Future (“long-term”) side effects from COVID vaccines are extremely unlikely

Concern over unknown, future side effects is by far the most common reason I hear people give for why they aren’t vaccinating against COVID. At a quick glance, that may seem reasonable, but when you start really looking into the science, it quickly becomes clear that there is simply no reason to suspect that there will be dangerous long-term consequences of these vaccines. Indeed, based on everything we know about the immune system, vaccines in general, and these vaccines specifically, it is extremely unlikely that they will cause unknown, serious, wide-spread side effects in the future, and the known risks from COVID far outweigh the hypothetical risks from the vaccines.

In this post, I’m going to carefully go over the science and logic that allows us to be so confident in the long-term safety of these vaccines, but before I do that, I want to briefly explain who my target audience is here, namely, the “vaccine hesitant.” I am refereeing to the people who usually would not consider themselves “anti-vaccers” and would usually vaccinate themselves and their children but have been swept up in the maelstrom of misinformation and fear about the new COVID vaccines. If you are someone who is truly seeking information and trying to think for yourself, then please hear me out and seriously consider the arguments and evidence that I am going to present. Do not give in to the baseless fearmongering that is rampaging through the internet and media.

To that end, I want to point out right at the start that this tactic of stirring up fear of future, unknown, long-term consequences is not new or unique to COVID vaccines. It is straight out of the traditional anti-vaccine playbook. It is something I was writing about long before COVID, and the argument is just as flawed now as it was then. So, if you are someone who eschews the title “anti-vaccer,” but are avoiding the COVID vaccines because of the arguments about unknown, long-term effects, please realize that these are not new arguments that arose out of legitimate concerns specifically about COVID vaccines. Rather, stoking fear of future unknowns is a standard (and flawed) anti-vaccer tactic that they have been using for decades and are now dressing up and presenting as if it is a novel concern for COVID vaccines. Do not be fooled by this tactic.

This post is necessarily long, because there’s a lot to talk about in order to cover this topic properly, but, again, I have written this for people who are truly trying to think for themselves, and are truly seeking information. So if that is you, please read this carefully in its entirety.

Because of the length of this post, I will summarize key points in bullets below, before elaborating on each of them.

Summary: TL;DR

  1. mRNA vaccines have been being studied for over a decade (including human trials).
  2. Current COVID vaccines have been extremely well studied, with sample sizes of hundreds of thousands of people, and studies have been compiled into large meta-analyses/systematic reviews. Thus, the short-term risks of the vaccines are extremely well-documented, and the benefits outweigh the risks. The only “unknown” is about long-term effects; however…
  3. No vaccine has ever caused the type of widespread, serious side effect years down the road that everyone is afraid of.
    1. Nearly all side effects occur shortly after vaccination (see #2).
    2. The only example of a sided effect that showed up months later appear within a year (whereas we’ve been using COVID vaccines for over a year) and was rare. The vaccine benefits still outweighed the risks.
  4. Vaccines rarely cause long-term (future) side effects because they use low doses over a short time.
    1. Vaccines simply train your immune system.
    2. Vaccines are quickly removed from the body.
    3. Most vaccine components were well-studied, and their safety is known.
    4. mRNA:
      1. mRNA does not alter your DNA.
      2. mRNA is very quickly broken down and removed.
      3. mRNA in vaccines cannot make your body produce entire viruses.
      4. You are constantly exposed to mRNA from viruses (e.g., from colds)
      5. If you catch COVID, your cells will use viral mRNA to make proteins just like they do from the vaccine, but…
        1. Your cells will make entire viruses, not just a single protein.
        2. You will be exposed to far higher levels of mRNA.
    5. Side effects from immune stimulation will usually happen right away and will usually be worse from actual infection with COVID.
  5. A demand for long-term studies is meaningless unless you can justify why a particular length of time is needed.
    1. No matter how long something has been studied, it is always technically possible that an effect won’t show up until slightly after the length of that study.
    2. This is true for all medications, foods, minerals, vitamins, etc., yet we don’t fear most of them.
    3. Therefore, you must provide actual evidence or reasoning to think that a futre side-effect is actually likely.
  6. Focusing on a highly-unlikely, unknown, hypothetical risk from the vaccine while downplaying the very real and serious risk from COVID is bad risk assessment.
  7. Fears over unknown long-term effects of the vaccines are baseless. The burden of proof is on anyone claiming that the vaccines are dangerous.

Not as new as you might think

Before we go into the details of the COVID vaccines, we need some background information to put them in context, and I think it is important to point out that these vaccine technologies are not as new as people are often led to believe. Sure, these exact vaccines were developed recently, but mRNA vaccines have been being developed and tested for years. Thus, the underlying technology is well-studied.

Let me direct you to a review paper published in 2018 (before COVID) titled, “mRNA vaccines — a new era in vaccinology” (Pardi et al. 2018). This review covers over a decade of research on mRNA vaccines, including safety and efficiency trials on mice (Fleeton 2001; Geall et al. 2012; Magini et al. 2016), ferrets (Brazzoli et al. 2015), pigs (Schnee et al. 2016), monkeys (Brito et al. 2014), and yes, even humans (Craenenbroeck et al. 2015; Bahl 2017; Alberer et al. 2017). As you’d expect in a rapidly growing field, even more studies were published following that review, (but prior to COVID). Feldman et al. (2019), for example tested mRNA influenza vaccines in over 200 people, including following them for a full year after the vaccines to assess safety and effectiveness. Similarly, studies like Alberer et al. (2017) followed patients for a year prior to publishing and continued to follow them after publication.

To be clear, those human trials were small trials; my point is simply that we were able to develop these COVID vaccines so quickly not by rushing, but rather by utilizing a robust body of research that had already been conducted. All of the information was there, waiting to be applied to something like COVID.

The way that people (including politicians and the media) are talking about these vaccines, you’d think that they represent totally uncharted territory. Reading the comments on my page, people are acting like we have almost no knowledge about them and are shooting in the dark, recklessly plowing into the unknown, but that’s simply not true. In reality, we knew a ton about mRNA vaccines before COVID, and that should really change your perspective on these vaccines.

It is so easy to give into fear of the unknown, particularly when you are so constantly bombarded with people’s concerns. I don’t blame anyone for that;’ it’s human nature, but it’s important that we use logic and facts to overcome our base fears, and if you step back and start to rationally look at the wealth of knowledge these vaccines were based on (including human trials spanning a year or more), that really should paint these vaccines in a different light and help to alleviate those fears.

Note that I only cited a small handful of the studies that had been conducted prior to COVID.


What we know: proximate (short-term) side effects

The crux of the concern over these vaccines is fear of the unknown, so before we can talk about the unknown, we need to be clear on what we do know, as well as clearly defining what we mean by “unknown, long-term effects.”

There are basically two categories of effects we need to talk about:

  • Proximate effects (short-term) = effects that first occur shortly after vaccination
  • Unknown future effects (long-term) = effects that do not show up for months or years after vaccination (Note 1)

It is important to make this distinction, because I often find that people meander back and forth between these two without having a clear understanding of what is actually known or how it is known. So let me try to be as clear as possible: we have an extremely robust understanding of proximate effects, and the fact that the vaccines are new is 100% irrelevant.

Proximate effects are fairly straightforward to test. First, scientists conduct phase 1–3 human trials using a randomized, placebo-controlled approach, where they follow thousands of patients for several weeks following vaccination. Then, once the vaccine is released to the general public, scientists continue to monitor it for side effects using things like large cohort studies and case-controlled studies. As the sample sizes increase, so does our ability to detect increasingly rare events. With tens of thousands of participants, we can detect events that occur every few thousand people, but we will miss events that happen once for every 10,000 people. At a few hundred thousand people, we can detect events that occur once per tens of thousands of people, but will miss events that happen once for every 100,000 people, etc. (Note: numbers are approximations).

There are two critical points here. First, because our ability to detect rare side effects is dependent on sample size, as the sample size increases, any new side effects will, by definition, be increasingly rare. By the time we are into the millions (as we are with COVID vaccines) we aren’t going to suddenly find a new common serious side effect, because those would have been picked up at much smaller sample sizes.

Second, the novelty of the vaccines is completely and totally irrelevant. Because we are talking about events that happen within a few weeks of being vaccinated, it does not matter if the vaccines have been available for two months or two hundred years. The only thing that matters is the sample size (i.e., number of participants). Let me say that again (in bold), our ability to confidently know the rates of proximate side effects depends entirely on the sample size; the age of the vaccine is 100% irrelevant.

In the case of COVID, we were able to get these sample sizes extremely quickly because there were so many cases of COVID and governments dumped so much money into massive vaccine campaigns. All of the currently recommended vaccines passed their initial phase 3 trials with large sample sizes. For example, Pfizer used over 43,000 participants (Polack et al. 2020), and Moderna used over 30,000 (Mahase 2020).

Following those phase 3 trials, numerous large studies have been released. Indeed, so many studies have been conducted that we can do systematic reviews and meta-analyses. As explained here, these combine the data from multiple studies to look for overarching effects and are the highest level of scientific evidence. Qianhui et al. (2021), for example, included 87 different safety studies, and concluded that, “Available evidence indicates that eligible COVID-19 vaccines have an acceptable short-term safety profile.”

Yet more studies have been conducted since that review/meta-analysis, and some of them are truly massive. Barda et al. (2021), for example, compared over 800,000 vaccinated individuals to over 800,000 unvaccinated individuals to look at the rates of adverse events from the Pfizer vaccine as well as the rates of those same events in people who develop COVID. Not only did the vaccine have low rates of serious side effects, but, for most conditions (including myocarditis and myocardial infarction), the rates of those events were higher in people who caught COVID than in people who received vaccines (Note 2).

Figure 4 from Barda et al. 2021 showing the risk of adverse events from the Pfizer vaccine and from COVID itself. For each side effect, risk was calculated by by matching over 100,000 people who had been infected with COVID with people who were not infected with COVID (for COVID side effect rates), and matching several hundred thousand people who had received the vaccine with people who had not received it (for the vaccine rates). In most cases, COVID infection itself carries more risk of these specific side effects than does the vaccine. The most obvious exception (lymphadenopathy) is merely a swelling of the lymph nodes, which is not generally a serious condition.

Figure 4 from Barda et al. 2021 showing the risk of adverse events from the Pfizer vaccine and from COVID itself. For each side effect, risk was calculated by by matching over 100,000 people who had been infected with COVID with people who were not infected with COVID (for COVID side effect rates), and matching several hundred thousand people who had received the vaccine with people who had not received it (for the vaccine rates). In most cases, COVID infection itself carries more risk of these specific side effects than does the vaccine. The most obvious exception (lymphadenopathy) is merely a swelling of the lymph nodes, which is not generally a serious condition.

Other calculations of the rates of specific adverse events have had even large sample sizes. For example, the Israel Ministry of Health used over 5 million people to calculate the rate of myocarditis following vaccination. Similarly, in the USA, the CDC has several hundred million vaccine doses to use in its calculations.

The point is that we are extremely confident about the short-term consequences of the vaccines. It’s hard to overstate the massive volume of data we have. Barda et al. (2021), for example, is one of largest cohort studies I have ever read. It is larger than most studies on the safety of well-established vaccines that have been available for decades. Indeed, we have been able to quickly collect so much data that our knowledge of the short-term safety of COVID vaccines is equal to or greater than our knowledge of the short-term safety of many standard vaccines.

