Claims and Consequences
This blog discusses misinformation - including deliberate disinformation - during the SARS-COV-2 pandemic. I won't link directly to anti-vaccine content to avoid adding search-engine credibility to material best left unfound.
I've written this because I was recently asked by a family member to review some claims around vaccination. The claims I read, although baseless, were sensational and manipulative, using words like "criminal", "killing people" and "holocaust". It turns out the short, sharp shock of anxiety and anger-generating lies will spread more successfully than comprehensive, moderate, boring truth. Consequently, anti-vaccine misinformation has developed an almost Tardis-like quality - in containing zero room for nuance but unlimited space for logical fallacy.
It transpires that one of the most common attacks falls within the bailiewick of this blog, being about mortality rates. The assertion is usually that vaccination is unnecessary for a virus with an X% mortality or recovery rate, and this is often made alongside exaggerations of the risk of adverse reactions from the jab itself. Of course, this kind of material appears in many guises, with a variety of values for X, depending on which data is being (mis-)read. However, three false propositions are generally discernable:
- Claim 1: Death is the only consequence we should vaccinate against,
- Claim 2: Vaccination is unneccessary for the average person, and,
- Claim 3: The dangers of vaccination are greater than the dangers from COVID-19 itself.
Of course, these short, simple propositions are riven with flaws.
With a fixation on mortality or recovery, Claim 1 ignores that vaccination has never been exclusively tied to fatal disease. Long-standing precedents exist for use against pathogens with a lower case fatality rate (CFR) than COVID-19. Consider poliovirus: although estimates vary, case fatality rates for COVID-19 dwarf the estimate of 0.1-0.5% for poliovirus infection, since polio attacks the nervous system in only a minority of cases. Despite this, the need for (and success of) mass vaccination for polio is an accepted fact of modern life. Placing survival to one side, mitigating another risk such as long-term disability was enough to justify vaccinating our populations, and it seems obvious that the same considerations apply for COVID-19. The appeal to selfishness inherent in much anti-vax material suggests those at low risk should remain unencumbered by public health responsibilities. A related heartlessness considers deaths involving comorbidities as not being "true" COVID-19 mortality, as though the lives of those with asthma, diabetes or high blood pressure literally don't count. Selfish or not, the erroneous idea that there existed pre-vaccination clarity on which individuals were and were not at serious risk runs us smack-dab into the next claim.
With respect to Claim 2, we immediately hit a problem of averages: who is this average person that needn't be vaccinated? By boiling down the complexity of COVID-19 mortality into a single mortality rate, the claim makes a sweeping generalisation, incorrectly passing off a general aggregate as appropriate for specific application. Reasonable precision would at least require:
- Distinguishing between case-fatality rate and infection-fatality rate,
- Describing the specifics of the populations, especially the biological sex and ages considered,
- Describing population density, prior infection and health characteristics of those populations.
Imprudently glossing over such considerations was prominent in even the pre-vaccine phase of the pandemic: advocates for adopting what they saw as a Swedish-style herd-immunity approach ignored that a "herd" with lower obesity rates and higher life expectancies (among many other differences) might not safely offer plug-and-play comparison. Since the target of any simplistic claim is unlikely to belong to the precise group that claim derives from (or may be an outlier in that group), such claims can dangerously mislead. The best protocol for applying general averages to your own specific case was first described in 1971 by Harry Callahan: do you feel lucky, punk? Many vaccine refusers, misled that they were not in fact gambling, will have lost in a game with the very highest stakes. Vaccine sceptics make a lot of noise about consequences without considering the potentially deadly outcomes of material they themselves publish.
Both Claim 1 and Claim 2 seem to simplify an unwieldy morass of facts into something easily digestible, but these over-simplifications omit complexities that are required for understanding and decision making. A model for COVID-19 risk suggesting infections proceed to one of two outcomes, namely death or pre-infection status-quo, is simply wrong. Many negative long-term consequences can arise in survivors of COVID-19, such as inflammatory clotting pathologies, multiple organ damage, measurable cognitive deficits, and higher rates of diabetes, including within children. Awareness of such issues grew as the pandemic proceeded, but history already told us to beware of post-pandemic consequences, such as the theorised cardiovascular impacts from 1918's influenza. A dangerous underpin to anti-vax thinking seems to be that viral infection is somehow more "natural" than vaccination, which seems akin to preferring a raging inferno over employing a chemical fire extinguisher. Barring those in rare medical categories, people weighing the long-term risks from a serious zoonotic virus like SARS-COV-2 against risks from a vaccine designed and tested for human safety should vastly prefer the jab. That we find ourselves powerless to stem persistent disinformation to the contrary is a poor starting point from which to face the inevitable next pathogen. Our world has proven itself too connected to believe another hundred years will pass before the next crisis.
