I sensed a disturbance in the antivaccine (i.e, the dark) side of the Force yesterday. No matter where I wandered online and on social media, I kept running into a new article, an article by Neil Z. Miller about vaccines. For example, the merry band of antivaccine propagandists over at Age of Autism seem to like Miller’s article very, very much. So did the vaccine truthers over at—where else?—VacTruth.org. I kept seeing it on Facebook and Twitter, too.
Even though the current vaccine schedule is safe and effective as well as evidence-based and the claim that we give too many vaccines too soon is an antivaccine myth, that doesn’t stop Miller from claiming otherwise in an article just published in the Journal of American Physicians and Surgeons (JPANDS) entitled Combining Childhood Vaccines at One Visit Is Not Safe. Now, the first thing I think whenever I see an article published in JPANDS is that it must be so horrible that no reputable journal would touch it with a ten foot cattle prod. After all, JPANDS is the house organ of a group, the Association of American Physicians and Scientists (AAPS), best known for its extreme right wing politics, its members’ belief that they are brave maverick doctors who don’t follow the “herd,” and its utter disdain for any evidence that conflicts with its ideology.
As hard as it is for me to believe, I first discovered the AAPS over ten years ago and was astonished at its embrace of antivaccine views, HIV/AIDS denialism, and other forms of pseudomedicine and medical conspiracy theories. Examples abound. Basically, the AAPS is known for being against vaccine mandates, against Medicare (calling it “unconstitutional”), and against any form of regulation of health care by government. It’s called public health programs “tyranny.” The AAPS has also published bad papers claiming to find that abortion causes breast cancer, has promoted the vile idea that shaken baby syndrome is a misdiagnosis for “vaccine injury,” supported HIV/AIDS denialism, and (of course!) done what all crank medical organizations like to do, attack evidence- and science-based medicine as placing unacceptable limits on physician autonomy. Perhaps my favorite example of AAPS crankery is when it published a blog post (now removed, no doubt in embarrassment) claiming that then-candidate Barack Obama was possibly “deliberately using the techniques of neurolinguistic programming (NLP), a covert form of hypnosis.” The AAPS doesn’t even limit itself to medicine in that it’s also published papers attacking anthropogenic global warming, as though physicians had the necessary expertise to judge the science in that field. Truly, the crank magnetism and arrogance of the AAPS know no bounds. If you don’t believe me, consider this. JPANDS has published articles by the father-son team of antivaccine “scientists,” Mark and David Geier.
We’ve also met Neil Z. Miller before. Well do I remember an article by him and co-author Gary S. Goldman published five years ago that resurfaces from time to time. Basically, it was an “analysis” purporting to show that infant mortality correlates with the cumulative number of doses of vaccine in the childhood vaccination schedule. Let’s just say Miller and Goldstein’s rationale, methods, and analysis were rather suspect. No, wait. Strike that. Let’s just say it was a giant, drippy, stinky turd of an article—and amateurish to bood—whose conclusions were not at all supported by the data or analysis. He and Goldman also teamed up for an equally inept attempt to show that more vaccination correlates with more hospitalizations and deaths. They failed.
So, given how much antivaccinationists like this latest article by Miller, combined with the abysmal (and well-deserved) reputation of JPANDS and Miller’s track record, I was not optimistic that this wouldn’t be more of the same. My pessimism was justified. The main difference is that this latest article is not an attempt at an original investigation, but rather an attempt at a systematic review. Unfortunately, in a systematic review, one is supposed to review the existing literature as comprehensively as one can, discussing its strengths and weakness, not cherry pick a few studies to misinterpret and cite one’s own “studies” far more prominently than they deserve. Let’s take a look.
Although health authorities including the Centers for Disease Control and Prevention (CDC) claim that childhood vaccines are safe and recommend combining multiple vaccines during one visit, a review of data from the Vaccine Adverse Event Reporting System (VAERS) shows a dose-dependent association between the number of vaccines administered simultaneously and the likelihood of hospitalization or death for an adverse reaction. Additionally, younger age at the time of the adverse reaction is associated with a higher risk of hospitalization or death.
Which is basically what Miller and Goldman tried, and failed, to show four years ago.