Again, to be 100% clear, the fact that the vaccines are relatively new simply does not matter for these short-term effects. Further, these studies aren’t the result of “rushing.” Rather, it is simply matter of vaccinating so many people so quickly that we were able to rapidly collect the data that would usually take years to accumulate. It is the size and volume of the studies that matters, and we have numerous truly massive studies unequivocally showing that serious side effects are rare and the benefits outweigh the risks.

To put it simply, the short-term side effects of the COVID vaccines have been thoroughly studied and are extremely well-documented. Scientifically, these vaccines are no longer experimental (with the exception of their application to young children, in some cases). They have already passed numerous experiments and the evidence is clear (Pfizer isn’t “experimental” legally either). Insisting that we haven’t studied the vaccines well-enough to know the short-term side effects is, at this stage, science denial.

See Note 3 regarding the supposed vaccine-related deaths and injuries reported in VAERS.


Vaccines don’t cause wide-spread, long-term adverse events

Now we can finally turn our attention specifically to the topic of unknown, long-term effects (which, remember, are effects that do not show up for months or years after vaccination; Note 1). I realize I took a long time getting here, but that background was important, because I have shown that we have a massive body of studies showing that the COVID vaccines have few serious side effects shortly after receiving them. Thus, the only way to doubt their safety without outright science denial is to raise concerns over presently unknown, long-term effects, but, as I will show, those concerns have no scientific merit.

The type of future long-term consequence that everyone seems so afraid of (i.e., the type that only manifests months or years down the road) is virtually unheard of from vaccines. I looked long and hard for examples of this occurring, and in the entire history of vaccines, I was only able to find one: Pandemrix, an H1N1 vaccine used in Europe for the 2009–2010 flu season was associated with an increased risk of narcolepsy that usually only manifested weeks or months after the vaccine. You can read more details on Thoughtscapism and Skeptical Raptor, but there are just three points I want to make.

  1. Depending on the study, the lag between vaccination and onset of narcolepsy was 0-242 days (median = 42; Partinen et al. 2012) or 0-360 days (median not reported; Nohynek et al. 2012). Pfizer and Moderna both began their phase 3 COVID vaccine trials on 27 July 2020 (~400 days ago) and received emergency use authorization (thus starting mass vaccination campaigns in the USA) in December 2020 (~260 days ago). Indeed, Israel had already administered over 1 million doses by the end of 2020. This means we are already past the time frame where we should have started picking up something comparable to the long-term effects of Pandemrix.
  2. As is so often the case with vaccine side effects, the disease they prevent (influenza in this case) also causes the same side effect.
  3. This side effects was rare (between 1 in 52,000 doses and 1 in 57,500 doses in England [Miller et al. 2013] and 1 in 16,000 in Finland [Nohynek et al. 2012; for unclear reasons, Finland had a high rate that could not be generalized to other countries), and the benefits of the vaccine still outweighed the risk.

That last point is really important, because for it to turn out that avoiding the COVID vaccines was the safer choice, unknown future side effects would not only have to exist, but they would have to be so common and so serious that they outweigh the enormous known benefits of the vaccines, and that is a situation that has never occurred for any vaccine (Note 4). For reasons that I’ll explain in the next sections, that’s simply not how vaccines interact with the body.

So, if you are avoiding the COVID vaccines because of a fear of unknown, serious, long-term side effects, ask yourself, is that fear really rational given that future, long-term side effects of vaccines are virtually unheard of, and there has never been a case where those side effects were widespread and outweighed the benefits of the vaccines?


Why vaccines don’t cause future long-term effects: Low dose, short exposure

Let’s now talk about why vaccine side effects nearly always show up shortly after vaccination. The type of long-term consequence we are talking about typically comes from one of two causes: a very large dose over a short time, or a small dose over a prolonged period of time. Vaccines don’t fit either of those categories. They are fundamentally different from most medications because they simply train your immune system before being quickly removed. Your own immune system is what provides a lasting benefit. Further, vaccines do this via low, non-toxic doses. Remember, the dose makes the poison. Everything, even water (Garigan and Ristedt 1999), is toxic at a high enough dose and safe at a low enough dose. So people who scream about “TOXIC CHEMICALS” in vaccines are ignoring basic chemistry. There is no such thing as a toxic chemical, there are only toxic doses, and the doses in vaccines are not toxic.

One of the most common arguments I hear people making to justify concerns over COVID vaccines is, “look at all the examples of drugs that were approved, then years later long-term effects were found.” Those examples are, however, nearly always for drugs that were taken repeatedly. It’s the cumulative effect that causes the risk (particularly for chemicals that persist in your body for long periods of time). Vaccines, in contrast, have limited exposure, and your body quickly eliminates them. Within a few days of receiving the vaccine, the vaccine itself has been totally eliminated from your body. The long-term protection comes from immune system memory, not from the vaccines themselves.

This is really important, because it means we don’t have a mechanism through which COVID vaccines would cause long-term harm. Because vaccines are a low doses given 2-3 times, we expect any consequences to happen quickly, which is exactly what we find. The most common side effects are things like soreness and moderate flu symptoms that start within a few hours or days of receiving the vaccine. These effects aren’t because the vaccine is “toxic” but rather because it is doing exactly what it was designed to do and stimulating your immune system. It’s that activation of your immune system that makes you feel unwell, but that activation is critical, because it is how your immune system learns to identify and fight COVID. Similarly, serious side effects from the vaccines, while rare, usually show up shortly after vaccination.

Side effects that don’t show up for months or years simply aren’t expected from vaccines because of how vaccines work. Nevertheless, in the following sections, let’s look more closely at the three main hypothetical sources of long-term harm: adjuvants/preservatives, mRNA, and immune activation.


#1: Adjuvants and preservatives

Vaccines typical consist of three basic components: a representation of the infectious agent (antigens, weakened viruses, virus particles, mRNA, etc.), an adjuvant that simulates the immune system and/or aids in delivery of the antigen, mRNA, etc., and preservatives (usually salts, metals, and sugars) to avoid contamination and stabilize the other components.

The later two categories (adjuvants and preservatives) are historically the things that anti-vaccers have targeted (e.g., the infamous, and completely false, accusation that thimerosal [ethyl-mercury] caused autism). These accusations have, however, never stood up to scrutiny. Vaccine components have been well-studied and are safe at the doses used in vaccines.

Specifically for COVID vaccines, their components differ from one vaccine to the next, but the safety of the components is well-known. Many of them use standard salts/metals that have been used in numerous previous vaccines and medications, and the non-mRNA vaccines usually use the adjuvants that have already been used in other vaccines.

Specifically for the mRNA vaccines, they use a different type of antigen known as a “lipid nanoparticle” (basically a small, fancy fat) that stimulates the immune system and serves as a delivery mechanism for the mRNA. These are new for a commercially available vaccines (because we’ve never had commercially available mRNA vaccines before), but that doesn’t make the nanoparticles themselves new, and there is a wealth of studies on them (including studies on other vaccines that have been being developed [see previous section on the history of mRNA vaccines]). See Hou et al. (2021) for an extensive review of the topic.

My point is simply that while the vaccines are “new,” their components have been well-studied, and there is simply no reason to think that they pose a long-term danger.


#2: mRNA

Now let’s turn our attention to the big one that has so many people worried: mRNA. At the outset, we need to be clear on what mRNA is and which it does. Your cells contain DNA stored in the nucleus. This provides the plans for your body and how it runs, and it is what you pass on to make your offspring when you procreate. For the actual day-to-day running of your body, however, it has to be transcribed into mRNA (aka “messenger RNA”). This is a single stranded copy of your double-stranded DNA. The mRNA can then leave the nucleus and go to the ribosomes (little protein factories in your cells) which translate the mRNA into amino acids which are then strung together and folded to form proteins. This is happing millions of times in your body each second. Importantly, the process does not alter your DNA. Your genetic code is unaffected. Think of it like taking a master copy of a recipe, photocopying it, then giving that photocopy to someone who then follows the instructions on it.

Viruses are actually pretty neat and replicate by tapping into this system. They can’t reproduce on their own. Instead, they insert their DNA or RNA into your cells and hijack your molecular machinery by making the ribosomes translate their RNA and build new virus (some viruses have DNA and require a transcription step, others [like COVID] store their genetic material as RNA).

The mRNA vaccines tap into this same process. They include a small fragment of the RNA from the SARS-CoV-2 virus (specifically for the spike protein), thus causing your cells to produce that spike protein. Your immune system is then stimulated to attack the spike protein, and in the process, it learns to attack the actual SARS-CoV-2 virus. Take a minute to stop and marvel at the ingenuity of this system, because it’s incredible.

There are several important points that need to be made here:

  1. This process does not alter your DNA. The viral mRNA does not get integrated into your DNA. This is not gene therapy. All that happens is protein production by ribosomes. Again, this is like handing your cells a photocopy of a set of instructions.
  2. mRNA is a very fragile, short-lived molecule. During my PhD, I worked in a laboratory where some people do RNA research, and they often joked that if you looked at the vials the wrong way the RNA would vanish. The point is that the mRNA from the vaccines very quickly breaks down and is removed from your body. Within a few days of receiving the vaccine, it is totally gone.
  3. The vaccines only contain the mRNA for a single protein. It is impossible for them to cause your body to make the full virus. They simply don’t contain that information.
  4. This is a process that is already happening constantly in your body. Right now, you almost certainly have some viruses (even if you are healthy), and those viruses are hijacking your cells with their RNA and forcing your cells to make virus for them. Indeed, unlike with the vaccine, they are making your body produce entire viruses, not just a single protein. Similarly, anytime you become infected with a cold, the flu, etc., your body is exposed to tons of viral RNA which it then translates into proteins (entire viruses)
  5. (related to #4) If you become infected with COVID, this process is going to happen anyway, but unlike with the vaccine, your cells are going to produce the entire virus, and, because the virus will be replicating, you will be exposed to substantially more viral RNA for a longer period of time.

That last point is incredibly important, because it means that any fears you have about the mRNA in the vaccine should be even greater for the virus itself. It doesn’t make any sense to simultaneously downplay the seriousness of COVID while fearing the mRNA in the vaccines, because if you catch COVID, you are going to be exposed to substantially higher doses of viral RNA!

As you can hopefully see, none of this lends credence to the idea that the vaccine will cause long-term effects. There is simply no mechanism through which the mRNA could cause long-term harm, and even if there was a concern over long-term effects, that concern would be even higher from actually catching COVID!


#3. Immune activation

The final way in which vaccines could, in concept, cause harm is as a side effect of the inflammatory immune response they stimulate. Indeed, that is the cause of most vaccine side effects. The vaccine sets off a cascade of immune responses, and sometimes your body gets caught in the crossfire, though this rarely causes serious problems.

Importantly, however, this happens while your immune system is being stimulated. This isn’t a pathway that we would expect to not cause any noticeable problems shortly after vaccination, then suddenly cause massive problems down the road. It could, in concept, cause a problem that starts shortly after vaccination and persists long-term, but it’s unlikely to cause problems that don’t appear until months or years later.

This is important because, again, problems that arise shortly after vaccination and persist aren’t what we are talking about. Those aren’t unknown. Rather, we already know that those are rare because we can detect them shortly after vaccination (see previous section on short-term studies).

Finally, as I’ve alluded to several times already, problems that arise as a result of immune activation should also arise as the result of actual infection with SARS-CoV-2, and they’d usually be expected to be worse or more common from an actual infection. Indeed, that’s exactly what Barda et al. (2021) found.