Whilst the first two claims traded at least partially in false reassurance, Claim 3 is all about fear, and it exploits the simple truth that every medical treatment can bring unwanted side-effects. Vaccine side-effects haven't been ignored, despite what misinformation might say, since we have adverse event recording mechanisms such as from the NHS Yellow Card or US VAERS systems. Unfortunately this is where the post-hoc fallacy often arises. Specifically, this fallacy is embedded in the anti-vax presumption that anything happening after a vaccination, happened because of the vaccination. Ironically, it seems more likely that many common adverse events happen after vaccination because of the fear stoked by anti-vaccine misinformation, but we'll leave that to one side. Adverse event recording systems exist to monitor for conditions occurring at higher than expected baseline rates, conditions that might therefore be attributable to the treatment. The phrase "expected baseline rates" really matters here: conditions such as heart attack, nerve disorder, spontaneous miscarriage etc, already occur under everyday circumstances, so this data is studied for signals that a vaccine changes the rate of occurence. The very rare but highly serious Astrazeneca/CHADOX1 unusual clotting effect was found this way, so we know these monitoring systems work.
Unfortunately, the rightly public nature of such data also inspires much bad-faith commentary. Where a regulatory report reviewed miscarriage data and concluded "no safety signals [...] emerged from review of these cases", the anti-vax reporting I encountered distorted this as: "the FDA knew that this mRNA vaccine would kill unborn babies". In fact, a major study of 87000 pregnancies in Scotland found the precise opposite - pregnant women (perhaps made fearful by misinformation) were less than half as likely as non-pregnant women to be vaccinated, while 98% of critical care admissions and all baby deaths were associated with unvaccinated women. Social media might be defended as a victory for freedom of speech, but in the bombardment of concerned mothers-to-be with falsehoods it is surely a pyhrric one.
Since peer-reviewed scientific research is not broadly accessible to the general population, we historically relied on medical practitioners, professional and regulated media and government public-information mechanisms to keep the population informed. The rise of social media brought powerful tools to disseminate information, but the commercial drive to maximise eyeballs and advertising often appears to amplify the wrong content. Over its relatively short history, the social media industry has repeatedly failed the Spiderman test, wielding huge power with insufficient responsibility. As a consequence, the rise of anti-vaccine content around COVID-19 wasn't just predictable, but predicted. Our societies are experiencing a painful transition to the provision of public health information via private corporations. What existed before wasn't broken, but our social-media "fix" happened anyway, and so now we face the consequences.
References:
Stock, S.J., Carruthers, J., Calvert, C. et al. (2022) SARS-CoV-2 infection and COVID-19 vaccination rates in pregnant women in Scotland, Nature Medicine (2022) doi: 10.1038/s41591-021-01666-2
Haas, J.W., Bender F.L., Ballou, S. et al (2020) Frequency of Adverse Events in the Placebo Arms of COVID-19 Vaccine Trials. A Systematic Review and Meta-analysis, JAMA Network Open. 2022; 5(1). doi: doi:10.1001/jamanetworkopen.2021.43955
Wilson, S. L., Wiysonge, C. (2020) Social media and vaccine hesitancy, BMJ Global Health 2020; 5(10). doi: 10.1136/bmjgh-2020-004206
Berger, J., Milkman, K.L. (2010) Social Transmission, Emotion, and the Virality of Online Content, Marketing Science Institute Working Paper Series 2010 Report No. 10-114
Previous posts
Measles - a short history lesson
The UK Health Security Agency recently issued a press release, warning that too few children were vaccinated against measles. With the benefits of vaccination and a developed healthcare system, it is easy to forget that measles was often a fatal disease for young children. Table 1 shows just how deadly measles was at the start of last century without vaccination:
Here is the nowcast
Everyone is familiar with the idea of a forecast. You have data on a phenomenon up to the current time, and want to forecast the phenomenon at some point in the future. The most obvious example is the weather forecast, but forecasting is also required in pension and annuity work. For example, when calculating reserves for pension payments, some kind of projection is required for future mortality improvements.
Add new comment