follow when more vaccines come. In the course of numerous webinars and advocacy activities, we have found the most effective way to convince these same doubters to change their minds is by appealing to their love of family and the greater good and making them understand herd immunity. 

The case of AstraZeneca vaccine is a classic case of an excellent vaccine developed by the finest scientists in Oxford and AstraZeneca gone awry by a series of unfortunate events and circumstances. It also has striking similarities with what went wrong with Dengvaxia here.

AstraZeneca’s vaccine was dogged by reports of supply debates within EU governments prompting an increase in vaccine nationalism. Then came reports of extremely rare clotting issues in Europe causing individual medical authorities to suspend its use, delaying even further the roll-out of vaccines.

While the World Health Organization is still sticking to its guns, the damage was done and thus the Philippine FDA has followed suit. So, what happens now to the thousands of people vaccinated with the first dose of AstraZeneca? Limbo? What happens to the millions of orders for the AstraZeneca COVID-19 vaccine ordered by the private sector presumably for millions of young workers in the private sector?

Delaying second dose vaccination to 12 weeks seems optimal, in terms of efficacy, so we hope the FDA will resolve this dilemma soon. The risk of developing severe COVID-19 in non-elderly adults, however, maybe rising thanks to the UK and other variants. So, is the risk of half-protection and dying from COVID-19 higher than 
deadly clots by a massive factor?

EBM champions concepts like Number Needed to Treat (NNT) and its inverse Number Needed to Harm (NNH), how many cases would you need to treat to show benefit or harm and at what level is this acceptable or not acceptable.

The cases were reported in an EU database (34 million including the UK, where around 25 million people have had the vaccine). 18 deaths from clots of the total 34 million vaccinated is 0.00000053 risk or Attributable Risk Increase or ARI.

The NNH is the reciprocal (i.e. 1/ARI) since people understand whole numbers rather than fractions or percentages after all. The NNH for the AstraZeneca COVID-19 vaccine is 1.9 million, or you need to inoculate almost 2 million people with the vaccine to risk a single death.

In the Philippines, unvaccinated COVID-19 patients have a 1.8% chance of dying. So for simplification, its NNH is 1/0.018 which is 55. So how do we now compare risk of dying from the vaccine related to clots and risk of dying of COVID-19 without the vaccine?

You have 34,000 times higher risk of dying of COVID-19 compared to the risk of dying from clots. The DOH assured the vaccination pause may not last two weeks, so why make us wait and cause more anxiety when the data has been there for some time? The further blow to vaccine confidence cannot be over-estimated.

The FDA seemed not have waited for the vaccine experts panel or overturned the Health Technology Assessment Council opinion and decided without consulting the vaccine czar to suspend the AstraZeneca vaccine for non-elderly people. The over-cautious attitude maybe related to fears of being sued or litigated in our toxic atmosphere post Dengvaxia.

So how do we move forward? Well for one, include and expand the current Inter-Agency Task Force, to include clinical epidemiologists. Almost all the COVID-19 Task Forces worldwide are led by epidemiologists and doctors.

The Department of Health’s vaccine confidence campaigns should be mindful of the lessons from Dengvaxia rather than ignore or scapegoat it, especially given the current parallels. The Department of Justice should stop filing cases against DOH officials two-years or more after vaccination.

It is widely accepted that adverse events usually happen within an hour of vaccination, but not two years later. Officials need to appreciate that how they talk about or respond to issues connected to vaccines have an impact on vaccine confidence.

As the recent Al Jazeera special report highlighted, the response to Dengvaxia harmed vaccine confidence in the Philippines rather than helping it. A more proportionate response would have delivered better outcomes.

Start vaccinating influential people like politicians, celebrities and even media personalities. I did not agree that political leaders should not have been prioritized. They are key influencers, they convince the “masa” and the bandwagon effect will begin. Make the private sector an equal partner in vaccination rollouts. It will also be good to promote public-private partnerships.

The pandemic has overturned completely the way EBM works. It is in this context that our FDA allowed Coronavac/Sinovac for seniors based not on Phase Three published literature but on Phase One and Two immunological studies and reports of no severe side effects in a few countries like Indonesia. I believe this would not be considered as high-level evidence-based decision making.

Dr. Charles Yu is a practicing pulmonologist at the De La Salle University Medical Center and clinical epidemiologist and full Professor of Medicine at the De La Salle College of Medicine. He is the vice-chancellor for Research and has more than a 30-year-history of doing clinical research in TB, vaccines and other respiratory disorders. He is a multi-awarded researcher and was recently awarded by the Phil. College of Physicians (PCP) Exemplar for Outstanding Researcher this year as well as Most Outstanding Fellow of the Phil College of Chest Physicians (PCCP) and one of the Ten Outstanding Filipino Physicians in 2009 (JCI, PHIC, DOH, PCSO). He is an internationally respected health expert, and is a member of Asia-Pacific Respirology Research Committee, worked as WHO consultant on tuberculosis, and was past president of PCP and PCCP.