So, once again, it makes no sense to fear this as a consequence of the vaccine while downplaying the seriousness of COVID, because infection with COVID is more likely to cause this problem (see Note 2 on absolute risk).


How long is long enough?

This is an issue that I’ve written about several times before (e.g., here and here), but in short, the demand for long-term data becomes extremely problematic unless “long-term” is carefully defined and justified beforehand. We already have over a year of data on COVID vaccines, plus many years of data on mRNA vaccines more generally. For most scientists, based on everything we know, that is plenty long enough to be confident in the safety of these vaccines, but if you are going to claim that it is not long enough, the questions become “why?” and “how long is long enough?”

As I said earlier, anti-vacces have used this argument against vaccines for ages, and the problem is that they constantly shift the goal posts. If you show them a 3-year study, they say, “well maybe effects don’t show up until 5 years.” If you show them a 5-year study, they say “well maybe effects don’t show up until 10 years.” If you show them a 10-year study, they switch to 15 years, 20 years, etc. They can keep extending it all the way until the end of the human life-span, and beyond the fact that continually shifting the goal posts is an ad hoc fallacy (and this whole thing is an argument from ignorance fallacy), demanding 15 years of data is only slightly more irrational than demanding 10 years, or even 5 years or 3 years.

Really think about this. Given that no vaccine has ever had a wide-spread, serious side effect that only shows up more than a year after vaccination, what is the justification for demanding 3 years of data instead of accepting the year+ of data we have? How is the demand for 3 years of data more logical than a demand for 10 years, or 20 years, or 60 years? All of those are time categories where we’ve never seen a vaccine suddenly cause new problems and for which we have zero reason to expect these vaccines to cause problems. The probability of a long-term effect only showing up over a year after vaccination is pretty close to zero, which means that it is close to zero for 3 years, 5 years, etc.


Bad risk assessment

As I’ve shown throughout this post, there is simply no good evidence to suggest that the COVID vaccines will have serious long-term consequences that only show up in the future. It’s a baseless fear. Meanwhile, we know that COVID itself is very serious. In the USA alone, it has killed over 650,000 people. In 2020, it was the third leading cause of death in the USA, and in early 2021, it briefly spiked to the #1 slot before dropping back to position #3. We should not be downplaying something that is so prevalent and deadly that it is the third leading cause of death. Further, beyond death, many people suffer serious complications from COVID, some of which will likely persist into the future (Mitrani et al. 2020; Fraser 2020).

Therefore, based on everything we know (which is a lot), risk assessment clearly shows that you are safer with the vaccine than without it, and while it is technically possible that there will be future unknown consequences of the vaccine, these would be even more likely from COVID itself, and it is incredibly unlikely that they will happen from the vaccines and be serious and widespread enough to alter the risk assessment.

By avoiding the vaccine, you are placing more weight on an unknown and unlikely hypothetical future risk than you are placing on a very real and serious known risk.

See the following posts for more details including the “99% survive” argument, precautionary principle argument, and COVID comorbidities


Long-term fears are baseless: The burden of proof

As I’ve explained throughout, there is not one shred of evidence nor a single logical argument that makes it likely that the vaccines will have unknown long-term consequences. This is a completely made-up concern. This isn’t a situation where we have preliminary data suggesting a concern, or a logical/scientific basis for thinking that there is a risk. Rather, this is a concern that was pulled out of thin air with absolutely no evidence behind it.

Further, to be clear, the fact that something is new does not make it likely that there are unknown long-term effects. Indeed, everything we know about vaccines and the immune system makes it extremely unlikely that there will be future, unknown, wide-spread, long-term consequences. Is it technically possible? Sure, but there are an infinite number of technically possible things that will probably never happen. “Technically possible” is not a valid justification for a fear, particularly if that fear will prevent you from taking a medication that greatly lowers your risk of disease and death.

For future unknown consequences to be a logically valid reason for not vaccinating, the probability of serious consequences occurring would need to be high enough to trump the massive known benefits of the vaccines. We would need some really compelling preliminary evidence to suggest that these future injuries will occur, and we simply don’t have it, not one scrap of it.

I say again, this is a made-up concern. Although it makes a certain amount of sense from the standpoint of the psychology of our panicky primate brains, it is a concern that is not based on any evidence or logic. You can’t just make up a concern, then demand action based on that concern. You need actual evidence to support the concern.

In medicine (and science more generally), it is not enough to simply say that something is technically possible. Rather, you have to show that there is a reasonable probability of it being true before it makes sense to treat it seriously (this is something known as the “prior probability”).

Imagine, for example, that I decide that taking aspirin while drinking soda is dangerous, and when asked to justify that fear, I simply say, “well we don’t know that it isn’t dangerous. It’s technically possible that it’s dangerous, and look at how many drugs have been recalled because of some complication with another chemical.”

I think that we can all agree that my fear would be irrational, right? In technical terms, it would be an argument from ignorance fallacy. The fact that something is unknown, doesn’t mean that I can act as if that thing is known to be dangerous. There are an infinite number of things that are unknown. There are an infinite number of potential interactions and long-term effects for all treatments, including vitamins, supplements, herbs, etc.  There haven’t been, for example, any 30-year studies on the effects of regularly taking most vitamins or supplements, and given that those are taken daily, they are far more likely to cause long-term issues. So why not be concerned about them?

Do you see the point that I am getting at here? The fact that we haven’t looked at 3-year effects of the vaccines (or 5 years, or 10 years, etc.) would only matter if we actually had evidence to suggest that there would be problems down the road, and we don’t have that evidence. Indeed, all of the evidence suggests the opposite. Therefore, this is a baseless fear and the burden of proof is on those who are avoiding the vaccine based on these concerns.

Now you could try to quibble with me and say that, “No one is saying that there definitely are long-term effects. We are just saying that we don’t know if there are and, therefore, we should not take the vaccine until we do know.” But, again, that doesn’t work for all the reasons that I’ve laid out. A lack of knowledge simply isn’t sufficient in and of itself. This is an abuse of the precautionary principle, and although you may not be claiming that there are, in fact, long term effects, by choosing to avoid the vaccines, you are, nevertheless, acting as if there will be those effects. As explained earlier, that’s bad risk assessment.

I want to conclude this with some questions. If you are not vaccinating because of concerns over unknown long-term effects, ask yourself, “why do I have those concerns?” Can you point to any actual data to justify them, or is it simply a fear of the unknown? If the latter, ask yourself how likely it is that those fears will come true. The fact that something is new or unknown doesn’t make it dangerous. Given the very real risk of COVID, the decade+ of research on mRNA vaccines, the decades of research on vaccines in general, the massive studies on the COVID vaccines, and the fact that no vaccine has ever had the type of serious, widespread, unknown, long-term side effect that everyone is so afraid of, does it really make rational sense to avoid the vaccines out of fear of the unknown? Does it really seem more likely that you will be injured by this totally hypothetical and unprecedented vaccine injury than by a virus that is currently the 3rd leading cause of death in the USA?



Note 1: When we talk about unknown long-term effects, we are not talking about adverse events that happen shortly after vaccination and continue to cause problems into the future (those are proximate events that have long-term consequences). We aren’t talking about something like myocarditis which, in rare cases, occurs shortly after vaccination and (in a small subset of the most extreme cases) can cause long-term damage. We already know that those events are extremely rare, because we’ve already been able to detect them. They aren’t unknown. In other words, because those events are first detected shortly after vaccination, we have been able to test them with the current short-term studies and have shown that they are extremely rare.

Note 2: Barda et al. (2021) was comparing rates among the vaccinated with rates among the infected, not absolute risk. Absolute risk depends on how likely you are to become infected. However, other analyses (e.g., Gargano et al. 2021) have shown that in high-risk countries like the USA, your absolute risk of serious injury and death is lower with the vaccine than without it, even if you are in a low-risk COVID group.

Note 3: There are many false claims floating around about thousands of deaths following vaccination. These claims are based on VAERS which includes anything observed following vaccination and does not establish causation. With millions of people receiving vaccines, it is inevitable that a few thousand will die shortly afterwards just by chance. In the vast majority of cases, there is simply no reason to think that the vaccines were responsible. Similarly, while some of the adverse events reported in VAERS may have been caused by vaccines, most probably weren’t. The database is self-reported (anyone can make entries), and some truly wacky submissions have been included. Further, again, the fact that something happened after vaccination absolutely does not mean that the vaccine caused it (that’s a post hoc ergo propter hoc fallacy; more details here and here). To quote the CDC “FDA requires healthcare providers to report any death after COVID-19 vaccination to VAERS, even if it’s unclear whether the vaccine was the cause. Reports of adverse events to VAERS following vaccination, including deaths, do not necessarily mean that a vaccine caused a health problem. A review of available clinical information, including death certificates, autopsy, and medical records, has not established a causal link to COVID-19 vaccines” (the bold was in the original). More details on VAERS here.

Note 4: Again, to be 100% clear, we are talking about injuries that won’t show up until later down the road. You certainly can find examples from decades ago where there were issues with a vaccine rollout (particularly concerning polio vaccines), but those issues were immediate, and that’s not what we are talking about here. The COVID vaccines all underwent massive randomized controlled trials and have been carefully monitored following release to the public, and with the hundreds of millions of doses that we have administered, we have a very clear picture of the immediate risks and benefits. Those aren’t unknowns.