The first mistake Miller makes is the same mistake he’s made time and time again: Dumpster diving the VAERS database. Why do I call it “dumpster diving”? Easy. The VAERS (Vaccine Adverse Event Reporting System) is a problematic database, whose contents are useful for a very limited range of analyses. However, it is also a public database that can be freely downloaded by anyone; so naturally it attracts antivaccine “scientists” like catnip attracts cats. As longtime readers of this blog and vaccine science advocates know, VAERS is a passive reporting system. Anyone can report a suspected vaccine injury to the database. You don’t have to be a medical professional to do it, and there doesn’t even have to be a plausible relationship between the vaccine and the suspected adverse reaction. Indeed, it’s long been known that vaccine litigation distorts the VAERS database, with plaintiffs’ attorneys encouraging their clients to report their children’s cases to VAERS. Consequently, VAERS cannot be relied upon for anything close to the true incidence of specific adverse reactions. If you don’t believe me, let me just mention once again that reports of vaccines turning people into The Incredible Hulk and Wonder Woman have been successfully entered into VAERS.
Why would the government want to set up a system like this? Simple. It’s an early warning system. Even though the database can be distorted and even though its contents are not a reliable estimate of the incidence of vaccine injuries, it’s still useful in that when scientists notice an uptick in the reporting of an adverse event due to a vaccine in VAERS it’s an impetus to study it more deeply and see if the increase is real.
Of course, at some level, Miller must know this, as he spends much of the first page trying to convince the reader that VAERS is valuable and citing CDC studies using VAERS data. (Of course, VAERS is valuable and useful, just not in the way Miller thinks it is.) First, Miller tries to convince us that adverse events are underreported in VAERS:
Since 1990, the VAERS database has received more than 500,000 reports of suspected adverse reactions to vaccines. Although this represents a large number of people who may have been hurt by vaccines, under-reporting is a known limitation of passive surveillance systems. This means that VAERS only captures a small fraction of actual adverse events. In fact, shortly after VAERS was established, a large vaccine manufacturer, Connaught Laboratories, estimated “about a 50-fold under-reporting of adverse events in the passive reporting system.”3 Perhaps 98% of all adverse reactions to vaccines are not included in the VAERS database, and up to 25 million U.S. citizens could have been adversely affected by vaccines in the past 25 years. This well-known disadvantage of a passive reporting system, as opposed to an active surveillance system in which medical workers are trained to systematically collect all cases of suspected adverse vaccine reactions, is rarely acknowledged by health authorities when vaccine safety is discussed.
A better way of looking at this would be that certain kinds of adverse events are underreported and certain others are likely to be massively overreported, thanks to the way lawyers petitioning the National Vaccine Injury Compensation Program insist that their clients enter their suspected adverse reaction, no matter how implausible. Miller cites the Institute of Medicine’s report on vaccines to bolster his view, but this is what the IOM actually wrote:
Care must be taken in interpreting information from passive surveillance systems. The extent of underreporting cannot be known. Duplicate reports of the same event for the same patient are common and are not always easy to detect, making totals questionable. Medical information provided on reporting forms is often incomplete. In general, passive surveillance systems are useful in flagging potential problems and suggesting hypotheses. See Chapters 2, 10, and 11 for further discussion.
Which is exactly what I just explained above. The IOM also noted:
From a comparison of spontaneous reports with postmarketing surveillance data, the company estimates about a 50-fold underreporting of adverse events in the passive reporting system. The distribution of types of events, however, was found to be approximately the same; in both cases, the majority of reported events were local reactions or fever. The company has seen a marked decrease in adverse event reports since the inception of VAERS late in 1991, because physicians are now requested to send reports directly to the VAERS contractor.
Miller then cites three studies by the CDC using VAERS: a 2015 study looking at the MMR vaccine in adults; a 2014 study looking at live attenuated influenza vaccine (LAIV3); and a 2013 study examining intussusception after a rotavirus vaccine. The first study is described thusly:
Although 5% of reports were serious, including several deaths, CDC researchers concluded that “in our review of VAERS data, we did not detect any new or unexpected safety concerns for MMR vaccination in adults.”
Note the implication that the MMR vaccine is very dangerous. What Miller neglected to mention (almost as if he didn’t read anything more than the abstract) is that the medical records of patients suffering serious events were reviewed. Of the seven deaths, two were due to cardiovascular disease, one due to a drug overdose, one due to chronic preexisting myocarditis, one due to pulmonary embolus, one due to an arrhythmia. The last one was a patient who had had a renal transplant and was on immunosuppression who died of disseminated varicella. In other words, none of them appear to have been due to the MMR vaccine.
