I don’t have many “rules” per se about blogging, but one informal rule that I do live by is that I never blog about a study if all I can access is the abstract. In general, I insist on having the complete study before I will blog it, because to me the abstract isn’t enough. Basically, if I’m going to blog a study, I generally want to do it right and be able to read the whole paper, because that’s the only way to properly analyze a paper. I find this rule particularly important when analyzing the latest bit of antivaccine pseudoscience, especially because most antivaccine activists don’t go past the abstract and even more because the abstract quite often doesn’t tell the whole story. Still, this is a rule that I have broken on a handful of occasions over the last 12 years, although I generally try not to break it without a compelling reason. The small number of times that I’ve blogged about just an abstract over such a long period of time attests to that.
It looks as though this is going to be one of those times.
This time around, as always, I have what I think are some very good reasons, not so much because the antivaccine study that I’m about to blog is anything particularly interesting, informative, or even more awful than the usual run-of-the-mill antivaccine study. Heck, it isn’t even a study. It’s just a survey that reminds me of the second coming of the German homeopath survey that tried to link vaccines to autism using such a horribly designed survey instrument. Indeed, crappy surveys with biased questions are a favorite tactic of antivaccine activists, dating back to J.B. Handley’s fun with phone surveys in 2007 and continuing on right up to this day. Yet, the antivaccine crankosphere has over the last week or so been hailing it as the definitive “vaxed versus unvaxed” study that they’ve always wanted, and—surprise! surprise!—it “shows” a higher rate of autism, allergies, ADHD, and learning disorders in vaccinated children. As a result, more of my readers than I can remember have been sending me links to the study. But there was a weird thing about it, something I’ve never seen before. There was just an abstract. There was no link to an actual manuscript.
I was almost at the point where I was going to write to the authors when I saw this article on Retraction Watch:
A study linking vaccines to autism and other neurological problems has been removed by a Frontiers journal after receiving heavy criticism since it was accepted last week.
The abstract — published online in Frontiers in Public Health after being accepted November 21 — reported findings from anonymous online questionnaires completed by 415 mothers of home-schooled children 6-12 years old. Nearly 40 percent of children had not been vaccinated, and those that had were three times more likely to be diagnosed with neurodevelopmental disorders such as autism, the study found.
After receiving criticism on Twitter, Frontiers released a public statement, noting that the study was only “provisionally accepted but not published,” and is being re-reviewed. When asked for a comment, a Frontiers spokesperson referred us to the statement.
This article was provisionally accepted but not published. In response to concerns raised, we have reopened its review. @70Hertz
— Frontiers (@FrontiersIn) November 28, 2016
Wow. That’s a first! I’ve never seen an editor giving a statement on Twitter as its only explanation for retracting an article—retracting an abstract, actually. For now, the abstract is still available via Google Cache. To preserve it for posterity, I quote it here:
Vaccination and Health Outcomes: A Survey of 6- to 12-year-old Vaccinated and Unvaccinated Children based on Mothers’ Reports
Anthony R. Mawson1*, Brian D. Ray2, Azad R. Bhuiyan3 and Binu Jacob4
1Epidemiology and Biostatistics, School of Public Health (Initiative), Jackson State University, USA
2National Home Education Research Institute, USA
3Epidemiology and Biostatistics, School of Public Health (Initiative), USA
4Former Graduate Student, Jackson State University, School of Public Health (Initiative), USA
Background: Vaccinations have prevented millions of infectious illnesses, hospitalizations and deaths among US children. Yet the long-term health outcomes of the routine vaccination program remain unknown. Studies have been recommended by the Institute of Medicine to address this question.
Specific Aims: To compare vaccinated and unvaccinated children on a broad range of health outcomes, and to determine whether an association found between vaccination and neurodevelopmental disorders (NDD), if any, remains significant after adjustment for other measured factors.
Design: A cross-sectional survey of mothers of children educated at home.
Methods: Homeschool organizations in four states (Florida, Louisiana, Mississippi, and Oregon) were asked to forward an email to their members, requesting mothers to complete an anonymous online questionnaire on the vaccination status and health outcomes of their biological children ages 6 to 12.
Results: A total of 415 mothers provided data on 666 children, of which 261 (39%) were unvaccinated. Vaccinated children were significantly less likely than the unvaccinated to have been diagnosed with chickenpox and pertussis, but significantly more likely to have been diagnosed with pneumonia, otitis media, allergies and NDDs (defined as Autism Spectrum Disorder, Attention Deficit Hyperactivity Disorder, and/or a learning disability). After adjustment, the factors that remained significantly associated with NDD were vaccination (OR 3.1, 95% CI: 1.4, 6.8), male gender (OR 2.3, 95% CI: 1.2, 4.3), and preterm birth (OR 5.0, 95% CI: 2.3, 11.6). In a final adjusted model, vaccination but not preterm birth remained associated with NDD, while the interaction of preterm birth and vaccination was associated with a 6.6-fold increased odds of NDD (95% CI: 2.8, 15.5).