Related posts

Literature cited

  • Alberer et al. 2017. Safety and immunogenicity of a mRNA rabies vaccine in healthy adults: an open-label, non-randomised, prospective, first-in-human phase 1 clinical trial. Lancet 390:1511–1520.
  • Bahl 2017. Preclinical and Clinical Demonstration of Immunogenicity by mRNA Vaccines against H10N8 and H7N9 Influenza Viruses. Molecular Therapy 25:1316–1327.
  • Barda et al. 2021. Safety of the BNT162b2 mRNA Covid-19 Vaccine in a Nationwide Setting. The New England Journal of Medicine
  • Brazzoli et al. 2015. Induction of Broad-Based Immunity and Protective Efficacy by Self-amplifying mRNA Vaccines Encoding Influenza Virus Hemagglutinin. Journal of Virology 90.
  • Brito et al. 2014. A cationic nanoemulsion for the delivery of next-generation RNA vaccines. Molecular Therapy 22:2118–2129.
  • Craenenbroeck et al. 2015. Induction of cytomegalovirus-specific T cell responses in healthy volunteers and allogeneic stem cell recipients using vaccination with messenger RNA-transfected dendritic cells. Transplantation 99:120–127.
  • Feldman et al. 2019. mRNA vaccines against H10N8 and H7N9 influenza viruses of pandemic potential are immunogenic and well tolerated in healthy adults in phase 1 randomized clinical trials. Vaccine 37:3326–3334.
  • Fleeton 2001. Self-replicative RNA vaccines elicit protection against influenza A virus, respiratory syncytial virus, and a tickborne encephalitis virus. Journal of Infectious Diseases 183:1395–1398.
  • Fraser 2020. Long term respiratory complications of covid-19. BMJ 370.
  • Gargano et al. 2021. Use of mRNA COVID-19 Vaccine After Reports of Myocarditis Among Vaccine Recipients: Update from the Advisory Committee on Immunization Practices — United States, June 2021. CDC 70:977–982.
  • Garigan and Ristedt 1999. Death from hyponatremia as a result of acute water intoxication in an Army basic trainee. Military Medicine 164:234–238.
  • Geall et al. 2012. Nonviral delivery of self-amplifying RNA vaccines. Proceedings of the National Academy of Science 109:14604–14609.
  • Hou et al. 2021. Lipid nanoparticles for mRNA delivery. Nature Reviews Materials
  • Israel Ministry of Health. 2-June-2021. Surveillance of Myocarditis (Inflammation of the Heart Muscle) Cases Between December 2020 and May 2021 (Including). Accessed 27-8-21.
  • Magini et al. 2016. Self-Amplifying mRNA Vaccines Expressing Multiple Conserved Influenza Antigens Confer Protection against Homologous and Heterosubtypic Viral Challenge. PLoS ONE 11: e0161193.
  • Mahase 2020. Covid-19: Moderna vaccine is nearly 95% effective, trial involving high risk and elderly people shows. BMJ 371.
  • Miller et al. 2013. Risk of narcolepsy in children and young people receiving AS03 adjuvanted pandemic A/H1N1 2009 influenza vaccine: retrospective analysis. BMJ 26.
  • Mitrani et al. 2020. COVID19 cardiac injury: Implications for long-term surveillance and outcomes in survivors. Heart Rhythm 17:1984–1990
  • Nohynek et al. 2012. AS03 Adjuvanted AH1N1 Vaccine Associated with an Abrupt Increase in the Incidence of Childhood Narcolepsy in Finland. PLoS ONE 7: e33536
  • Pardi et al. 2018. mRNA vaccines — a new era in vaccinology. Nature Reviews Drug Discovery 17:261–279
  • Partinen et al. 2012. Increased Incidence and Clinical Picture of Childhood Narcolepsy following the 2009 H1N1 Pandemic Vaccination Campaign in Finland. PLoS ONE 7:e33723
  • Polack et al. 2020. Safety and efficacy of the BNT162b2 mRNA Covid-19 Vaccine. New England Journal of Medicine 383:2603–2615.
  • Schnee et al. 2016. An mRNA vaccine encoding rabies virus glycoprotein induces protection against lethal infection in mice and correlates of protection in adult and newborn pigs. PLoS Neglected Tropical Diseases 10:e0004746.
  • Qianhui et al. 2021. Evaluation of the safety profile of COVID-19 vaccines: a rapid review. BMC Medicine 19:173.

Posted in Vaccines/Alternative Medicine | Tagged , , , , | Comments Off on Future (“long-term”) side effects from COVID vaccines are extremely unlikely

The “99% survive COVID” argument is deceptive and completely misses the point

covid vaccineThroughout the COVID-19 pandemic, the argument that “99% of COVID patients survive” has been repeatedly used as an excuse to oppose public health measures including mask mandates, lockdowns, and, more recently, vaccines. It asserts that because COVID has a ~99% survivorship, is not really that big of a concern and vaccines, masks, lockdowns, etc. are not needed.

There are many issues with this argument, but before I talk specifically about COVID-19, I want to discuss a hypothetical disease. Imagine that the majority of people who become infected with this hypothetical disease are fine. Roughly 70% of people are asymptomatic. Really serious consequences that could alter the rest of the victim’s life only occur in about 0.5% of cases, and over 99.5% of patients make a full recovery. Actual mortality is even lower. The exact number varies, but even on the high end, over 99.9% people survive.

Now, when I present this hypothetical disease like that, it probably doesn’t sound that bad. Simply giving you those numbers without any additional context makes the disease sound pretty benign, and you might be tempted to use this “>99% live” argument and argue that we don’t really need a massive vaccine push and should just wait for natural herd immunity.

Here’s the thing though. I lied when I said this was a hypothetical disease. Those are the stats for polio (CDC Pinkbook), a notoriously horrible disease that was once a scourge on our planet causing thousands of cases of paralysis and even death. It is a disease that, thanks to vaccines, we very rightly sent into oblivion in nearly every part of the world. If, however, instead of using vaccines, we had listened to this “99% live” argument, we would still have polio in countries like the USA, and children would still be getting infected, paralyzed, and even killed by this disease. Ninety-nine percent survival is not as high as it sounds, nor is it the only factor that has to be considered, but it can sound very persuasive when it is presented without the other necessary pieces of information.

Please, keep this in mind as we now turn our attention to COVID. If you are tempted to dismiss the severity of COVID or the necessity of vaccines because COVID “only” kills 1% of patients, keep in mind that polio “only” paralyzed 0.5% of its victims and killed less than 0.1%.

Note: The actual COVID case fatality rate is affected by many factors (e.g., age). As such, it is far more complicated than a simple “99%,” and you should not confuse that number with the notion that you personally have a 99% chance of survival. Nevertheless, 99% is the number I keep seeing people use, and it is a rough approximation of an average for many parts of the world, so for simplicity, I will continue using it here.

Infection prevalence matters

One of the key problems with the “99% survive” argument is that it completely ignores the importance of transmissibility and infection prevalence (i.e., how easily a disease can spread). You see, a disease can be a serious threat to a population (in terms of mortality) either by having a high case fatality rate (i.e., killing a large percentage of patients) or by being highly contagious (i.e., infecting a large number of people), and case fatality rates can be very misleading by themselves.

Consider, for example, two hypothetical diseases (A and B). A has an extremely high case fatality rate of 100%, but it is not very contagious and has a very low prevalence in the population. Only about 1 in 1 million people will catch it.  In contrast, disease B has a much lower case fatality rate (1%), but is extremely contagious, and about 1 in 10 people will catch it.

This means that in a population of 10 million people, disease A will kill 10 people, whereas disease B will kill 10 thousand people! So, which of those diseases should you be more concerned about? Obviously disease B is more concerning, right? You are at a higher risk of dying form disease B, even though 99% of patients survive. Why? Because it is so common.

If you admit this basic fact, that disease B would be more concerning than disease A, then you have just admitted that case fatality rate is not sufficient in and of itself to tell you how concerning a disease is. If we had relied only on case fatalities, we would have incorrectly concluded that disease A was the bigger risk. This clearly demonstrates that the “99% survive” argument is not a good argument, because we have to look at other factors besides just survivorship (the opposite of case fatalities), and even a disease with a 99% survival rate can be very dangerous to both populations at large and you personally if it is highly contagious.

To put it simply, when considering risk, you need to consider both how likely you are to be infected, and how likely you are to die (or suffer serious consequences) if you are infected.

In the case of COVID19, a large part of why it is so dangerous is its high rate of transmission, with new variants like the Delta strain making the situation even worse. That is why it has been able to kill over 4 million people world-wide, with over 600,000 deaths in the US alone.

Those are sobering numbers, which should really make you stop and think long and hard about the veracity of the “99% survival” argument. To me, it seems self-evident that a disease that has already killed over 600,000 people in the US alone is a serious problem that merits something like a massive vaccination campaign, and the argument that the disease is not dangerous/doesn’t merit vaccines is clearly flawed.

Indeed, in 2020, COVID was the 3rd leading cause of death in the USA, behind only cancer and heart disease (Ahmad and Anderson 2021), and in early 2021, it spiked to the #1 cause of death in the US (a situation that is no longer the case thanks in part to vaccines). That’s what COVID looks like when it is not controlled by measures like vaccines, and it should be abundantly apparent from that situation that simply having a 99% survival rate does not mean that a disease isn’t a serious threat. Indeed, if you are going to argue that we don’t need to be that worried about COVID 19, then you had better make that same argument about literally every other cause of death, because when allowed to spread, COVID kills more than any of them (or if you want to arbitrarily restrict yourself to the 2020 data, you’d better make that argument about every cause of death other than cancer and heart disease).

To put all of that another way, a 99% survival rate is only comforting if the spread of the disease is being controlled by something like vaccines. A 99% survival rate without vaccines still results in hundreds of thousands of people dying. That is a simple fact.

Natural herd immunity is a misnomer

Proponents of the “99% survive” argument often say that we will eventually achieve natural herd immunity, at which point people will be protected even without vaccines. Let me briefly explain why that is a terrible idea.

The term “herd immunity” really only makes sense within the context of vaccines. Vaccine-induced herd immunity = a population that is protected from a disease. “Natural herd immunity” = a survivorship bias in which the survivors of the outbreak are protected from future infection and many previous herd members died. “Natural herd immunity” is a bloodbath. It’s a nonsense term.

Estimates for what it would take to achieve herd immunity to COVID vary, but even at the (probably) unrealistically low end of 60% immune, getting there in the US without vaccines would involve the deaths of roughly 2 million Americans! At what is probably the more realistic threshold of 90% immune, without vaccines, we are looking at about ~3 million dead Americans.

That’s not herd immunity; that’s herd culling. Alternatively, we can achieve actual herd immunity with vaccines without killing 2–3 million people.

Disease risk is more than just death

One of the many deceptive aspects of the “99% survive” argument is that it acts as if COVID infection is a strictly binary situation: either you die or you are totally fine. That is, however, a completely false dichotomy. Since when is death the only outcome we care about? Many people who catch COVID live, but still undergo tremendous suffering physically, mentally, and financially. When we are considering the risks and benefits of something like a vaccine for COVID, we should not just look at whether or not you are likely to die. Things like time off work and physical suffering matter.

There is a vast spectrum of responses from “totally fine” to “alive, but extreme suffering (and in the US, often financial difficulty).” Good estimates of hospitalization rates are a bit hard to come by, and are usually expressed in terms of rate per population size, not rate per infection, but by taking the cumulative overall hospitalization rate per 100,000 people from the CDC, the US population size, and the cumulative cases in the US, we find that roughly 5% of COVID cases result in hospitalizations. A scientific study using more sophisticated methods provides a similar estimate of 4.5% (Reese et al. 2020) [note: as with the case fatality rate, these are crude rates for the whole population and actual rates vary by age group]. Indeed, COVID is so rampant, that roughly 0.5% of the entire US population has been hospitalized for it. That’s actually an incredible number. One out of every 200 people in America has been hospitalized for COVID (again if we average across the population), and that’s just actual hospitalizations, many more have been at home sick, took time off work, etc.

Further, the complications from the disease can be serious. COVID can set off a cytokine storm that affects multiple of your body’s major systems. It’s too early to know exactly what that means long-term, but based on what we know about COVID and other viral diseases, there is good reason to be concerned that many people who survive will have long-term lung and heart issues (Mitrani et al. 2020; Fraser 2020). To be clear, we need to wait for more data before knowing for sure if there will be long-term complications, but the reasonably high probability that there will be is certainly something to think about when considering the risks and benefits of the vaccine, and it’s something that is completely ignored by this unreasonably simplistic “99% survive” argument.

Vaccines: benefits outweigh the risks

Finally, let’s briefly turn our attention to the COVID vaccines. This would take many posts to cover in proper detail, so I’ll try to be brief and just hit the key points.

All vaccines have side effects, but serious side effects are rare, and your risk of a getting COVID and having a serious complication from it is much higher than your risk of a serious complication from vaccines. Countries with large vaccination programs (e.g., Israel; Rossman et al. 2020) are seeing massive drops in COVID infection, hospitalization, and death rates thanks to vaccines, and there has not been a comparable increase in injuries, hospitalizations, and deaths from vaccines. In other words, the vaccines are reducing risk.