Next, Miller characterizes the LAIV3 study thusly:
Although 8.9% of reports were classified as serious (e.g., cardiovascular events, neurological debilities, and fatalities) CDC researchers concluded that “review of VAERS reports are reassuring, the only unexpected safety concern for LAIV3 identified was a higher than expected number of Guillain-Barré syndrome reports in the Department of Defense population, which is being investigated [sic].”
Note again how Miller tries to paint the LAIV3 as dangerous. This paper is actually a bit more nuanced. For one thing, the rate of serious adverse events reported was generally under 2 per 100,000 doses, which one would note if one reads the whole paper. Be that as it may, this paper only found a possible increased incidence of Guillain-Barré syndrome in the Department of Defense population, which is being investigated—which is exactly what VAERS is for: Early warning and hypothesis generation.
Finally, the last paper cited looked at intussusception after rotavirus. This one was interesting in that it looked at the number of days after vaccination that reports of intussusception were made and found a clustering between 3-6 days, which suggests that there might be a causal relationship based on what is known about intussusception and rotavirus. Also:
In summary, after distribution of 47 million doses of RV5 in the United States, we observed a persistent clustering of intussusception events during days 3 to 6 after the first-dose vaccination. When we combined all 3 doses of RV5, we estimated a small overall excess risk of ∼0.79 intussusception event for every 100 000 vaccinated infants. This level of increased risk in the United States would translate to 33 excess annual intussusception events after rotavirus vaccination with the coverage expected for a fully mature rotavirus vaccine program. This is substantially lower than the number of diarrhea hospitalizations prevented annually (∼40 000) since rotavirus vaccine introduction.
Emphasis mine. Risks versus benefits. Risks versus benefits.
Perhaps the most hilarious part of Miller’s paper is this:
These studies and others confirm that CDC considers VAERS an important post-marketing vaccine safety surveillance tool. Therefore, nobody should be swayed into believing the VAERS database does not contain immensely valuable raw data to be used by independent researchers conducting studies that evaluate the safety of U.S. mandated vaccines. For example, Mark Geier, M.D., Ph.D., independent researcher and former professional staff member at the National Institutes of Health (NIH), published several studies utilizing the VAERS database showing that vaccines containing thimerosal (mercury) significantly increase the odds of developing neurological disorders, including autism.
That’s right. Miller actually cited the Geiers as though they were real researchers and not antivaccine advocates pretending to do research. There’s a reason why I refer to anything the Geiers do with VAERS as “dumpster diving.” In fact, numerous studies have failed to find an association between thimerosal-containing vaccines and autism or other neurological conditions.
All of this, of course, was just a lead up to Miller regurgitating his 2012 “study.” I deconstructed that incompetently performed study in detail back when it was originally published; so I don’t really feel the need to go into too much detail, given that you can simply click on this link if you want to go into the truly Oracian-length Insolent deconstruction. there were some truly hilarious abuses of statistics, a seeming fetish for trying to fit Miller’s data to a straight line in the absence of a plausible rationale why it should fit to a straight line, a failure to control for historical trends in infant mortality, and a failure to control for obvious potential confounders, such as birth cohort. It was truly a crappy “study.”
Of course, to Miller, his Truth is being suppressed by The Man:
Our study showed that infants who receive several vaccines concurrently, as recommended by CDC, are significantly more likely to be hospitalized or die when compared with infants who receive fewer vaccines simultaneously. It also showed that reported adverse effects were more likely to lead to hospitalization or death in younger infants.
Why is The Truth being suppressed by The Man? I think you know. It’s big pharma, of course:
These findings are so troubling that we expected major media outlets in America to sound an alarm, calling for an immediate reevaluation of current preventive health care practices. But 4 years after publication of our study, this has not happened. Could it be because, according to Robert Kennedy, Jr., about 70% of advertising revenue on network news comes from drug companies? In fact, the president of a network news division admitted that he would fire a host who brought on a guest that led to loss of a pharmaceutical account. That may be why the mainstream media won’t give equal time to stories about problems with vaccine safety.
I wonder if Miller knows just how pathetic it is to whine about how a paper he wrote four years ago failed to catch the attention of the press or the scientific community when it was originally published. I wonder if he knows just how much more pathetic it is that he published his complaint in JPANDS, one of the crankiest of crank journals. Obviously, it never occurred to Miller that maybe—just maybe—the reason that his “study” wasn’t picked up by major media is because it was a stinking heap of crap. Unfortunately, so far, it’s worked because memories are short and antivaccine activists don’t care about any evidence that doesn’t support their ideology. Hopefully, this post will serve as a reminder of why Miller’s study is not evidence that the current vaccine schedule is unsafe.