In the USA, an analysis of data in May (by which time roughly 40% of the country had received both doses of the vaccine and 50% had at least one dose) found that out of over 107,000 COVID hospitalizations, fewer than 1,200 were for vaccinated individuals, and out of 18,000 COVID deaths, only ~150 of them were for vaccinated individuals. In other words, over 99% of COVID deaths were from unvaccinated individuals, as were nearly 99% of COVID hospitalizations. That trend has continued, and now with over half the country fully vaccinated, officials like Dr. Fauci are likewise reporting that over 99% of COVID deaths and over 97% of COVID hospitalizations are from the unvaccinated. The vaccines clearly work.

Further, we have not seen hospitalization or mortality rates from vaccine side effects rising to meet those numbers. In other words, with roughly half the population vaccinated, if the vaccine was truly more dangerous than the disease, we should see vaccine deaths and serious injuries at a rate that exceeds the rate for COVID. So where are the >18,000 individuals in May who died from the vaccine? Where are the >100,000 hospitalizations from the vaccine? They don’t exist, because serious side effects are extremely rare.

This clearly shows that your total risk is lower with the vaccine than without it. These numbers make that undeniable. Yes, the disease has a ~99% survival, and yes, the vaccine does have side effects, but the disease is exceptionally common, and series side effects of the vaccine are exceptionally rare, which, when taken together, results in your risk of serious injury or death being substantially lower when you have the vaccine.

Just to really drive this home, let’s look specifically at myocarditis following the mRNA vaccines. It is still not entirely clear if this is actually a side effect of the vaccines, but there is growing evidence that it is causal. Exact rates vary, but every calculation shows them to be low (Shay et al. 2021). A large data set from Israel reported a mere 148 cases within 30 days of vaccination out of over 5 million people who were vaccinated (Israel Ministry of Health). Even if all of those were caused by the vaccine, that would be a rate of 0.003%! Further, 95% of those cases were mild.

In the USA, with over 150 million people vaccinated, using data from VAERS (much of which has not been verified), the WHO calculated the following myocarditis rates per fully vaccinated individual: 0.041% for males aged 12–29, 0.004% for females aged 12–29, and 0.002% for males and females over 29. Further, again, most of those cases were mild and did not require serious medical intervention. I have not so far seen any confirmed deaths from this. In contrast, COVID has a case fatality rate of ~1%.

Now, I hear you saying, “but I am a male under 30, which means I’m at a low risk of death from COVID and higher risk of myocarditis, so for me the risk outweighs the benefit,” however, that math still doesn’t work. From the start of the outbreak in the US (based on the current CDC data), there have been 2,462 deaths from COVID in males under 30, and 1,594 deaths in females under 30. There is no evidence to suggest that the vaccines are causing comparable carnage.

Scientists actually ran the numbers on this and calculated that for males under 30, every 1 million second doses of an mRNA vaccine will result in 39–47 cases of myocarditis, but will prevent 11,000 COVID cases, 560 hospitalizations, 138 ICU admissions, and 6 deaths (Gargano et al. 2021). Also, remember again that most of these myocarditis cases are mild, whereas for COVID, even just the expected number of ICU admissions is 3 times the number of mostly mild myocarditis cases! Further, that is for a group with a relatively low COVID risk and relatively high myocarditis risk. The benefits of the vaccine are even more exaggerated for other groups. In men over 29, for example, 1 million second doses will only result in 3–4 cases of myocarditis, and will prevent 15,300 COVID cases, 4,598 hospitalizations, 1,242 ICU admissions, and 700 deaths (Gargano et al. 2021)!

Also, notice how those who like to cast aspersions on vaccines try to downplay death from COVID while hyping mostly mild injury from the vaccine. Indeed, this “99% survive” argument would have us believe that a 1% fatality is too low to be seriously concerned, but a 0.002–0.041% rate of generally mild myocarditis is unacceptably high. It is crazy to think that 3–4 mostly mild cases of myocarditis is worse than 700 deaths! That’s simply not how math works.

The numbers are undeniable: your personal risk* is lower with the vaccine than without the vaccine, even though “99% of people survive.”

I’ve been focusing on the concerns around Pfizer and Moderna because those are the prominent vaccines where most of my readers are, but we can do the same sorts of calculations with any of the vaccines currently available. They all have various risks, but in every case, the risks associate with not getting the vaccine outweigh the risks associated with getting it.

*Obviously some people have pre-existing conditions that make vaccines dangerous for them. That is not what I am talking about here. I am addressing people who have no known health conditions that would prevent them from getting the vaccine.


In summary, the “99% survive” argument completely ignores the high prevalence of COVID and completely ignores the importance of non-lethal effects (including hospitalization, possibly long-term effects, time off work, etc.). Further, using this argument against the necessity for vaccines massively and inappropriately downplays the risk from COVID while exaggerating the risks from vaccines. The empirical reality is that COVID is very dangerous, with over 600,000 dead in the US alone, whereas the vaccines are very safe. Given that over 50% of the US population is currently vaccinated, if the vaccines were truly more dangerous than the risk from getting COVID, we’d expect the rates of vaccine deaths and serious injuries to be surpassing the rates of deaths and serious injuries from COVID. In reality we are, of course, not seeing anything even remotely like that. Deaths and serious injury from COVID remain common in the unvaccinated, whereas serious injuries from the vaccines are extremely rare and deaths virtually unheard of.

I think part of the problem is that we often view taking an action as the inherently riskier choice. After all, you can’t get myocarditis from the vaccine if you never get the vaccine. That reasoning, while understandable, ignores the fact that not getting the vaccine is also an action, and that action puts you at a much greater risk of death or serious illness. Indeed, that risk from COVID is so great, and the risk from vaccines is so small, that even if you are young and healthy, the risk associated with not getting the vaccine is far higher than the risk associated with getting the vaccine. This is a mathematical fact.

Please read this post before making an argument about “unknown long-term effects” from vaccines.

Related Posts

Literature Cited

  • Ahmad and Anderson 2021. The Leading Causes of Death in the US for 2020. JAMA
  • CDC. Provisional COVID-19 deaths by sex and age. Accessed 26-July-2021
  • COVID.NET. Laboratory-confirmed COVID-10Associated Hospitalizations. Accessed 25-July-2021.
  • CDC Pinkbook. Poliomyelitis. Accessed 26-July-2021
  • Fraser 2020. Long term respiratory complications of covid-19. BMJ 370
  • Gargano et al. 2021. Use of mRNA COVID-19 Vaccine After Reports of Myocarditis Among Vaccine Recipients: Update from the Advisory Committee on Immunization Practices — United States, June 2021. CDC 70:977–982
  • Johnson and Stobbe. 30-June-2021. Nearly all COVID deaths in US are now among unvaccinated. AP
  • Israel Ministry of Health. 2-June-2021. Surveillance of Myocarditis (Inflammation of the Heart Muscle) Cases Between December 2020 and May 2021 (Including). Accessed 26-July-21.
  • Mitrani et al. 2020. COVID19 cardiac injury: Implications for long-term surveillance and outcomes in survivors. Heart Rhythm 17:1984–1990
  • Reese et al. 2020. Estimated Incidence of Coronavirus Disease 2019 (COVID-19) Illness and Hospitalization—United States, February–September 2020. Clinical Infectious Diseases 72:e1010-e1017
  • Rossman et al. 2020. COVID-19 dynamics after a national immunization program in Israel. Nature Medicine 27: 1055–1061.
  • Shay et al. 2021. Myocarditis Occurring After Immunization With mRNA-Based COVID-19 Vaccines. JAMA
  • Sullivan, 16-Jul-2021. U.S. COVID Deaths Are Rising Again. Experts Call It A ‘Pandemic Of The Unvaccinated’. NPR
  • WHO. 9 July 2021. COVID-19 subcommittee of the WHO Global Advisory Committee on Vaccine Safety (GACVS): updated guidance regarding myocarditis and pericarditis reported with COVID-19 mRNA vaccines. Accessed 26-July-2021.
  • Worldometer. Coronavirus. USA. Accessed 25-July-2021.
Posted in Uncategorized | Tagged , , , , | Comments Off on The “99% survive COVID” argument is deceptive and completely misses the point

Jon Stewart’s irresponsible, anti-science, COVID conspiracy theory rant

stewartI’ve been a fan of Jon Stewart for a long time. I usually find him to be both funny and insightful. It was, therefore, with great dismay that I watched him spread a conspiracy theory and inaccuracies about science on The Late Show with Stephen Colbert, and I regret that it is now my duty to explain his errors.

In the clip, which you can see here, Stewart begins by saying that we owe a great debt to science for helping to ameliorate the COVID19 crisis. That much I absolutely agree with. Scientists deserve an enormous amount of gratitude for developing the vaccines that are currently saving lives and reducing suffering around the world. Unfortunately, the interview quickly took a turn for a worse as Stewart endorsed the conspiracy theory that the virus that causes COVID19 escaped from the lab in Wuhan and made numerous dangerous, false statements about how scientists operate.

Mischaracterizations of science and scientists

Before I get to the conspiracy theory, I want to respond to a more general comment he made which really bothered me, because it is a common and damaging misconception about science and scientists. He said,

“This is the problem with science. Science is incredible, but they don’t know where to stop, and nobody in the room with those cats ever goes, ‘ya know, I don’t know if we should do that.’”

This is an inherently untrue statement that is, unfortunately, a very common misconception about science, perhaps most famously stated by Dr. Ian Malcom in Jurassic Park as, “Your scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should.”

In reality, most of us scientists spend a great deal of time thinking about the implications of our work and whether it is a good idea, and even if we didn’t want to do that, we are forced to do that by ethics committees. The days of “mad scientists” operating in isolation doing dangerous, unethical experiments are long gone. In modern science, you have to submit proposals before you do research, and those proposals get reviewed by ethics boards (which, btw, usually include scientists and non-scientists) who will approve, reject, or insist on modifications to your proposal. If you get approved, you then have to submit reports on what you actually did, and there are usually audits. Further, if you want to work with something dangerous, you often need additional government clearance and approval. Things are checked and regulated, particularly in a major lab like the one in Wuhan.

Indeed, Stewart went on to give an example that undercut his whole argument. In a deranged monologue in which he asserted (I hope jokingly) that scientists would be the death of humanity, he inaccurately described scientists’ re-animation of the virus that caused the 1918 flu,

“They had a little sample of [the virus], and it hadn’t been a scourge in the earth for 100 years, and they thought to themselves, ‘what if we just, I don’t know, woke it up?’ And nobody in the room was like, ‘No. Let’s not do that.'”

The reality of the situation was quite different. The research was conducted by the CDC and went through both an Institutional Biosafety Committee review and an Animal Care and Use Committee review before being conducted. The risks and benefits were carefully thought out, evaluated, and discussed by several groups of people, and very strict bio-safety guidelines were put in place and followed. This wasn’t a scientist working in their garage doing whatever popped into their head. It was, carefully planned, well-thought out, closely monitored, rigorously controlled research in a secure, state-of-the-art laboratory (you can read the details on the CDC’s website).

To be 100% clear, that obviously does not mean that no one ever goes rogue and operates outside of the system. There will always be people who break the rules, but those cases are rare, they are pretty close to impossible at a lab facility like Wuhan, and when those people get caught, the consequences are generally serious (such as loss of funding, loss of job, and sometimes prosecution). So, this notion that no one is telling scientists not to do things is simply false. There are constantly people checking our proposals and telling us not to do things.

This misconception matters, because it paints scientists as people who are totally disconnected from reality, blindly doing whatever interests us regardless of the dangers, but that’s a caricature, not reality. Scientists are normal people. We think about things like ethics and broad implications, and we have people checking us. Further, this sort of caricature downplays the importance of labs like the Wuhan lab. It’s not full of “mad scientists.” It’s full of incredibly intelligent, thoughtful people doing important work that needs to be done if we are going to prevent future disease outbreaks and mitigate the ones that we fail to prevent. Indeed, that Spanish Flu research that Stewart mocked gave us important insights into how highly virulent pathogens operate, and those insights help us to prepare for future outbreaks. This sort of mischaracterization of scientists undermines the critical work that scientists do.

The lab leak conspiracy theory

Moving on, I’m not going to go into great detail debunking the Wuhan lab leak conspiracy theory, because it has been thoroughly covered elsewhere (e.g., Siegel 2021). Instead, I’ll just hit a few highlights and point out some errors in Stewart’s reasoning. Before I start, I do realize that much of what he said was deliberate hyperbole for comedic effect, so I have no intention of nitpicking his jokes and will instead focus on the broad strokes.

The crux of his argument was basically just proximity:

  1. There is a lab in Wuhan that studies novel coronaviruses
  2. The novel coronavirus originated in Wuhan
  3. Therefore, it escaped from the lab

To anyone familiar with logic, this reasoning is clearly fallacious. It’s a non sequitur fallacy. Just because the two things occurred in proximity doesn’t mean that there was any relationship between them (note: you might argue that this is a correlation fallacy, but that generally involves trends, not isolated incidents; it is arguably a post hoc ergo propter hoc fallacy [i.e., there was a lab in Wuhan, then there was an outbreak; therefore the lab caused the outbreak]).

It is entirely possible for the co-occurrence of two things to be a coincidence, or for there to be some other factor that caused those two things to line up. In this case, the latter is true. Infectious disease labs very often focus on the pathogens around them, and as Colbert tried to point out to Stewart, China has many coronaviruses, and Wuhan is a massive city with an active wildlife trade. In other words, if a novel coronavirus was going to jump to humans, we would expect it to be in a city like Wuhan, which is why there was a coronavirus lab in Wuhan. As one virologist (Vincent Munster) commented,

“Nine out of ten times, when there’s a new outbreak, you’ll find a lab that will be working on these kinds of viruses nearby” (Maxem and Mallapaty 2021).

So, no, the fact that the outbreak originated in a city that has a coronavirus lab does not suggest that the outbreak was a result of a lab leak. That is extremely circumstantial evidence and totally ignores the fact that labs are often positioned in the areas where we’d expect an outbreak.

Similarly, Stewart dismissed the abundance of bats around Wuhan by pointing out that there are also lots of bats in Texas, but the outbreak didn’t start there. This is more specious logic (indeed, it’s a reductio ad absurdum fallacy). It’s not simply that there are bats in China, but rather that China is  a hot spot of human-wildlife interactions. It is those interactions that are problematic and, it is something scientists have been warning about for years. Back in 2007, a paper in Clinical Microbiology Reviews said,

“The presence of a large reservoir of SARS-CoV-like viruses in horseshoe bats, together with the culture of eating exotic mammals in southern China, is a time bomb” (Cheng et al. 2007).

This isn’t a virus that came out of nowhere with no warnings. It’s something scientists have worried about for a long time, which, once again, is why labs that study these diseases are so important, and why it is so dangerous to undercut them by spreading this kind of conspiracy theory. It’s also, again, why this lab was in China.

As you can hopefully see, Stewart’s arguments are bogus, but what should we make of the lab leak hypothesis more generally? As I said, I’m not going to go into a ton of detail, because that is a whole series of posts in itself, but here are a few highlights.

First, the WHO investigation into the Wuhan laboratory concluded that it was “extremely unlikely” that the virus had escaped from the lab. Second, scientific assessments of the sequences in SARS-CoV-2 have shown that it is most likely natural. Andersen et al. (2020) found no evidence to suggest that it was man-made and,

“strong evidence that SARS-CoV-2 is not the product of purposeful manipulation.”

If the virus was engineered, we’d see evidence of that in the genome, but we don’t see that evidence. This is critical, because it means that even if it escaped from a lab, it was probably a natural virus that was being stored there, not one that was engineered. As you will see, however, the arguments for the lab leak hypothesis generally center around the virus not being natural (you can read more about this research and how we know the virus wasn’t man-made in these articles: Bryner 2020; Hayes 2020; Saey 2020).

So, where does this leave the lab-leak hypothesis? It leaves it solidly in the territory of conspiracy theories that ignore official reports, cherry pick experts, and spin fanciful tales that attempt to “connect the dots” based on circumstantial evidence. This hypothesis inherently requires that scientists and governments have lied and falsified records, but there is no actual evidence of that happening. It is just something that has to be blindly assumed in order for this conspiracy theory to be possible, and as a scientist, I have an aversion to making more assumptions than necessary.

The more scientific arguments for the lab leak hypothesis have tended to focus either on the fact that we have yet to find the original virus in the wild yet or on an unusual CGG-CGG sequence. The fact that we have not confirmed the original host is, however, hardly surprising. It often takes many years to track novel virus back to the original host. It took 14 years to find the origin of the SARS epidemic (Maxem and Mallapaty 2021). So, 1.5 years of failed searching is hardly enough to conclude that it probably isn’t natural. The other argument is that the furin cleavage site of SARS-CoV-2 has a CGG-CGG sequence that is unusual. CGG codes for the amino acid arginine, but there are several other ways to produce arginine, and CGG is relatively uncommon in coronavirus. This led David Baltimore to famously refer to the double CGG as a “smoking gun” that the virus was man made. That statement was, however, misleading, because about 3% of the arginine sequences in SARS-CoV-2 are CGG, and CGG is found in plenty of other coronaviruses. It is, therefore, entirely possible for CGG-CGG to arise naturally, and even Baltimore has now acknowledged that it could be natural (Maxem and Mallapaty 2021). Dr. Anderson, who wrote the study I mentioned earlier, explained the situation in more detail, here.

Finally, what are we to make of various governments and even Dr. Fauci saying there should be thorough investigations? First, as I explained in a previous post this week, we should never conflate government decisions with scientific evidence. Second, if people like Fauci want more investigations, I don’t really object (I don’t think they are necessary, but I’m not going to rant in opposition of them either). However, critically, as it stands right now, the weight of evidence is strongly against the lab leak hypothesis, and arguments to the contrary inevitably stray into the realm of conspiracy theories, conjecture, and baseless assumptions. As such, it is extremely problematic for a man like Stewart to insist on national television that the virus originated in a lab. Look, if you want to say, “the current evidence suggests that it did not escape the lab, but out of an abundance of caution and in the interest of being as certain as possible, we should conduct another investigation” I don’t necessarily have a problem with that, but that is a very, very different thing from the type of proclamation that Stewart made or the types of conspiracy theories that I’m seeing online. Further, it absolutely does not justify Stewart’s rant about what he perceives as the dangers of science.

Let me conclude this with a question. If you doubt the official reports and want another investigation, will you accept the results of that investigation if it shows the disease was natural? To put that another way, how many investigations are required to convince you? This is a question that is worth thinking about. It is always a good idea to ask yourself what it would take to convince you that you were wrong.

Update 1-Sept-2021: A new review paper (Holmes et al. 2021) which was released after this post has reviewed the evidence for the lab leak hypothesis in even more detail, and concluded that the evidence strongly points away from that hypothesis.


  • Andersen et al. 2020. The proximal origin of SARS-CoV-2. Nature Medicine 26:450-452
  • Bryner 2020. The coronavirus was not engineered in a lab. Here’s how we know. Live Science (21 March 2020)
  • Cheng et al. 2007. Severe Acute Respiratory Syndrome Coronavirus as an Agent of Emerging and Reemerging Infection. Clinical Microbiology Reviews 20:660–694.
  • Hayes 2020. Here’s how scientists know the coronavirus came from bats and wasn’t made in a lab. The Conversation (13 July 2020).
  • Holmes et al. 2021. The Origins of SARS-CoV-2: A Critical Review. Cell
  • Maxem and Mallapaty 2021. The COVID lab-leak hypothesis: what scientists do and don’t know. Nature: News Explainer (8 June 21)
  • Saey 2020. No, the coronavirus wasn’t made in a lab. A genetic analysis shows it’s from nature. Science News (26 March 2020).
  • Siegel 2021. The Wuhan Lab Leak Hypothesis Is A Conspiracy Theory, Not Science. Forbes (3 June 2021)
Posted in Nature of Science | Tagged , , , | Comments Off on Jon Stewart’s irresponsible, anti-science, COVID conspiracy theory rant

The “it’s banned in Europe” fallacy

banned in europe

The title of this post is written somewhat in jest because this is not a formally recognized fallacy; nevertheless, it is a very common line of reasoning that is logically flawed and very closely aligned with multiple fallacies. The argument, in a nutshell, asserts that something is dangerous or likely dangerous simply because it is banned in some (usually European) countries. It often takes forms such as, “if X is so safe, then why is it banned in so many countries in Europe? All of those countries can’t be wrong.” Most commonly, I see this flawed reasoning applied to GMOs, certain pesticides (e.g., glyphosate), and some pharmaceuticals, but the flaws are inherent to the argument structure, regardless of the topic.

Recently (and why I’m writing this now), I have been seeing it applied to the AstraZeneca COVID 19 vaccine, with the claim that the vaccine was banned in Europe due to concerns about blood clots, therefore, according to the argument, the vaccine must be dangerous. Before I go any further and talk about the general problems with this line of reasoning, let me be 100% clear that the vaccine was not “banned.” Rather, it was only temporarily halted while scientists investigated the reports of blood clots. Those investigations found that the rates of blood clots following vaccination were below or only slightly above background rates (Østergaard et al. 2021; Pottegård et al. 2021), and the European Medicines Agency (EMA 2021) concluded that while there are potential side effects that are being monitored (as is true of all medicines), any serious side effects are extremely rare, and the benefits outweigh the risks. As a result, the vaccine has been resumed in almost all countries that had temporarily halted it.

Moving on, let’s look at the more general problems with this line of reasoning. First, it is inherently an appeal to authority fallacy. It is asserting that something must be dangerous simply because of the people (governments) who say that it is dangerous. This is problematic for multiple reasons. First, government decisions are often influenced by political motivations and lobbyists, and in many cases, the people making decisions are politicians, not scientists. Indeed, this argument is very similar to the flawed tactic of citing court cases as scientific evidence, as if judges, lawyers, and jurors are somehow scientific experts who are qualified to evaluate scientific evidence and make scientific decisions. Non-scientists (and sometimes scientists) make bad calls about science all the time. Do I really need to explain that governments often get the science wrong? Isn’t that something that everyone intuitively knows? After all, Saudi Arabia bans the teaching of evolution in public schools, does that mean evolution is wrong? Obviously not. We have to look at facts and evidence, not just whether or not a government has banned something.

Nevertheless, you may be wondering about the subset of cases where the ban actually was the result of a scientific assessment by an agency such as the FDA, CDC, EMA, etc. Shouldn’t we listen to those organizations? That is a totally fair question, and in that subset of cases, the argument should be treated more seriously. However, it still ultimately has to be about the evidence. Even scientists on government agencies can make bad calls. So, ideally, we should base our views on the evidence, not the source of the evidence. Having said that, for most people who don’t have the time or training to wade through a mountain of scientific literature, it does often make sense to default to the relevant scientific agencies agencies. I would not, as a general rule, encourage you to challenge professional scientific organizations unless you have a considerable amount of scientific training and expertise yourself.

So, what do we do when a group like the EMA recommends halting a vaccine? There are two important considerations here. First, we need to look at precisely what they said, and why they said it, because usually, the situation is far more nuanced than simply, “they banned it.” The AstraZeneca situation illustrates this wonderfully. The vaccine was not initially “banned.” Rather, it was temporarily halted while the evidence was reviewed. This was done out of an abundance of caution, and once the possible harms had been evaluated, it was resumed in nearly all countries (as a side note, this actually shows just how seriously vaccine safety is taken; if anti-vaccers were right that governments/Big Pharma push vaccines with little regard for safety, then such halts wouldn’t take place).

The second consideration is the extent of the consensus on the topic. Was the ban recommended by a handful of organizations or was it recommended by the vast majority of scientific organizations? This is where we get to another one of the underlying flaws in the argument. Namely, it has a tendency to cherry pick experts. What I see happen over and over again is someone insisting that something must be dangerous because some countries banned it while totally ignoring the fact that just as many (often more) countries have approved it! In other words, if you can appeal to authority and claim that something is dangerous because some countries ban it, then it is equally logically valid* for me to claim that it is safe because some countries approve it (*both lines of reasoning are invalid, my point is simply that they are equivalent).

See how that works? If we are going to go down this road of blindly trusting authority, why should we blindly assume that the countries that ban something are right rather than blindly assuming that the countries that approve it are right, especially when (in a great many cases) the countries that approve it are more numerous than the countries that ban it?

Indeed, the AstraZeneca situation illustrates this well. While a handful of countries continue not to use it, most have resumed using it, and the EMA now recommends it. So why should we cherry pick the handful that don’t use it and hold them up as proof while totally ignoring the ones that do use it? What makes the countries that don’t use it right and the far more numerous countries that do use it wrong?

The answer is simple, there is no logical reason to blindly assume that something is dangerous just because a country banned it. We need to look at the actual evidence and the reasoning behind why it was banned. We shouldn’t determine whether something is safe by flipping a coin between lists of countries that do and do not use it, and we certainly shouldn’t decide by cherry picking the list that fits our biases and preconceptions. Rather, we need to look at the actual evidence and the logic of why it is banned or approved.

Related posts

Literature cited

  • EMA 2021. COVID-19 Vaccine AstraZeneca: benefits still outweigh the risks despite possible link to rare blood clots with low blood platelets. 18-March-21.
  • Pottegård et al. 2021. Arterial events, venous thromboembolism, thrombocytopenia, and bleeding after vaccination with Oxford-AstraZeneca ChAdOx1-S in Denmark and Norway: population based cohort study. BMJ 373:n1114
  • Østergaard et al. 2021. Thromboembolism and the Oxford–AstraZeneca COVID-19 vaccine: side-effect or coincidence? The Lancet 397: 1441–1443.

First posted on 14-June-2021

Posted in Rules of Logic, Vaccines/Alternative Medicine | Tagged , , , , , , | 16 Comments

The problems with anti-vaccers’ precautionary principle arguments

Invoking the precautionary principle is a favorite tactic of anti-vaccers, anti-GMO activists, and various other groups that are prone to opposing scientific advances, but there are numerous issues with this strategy. The exact definition of the precautionary principle is a bit amorphous and variable, but the general concept is that before taking an action that has potential risks (particularly with the application of something new), the burden of proof should be on the proponent to demonstrate the safety of that action before taking it. There is certainly an element of validity to this. I absolutely agree, for example, that new medications should be tested and shown to be safe before being made publicly available (indeed, that is part of why we have organizations like the FDA that regulate the approval of new drugs). Similarly, months ago when COVID vaccines were in early stages of testing, I and many other ardent pro-vaccers stated that while the vaccines were promising, we wanted to see the results of the tests before drawing any conclusions. That is a completely rational, evidence-based way to approach the situation. New drugs, vaccines, and technologies certainly have the potential to cause harm. So, it is not unreasonable to want safety testing prior to their use.

Where this becomes problematic, however, is in determining what constitutes reasonable evidence. Groups like anti-vaccers have a tendency to stretch the precautionary principle to unreasonable limits and use it as an excuse for science-denial, rather than a legitimate decision-making tool. Usually, what I see is an incorrect insistence that the precautionary principle demands absolute assurance of safety and that it should be used in a strictly binary decision tree where any possible risk means that the thing in question should not be used, regardless of the known benefits. This is very bad risk assessment and ignores important aspects of how science actually works. As I’ll explain, science is about probabilities, not absolutes, and you must consider both the risk and benefit of an action.

Note for clarity: Just to be sure that I have been clear, I am not attacking the general concept of the precautionary principle. Rather, I am taking issue with the way that many people apply it (and you could make an argument that they are not actually applying it at all and are instead using something else entirely that they are incorrectly calling the precautionary principle. Indeed, many definitions of the precautionary principle explicitly state that it only applies in the absence of scientific evidence).

Science is about probabilities, not proof

This is a really important concept that I have written about frequently. Science is an inherently probabilistic endeavor. It shows us what is most likely true given the current evidence, not what absolutely is true. It always leaves open the possibility that the current results are wrong or some other piece of evidence has been missed. Indeed, the statistics we use to determine things like drug safety and efficacy are inherently probabilities. When we say that a result is statistically significant, what we really mean is that if there is actually no effect (i.e., all results are from chance), there is a low probability that a result as great or greater than the one we observed could arise (i.e., if there is no effect of the thing being tested and we did the experiment again, we’d be unlikely to get such a strong result). Probabilities are inherent to modern science.

Ultimately, this is a good thing. It forces skepticism and avoids dogmatism. Unfortunately, opponents of science seize that inherent and entirely justified skepticism and erroneously conflate it with practical doubt. The fact that we can’t be 100% about a result does not inherently mean that we should have any practical doubt about it. Am I 100% certain that smoking causes cancer? No, but the topic has been so well-studied and the results are so consistent that I’m 99.999% sure, and for all practical purposes, there is nothing wrong with making a statement like, “smoking causes cancer.” Similarly, it is technically possible that scientists are wrong about gravity, but, to borrow from Tim Minchin, you would be pretty foolish to act as if there is any practical doubt about gravity when deciding whether to exit your apartment through your door or a second floor window.

Even so, for many vaccines, we have an extremely high level of certainty that they are safe and effective. Take the notion that vaccines cause autism, for example. As I’ve written about at length, this hypothesis has been extremely well-studied. Multiple studies with tens of thousands of participants have been conducted (e.g., (Hviid et al. 2019 [657,461 children]; Madsen et al. 2002 [440,654 children]; Anders et al. 2004 [109,863 children]; and Jain et al. 2015 [95,727 children]), including a meta-analysis with over 1.2 million children (Taylor et al. 2014), and no large, properly controlled study has found any evidence of vaccines causing autism. Nevertheless, I frequently encounter anti-vaccers who try to ignore that evidence by inappropriately invoking the precautionary principle and asserting that, “since we can’t be 100% sure that vaccines don’t cause autism, we should err on the side of caution and act as if they do cause autism.” That is extremely faulty logic and is nothing more than science denial dressed up as a cogent decision-making principle.

The other related issue is the never-ending string of possible mechanisms of harm. Anti-vaccers frequently concoct an ever-shifting litany of things that scientists need to test before they will accept that vaccines are safe, and they often do this under the guise of simply adhering to the precautionary principle. Sticking with autism for a minute, for a long time, mercury was the main anti-vax boogeyman (and it still is in some circles), and anti-vaccers insisted that the burden of proof was on pro-vaccers to show that the mercury wasn’t causing autism (fundamentally a precautionary principle argument; see note on the burden of proof). So, scientists did lots of tests and even removed thimerosal (the form of mercury in vaccines) from nearly all childhood vaccines. The results of those studies consistently showed that thimerosal doesn’t cause autism (Hviid et al. 2003; Verstraeten et al. 2003;  Taylor et al. 2014), and removing it did not reduce autism rates.

So the burden of proof has been met and the precautionary principle satisfied, right? Not according to anti-vaccers. According to them, maybe its actually the age at vaccination, and the precautionary principle says that we need to demonstrate that it isn’t age at vaccination before we consider vaccines to be safe. Then, when studies show that it isn’t the age at vaccination (Uno et al. 2015; Destefano et al. 2004; Smeeth et al. 2004; Madsen et al. 2002), they switch to number of antigens, and when studies discredit that (DeStefano et al. 2013), they switch to it being the number of doses, and when studies discredit that (Fombonne et al. 2006; Hviid et al. 2003), they switch to aluminum or countless other fantasies.

It’s like fighting the hydra. No matter how many arguments you test and defeat, more crop up to take their place. This is the problem with the practical application of the precautionary principle. There will always be other possibilities. No matter how many things we test, there will always be things that haven’t been tested. This is why the burden of proof is usually on the person making the claim, and by switching the burden of proof, the precautionary principle opens a can of unending worms (see note on burden of proof).

Again, to be clear, I agree with a reasonable level of testing before something like a vaccine goes to market, and even after it goes to market, if reasonable evidence arises that it is causing a problem, I agree that the evidence should be investigated and proper trials should be done. If there is a legitimate, science-based reason to suspect that a risk might exist, it should be investigated. That is a totally reasonable application of the precautionary principle. The problem is that many people try to use it unreasonably and insist that all possibilities must be tested, even if they can’t present any good evidence to show that a danger is likely. Actually testing all possibilities is, however, impossible. Thus, anti-vaccers can be immune to studies, because no matter how many we conduct, there will inevitably still be things we haven’t tested.

A similar issue arises with an insistence for “long-term” studies. As I’ve argued before, “long-term” is meaningless unless it’s carefully defined beforehand. We have studies that followed patients for multiple years (Idbal et al. 2013; Ferris et al. 2014; Vincenzo et al. 2014), which would fit the definition of “long-term” for most scientists, but this never seems to satisfy anti-vaccers. No matter how long the study is, they will always retort that the negative effects might come at some later age. This flawed reasoning is fundamentally just an argument from ignorance fallacy. It is saying that we don’t know for sure that there aren’t problems 35 years later, therefore we should act as if there are. That is bad logic unless we have some compelling reason to think that there would be negative effects 35 years later, which we don’t. We do, however, have plenty of evidence that there are enormous benefits to vaccines and the risk from not vaccinating is much higher than the risk from vaccinating. This brings me to the next topic: risk assessment.

Bad risk assessment

Good risk assessment needs to consider both the risks and the benefits within a probabilistic framework. All actions carry some level of risk, and people often ignore the fact that inaction can be riskier than action.

So, when it comes to topics like vaccines, we need to consider both the risks and the benefits. Vaccines do have side effects, but serious ones are very rare, and we know that the benefits far outweigh those risks because numerous studies have shown that vaccines are extremely beneficial at saving lives (Clemens et al. 1988; Adgebola et al. 2005; Richardson et al. 2010). Indeed, the WHO estimates that from 2000 to 2018, the measles vaccine alone prevented 23.2 million deaths! That benefit absolutely has to be considered when evaluating the risk of vaccines.

This is another place where anti-vaccers misuse of the precautionary principle becomes problematic. They focus on hypothetical potential risks that have somehow eluded all previous studies and argue that we should “err on the side of caution” before potentially injuring children with vaccines. That is horrible risk assessment because it totally ignores the massive, well-established risk from not vaccinating. In other words, the cautious approach is actually to use vaccines that have passed reasonable safety testing, because the known risk from not vaccinating is so high.

Is it possible that there is some unknown danger from vaccines that we have missed? Yes. As explained above, it is technically possible, but it is extremely unlikely. Meanwhile, we know that there are massive dangers from not vaccinating. Studies have repeatedly shown that vaccines save countless lives. Therefore, it is absurd to knowingly sacrifice those millions of lives out of a fear of some unknown danger which probably doesn’t even exist! That’s not erring on the side of caution, and it is not a proper application of the precautionary principle.

Similarly, when it comes to the issue of long-term studies, is it technically possible that there is some long-term effect that we haven’t found yet? Yes, but it is very unlikely. Meanwhile, we know that vaccines prevent diseases which are often fatal for children and can have long-term consequences for survivors. Here again, anti-vaccers are asking us to put more weight on an unknown and unlikely risk than on a very real and well-known risk. Further, even beyond the known risks, it is very possible that measles and other childhood diseases have additional long-term complications that we are unaware of. Indeed, based on our understanding of physiology and diseases, it is more likely that vaccine-preventable diseases have additional unknown long-term consequences than it is that vaccines do. Measles, for example, negatively affects children’s immune systems for years (Petrva et al. 2019; Mina et al. 2019; more details and sources here and here), thus opening the door for all manner of secondary infections and long-term complications. So if we are going to play this game of fearing the unknown, why should vaccines be the unknown that we fear rather than the dieses?

The answer given to that question often involves some variant of the appeal to nature fallacy, and asserts that we shouldn’t play God, or can’t improve on nature, or humanity survived for millennia with these disease, etc. These responses are all obviously flawed for numerous reasons that I have elaborated on elsewhere, so I won’t waste any more time here (see posts here, here, here, and elsewhere).

What about COVID vaccines?

Finally, let’s apply all of this to the new COVID vaccines, because I have recently seen countless people using some variant of the precautionary principle to argue against receiving the COVID vaccine.

First, it is true that, by the very nature of being new vaccines, the COVID vaccines have not yet received as much testing as most vaccines currently on the market*, but they have received just as much or more testing as those vaccines had when they first entered the marked, and the amount of testing they have received is very good. The Pfizer trials used over 43,000 participants (Polack et al. 2020), Moderna used over 30,000 (Mahase 2020), etc. Those are very large trials with good power to detect adverse effects, and they found that the vaccines were very effective (generally with effectiveness in the high 80s or 90s), with few serious complications (comparable to existing vaccines). This is very good evidence that the vaccines are safe and effective, and it meets any reasonable application of the precautionary principle. The problem is precisely that anti-vaccers’ application is unreasonable, so instead of accepting the results, they are spreading baseless fear about possible unknown dangers and future long-term consequences.

*Update 1-Sept-2021: This is no longer correct. The COVID vaccines are now extremely well-studied. See this post for details.

Again, I agree that those trials were necessary before the vaccines went to market. I don’t blindly support vaccines. Rather, I base my views on the evidence, and the evidence shows that these vaccines are safe and will save countless lives. Is it possible that scientists are wrong about these vaccines? Yes, but based on all the available evidence, it’s not likely. Nevertheless, many insist that because it is possible that we have missed something, we shouldn’t vaccinate and should instead apply “caution.” As I’ve been trying to explain throughout this post, that sets up a false dichotomy, because not vaccinating has substantially more risks than vaccinating.

Let’s just compare the two for a minute. The vaccines have passed large, well-conducted trials. Further, they are based on technologies that have been being researched for many years, and many of their components are the same as other vaccines that have passed numerous, repeated tests. All of this gives us very good reason to think that they are safe, and the probability that they are actually dangerous and we just haven’t found out yet is extremely low. In contrast, we know that COVID is highly contagious and highly deadly. Total global deaths are over 2.5 million, and in the USA, COVID is currently one of the leading causes of death (it spiked all the way to the number 1 spot during the large outbreak at the beginning of 2021). So, the known risk from not taking the vaccine is extremely high (see notes at the end before responding with the inane “it *only* kills 1 in 100” or “it only kills the elderly” arguments).

Further, if we want to go down this road of fearing unknown long-term complications, unlike the vaccine, there is very good reason to think that COVID will cause long-term problems. There is growing evidence that many patients have complications long after being infected (though the disease is recent enough that the data are still being collected and a clear picture hasn’t emerged yet), and given the damage that COVID is known to cause to the heart, lungs, and other organs, and our knowledge of other diseases, it is very reasonable to think that there might be long-term problems (Mitrani et al. 2020; Fraser 2020).

So, on the one hand, we have few known risks from the vaccine and no good reason to suspect unknown long-term complications, and on the other, we have an extremely high known risk from COVID, as well as good reason to suspect that there might be long-term damage.

Thus, proper application of risk assessment absolutely does not support avoiding the vaccine. Reasonable concerns have already been tested (thus appropriate precautions were taken), the known risks of COVID are substantially higher than the known risks of the vaccine, and there is far more reason to suspect unknown future complications from COVID than from the vaccine.

If your concern is really that there might be currently unknown long-term damage, then really think about which of the following is more likely to cause such damage: a vaccine that has been well-tested and simply stimulates your immune system and prepares it to fight a single pathogen, or a deadly virus that sets off a cytokine storm and is known to cause serious damage to your heart, lungs, and other organs. Which one actually seems riskier to you?

Part of the problem here is that we often perceive a decision not to take action as the safe option or “erring on the side of caution,” but that’s not always true. Not taking action still has risks, and taking action is not automatically “erring on the side of caution.” The precautionary principle no longer applies to approved vaccines because they have already passed testing. At this point, it is a simple matter of risk assessment, and the risks from not vaccinating are far higher than the risks from vaccinating.

See this post for details on why vaccines are highly unlikely to cause long-term side effects.

Note on the burden of proof: The burden of proof always lies with the person making the claim. In other words, the person saying something exist has to provide evidence that it exists, and the other person does not have to discredit its existence (e.g., the burden of proof is on someone claiming that bigfoot is real, not someone claiming it isn’t). The precautionary principle inherently flips this by saying that we need to provide evidence that a danger doesn’t exist, rather than the burden being on the person claiming that it does exist. As explained earlier, given the very real possibility of injury form something like a new medicine, some basic safety testing is rational prior to approving the drug. However, once reasonable testing has been done, the burden of proof then falls to anyone who claims that there is a risk that those tests missed. In other words, if you want to say that those tests were wrong, then the burden of proof is on you. This is fundamentally why anti-vaccine arguments fail. They try to shift the burden of proof rather than presenting actual evidence.

Note on the “it only kills 1 in 100” argument: I frequently hear people make the argument that we don’t need the vaccine because most people survive COVID. This is a very bad argument for numerous reasons, which I will only briefly outline (see detailed explanation here). First, 1 out of every 100 infected individuals is actually a very high death rate. Second, you also have to consider how infectious the disease is. Even if a disease only kills a minority of infected individuals, that can still result in millions of deaths if lots of people become infected. This absolutely is the case with COVID. Again, it is currently the leading cause of death in the USA. Third, this totally ignores all the non-lethal effects. Fourth, this totally ignores the massive economic harm being caused by the virus. Fifth, this totally ignores the fact that the odds of a currently unknown future side effect from COVID are still much higher than the odds of a currently unknown future side effect from the vaccine.

Note on the “it only kills the elderly” argument: First, no it doesn’t. Yes, mortality rates are much higher for the elderly, but they still exist for all other age groups. In contrast, the mortality rate from the vaccine is 0 for all age groups. So the vaccine reduces risk for all age groups. Additionally, if we want to protect the elderly, the rest of us need to vaccinate to reduce the spread of the disease (vaccines are often less effective for the elderly, not to mention that many vulnerable people simply cannot receive the vaccine). Finally, this argument suffers most of the same flaws as the “it *only kills 1 in 100” argument, so see that note as well. 

Related posts

Literature cited

(note: if these hyperlinks break and/or you have trouble accessing articles for free, see this post for suggestions about how to access them)

  • Adegbola et al. 2005. Elimination of Haemophilus influenzae type b (Hib) disease from The Gambia after the introduction of routine immunisation with a Hib conjugate vaccine: a prospective study. The Lancet 366:144–150
  • Anders et al. 2004. Thimerosal exposure in infants and developmental disorders: a retrospective cohort study in the United Kingdom does not support a causal association. Pediatrics 114:584–591
  • Clemens et al. 1988. Mesles vaccination and childhood mortality in rural Bangladesh. American Journal of Epidemiology 128:1330–1339
  • DeStefano et al. 2013. Increasing exposure to antibody-stimulating proteins and polysaccharides in vaccines is not associated with risk of autism. J Ped 163:561–567
  • Ferris et al. 2014. Long-term study of quadrivalent human papillomavirus vaccine. Pediatrics 134: e657-665.
  • Fombonne et al. 2006. Pervasive Developmental Disorders in Montreal, Quebec, Canada: Prevalence and Links With Immunizations. Pediatrics 118
  • Fraser 2020. Long term respiratory complications of covid-19. BMJ 370
  • Hviid et al. 2003. Association between thimerosal-containing vaccine and autism. JAMA 290:1763–1766.
  • Hviid et al. 2019. Measles, mumps, rubella vaccination and autism: A nationwide cohort study. Annals of Internal Medicine.
  • Idbal et al. 2013. Number of antigens in early childhood vaccines and neurophsychological outcomes at age 7–10 years. Pharmacoepidemiology and Drug Safety 22:1263–1270.
  • Jain et al. 2015. Autism occurrence by MMR vaccine status among US children with older siblings with and without autism. JAMA 313:1534–1540
  • Madsen et al. 2002. A population-based study of measles, mumps, and rubella vaccination and autism. New England Journal of Medicine 347:1477–1482
  • Mahase 2020. Covid-19: Moderna vaccine is nearly 95% effective, trial involving high risk and elderly people shows. BMJ 371
  • Mina et al. 2019. Measles virus infection diminishes preexisting antibodies that offer protection from other pathogens. Science 366:599–606
  • Mitrani et al. 2020. COVID-19 cardiac injury: Implications for long-term surveillance and outcomes in survivors. Heart Rhythm 17:1984–1990
  • Petrva et al. 2019. Incomplete genetic reconstitution of B cell pools contributes to prolonged immunosuppression after measles. Science Immunology 4: eaay6125
  • Polack et al. 2020. Safety and efficacy of the BNT162b2 mRNA Covid-19 Vaccine. New England Journal of Medicine 383:2603–2615.
  • Richardson et al. 2010. Effect of rotovirus vaccination on death from childhood diarrhea in Mexico. New England Journal of Medicine 362:299–305
  • Smeeth et al. 2004. MMR vaccination and pervasive developmental disorders: a case-control study. Lancet 364:963–969
  • Taylor et al. 2014. Vaccines are not associated with autism: and evidence-based meta-analysis of case-control and cohort studies. Elsevier 32:3623-3629
  • Uno et al. 2015. Early exposure to the combined measles-mumps-rubella vaccine and thimerosal-containing vaccines and risk of autism spectrum disorder. Vaccine 33:2511–2516
  • Verstraeten et al. 2003. Safety of Thimerosal-Containing Vaccines: A two-phased study of computerized health maintenance organization databases. Pediatrics 112:1039–1048
  • Vincenzo et al. 2014. Long-term efficacy and safety of human papillomavirus vaccination. International Journal of Women’s Health 6:999–1010.
Posted in Uncategorized | Tagged , , , , | 6 Comments