Conclusions: In this study based on mothers’ reports, the vaccinated had a higher rate of allergies and NDD than the unvaccinated. Vaccination, but not preterm birth, remained significantly associated with NDD after controlling for other factors. However, preterm birth combined with vaccination was associated with an apparent synergistic increase in the odds of NDD. Further research involving larger, independent samples is needed to verify and understand these unexpected findings in order to optimize the impact of vaccines on children’s health.
Keywords: Acute diseases; Chronic diseases; Epidemiology; Evaluation; Health policy; Immunization; Neurodevelopmental disorders; Vaccination, Acute diseases, chronic diseases, Epidemiology, Evaluation, Health Policy, Immunization, Neurodevelopmental disorders, Vaccination
Citation: Mawson AR, Ray BD, Bhuiyan AR and Jacob B (2016). Vaccination and Health Outcomes: A Survey of 6- to 12-year-old Vaccinated and Unvaccinated Children based on Mothers’ Reports. Front. Public Health 4:270. doi: 10.3389/fpubh.2016.00270
Received: 17 Sep 2016; Accepted: 21 Nov 2016.
Amit Agrawal, Gandhi Medical College, India
Kelly Hsieh, University of Illinois at Chicago, USA
Linda Mullin Elkins, Life University, USA
Before I move on to discuss the provenance of this article, as well sa the kerfuffle its publication provoked, let’s take a look at what a piece of crap this survey is. It’s pretty obvious even from the abstract. First, this survey questioned 415 mothers of 666 children educated at home. Not only is that not a representative sample, given that all the children are home-schooled, it’s not even a very big sample. Remember when I discussed the statistical issues in doing even an epidemiological “vaxed versus unvaxed” study? To find any statistically significant, much less clinically significant differences in health outcomes between vaccinated and unvaccinated children would require huge numbers. I’m half tempted to make jokes about the number of children being 666 (hey, it’s me), but, jokes or no jokes, this is not a large sample.
Also, arents who choose to home school are not like your average parents. There are likely to be a lot of confounding factors that go along with home schooling, including the association between home schooling and antivaccine views. This association shows up in this very survey in that it reports that 39% of the children in the survey were completely unvaccinated. This is not representative of the general population, by any stretch of the imagination, where in general the number of totally unvaccinated children number in the low single digits. Add to that the likelihood of selective memory and reporting, and the likelihood of this survey providing useful information is vanishingly small. Of course, surveys are not the best means of gathering health data. Yes, I know. The NIH does surveys. I’ve even discussed one of them, specifically in relationship to how much “complementary and alternative medicine” (CAM) people use. However, while such surveys can be useful for assessing the sorts of treatments people partake in, they’re not quite as useful for assessing whether there are correlations between health practices (e.g., vaccination) and health outcomes (e.g., autism and ADHD).
There’s another interesting wrinkle to this paper, and that’s who peer reviewed it. One of the peer reviewers was Linda Mullin Elkins, a chiropractor at Life University, a “university” that portrays itself as:
We are at the forefront of the vitalistic health revolution by offering studies within the fields of Chiropractic, Functional Kinesiology, Vitalistic Nutrition, Positive Psychology, Functional Neurology and Positive Business, using entrepreneurship for social change.
“Vitalistic nutrition”? “Functional kinesiology”? “Vitalistic health revolution”? Yes, there’s some serious, serious woo there. Elkins is not a qualified peer reviewer for a paper like this—or for any peer reviewer. Then there is the issue of the journal itself. Frontiers in Public Health is published by Frontiers Media, which is on Beall’s list of predatory open access publishers. Such predatory open access journals follow a business model that involves charging publication fees to authors to publish just about anything.
Yes, basically, this paper is crap, so much so that even a predatory open access publisher pulled it, at least temporarily.
Among my interactions on social media, people kept pointing out to me that Anthony Mawson is a thrall of Andrew Wakefield, that he’s bought into the antivaccine Kool Aid. I responded that I really wasn’t familiar with him at all. It turns out that I was mistaken. I was. It’s just that I didn’t remember. However, all it took was to type his name into the trusty search box of this blog to find a post from four years ago that takes notice of our dear, dear Dr. Mawson. In fact, it turns out that I took notice of efforts by antivaccinationists to fund this very study to the tune of $500,000. Specifically, the antivaccine crank blog known as Age of Autism was soliciting funds to do this very study. Indeed, with the advantage of four years of perspective, I have a hard time understanding how doing such a survey could cost a whole half a million dollars. On the other hand, I think it’s worth citing my words from four years ago:
It all sounds rather innocuous, but looking deeper, I find that this “study” is not much of a study at all. In fact, it’s just an Internet survey, and not even a particularly informative survey. Why it will cost $500,000 to complete, I have no idea. It sures seems like a lot of green for a relatively easy study. It’s not as though a bunch of people to interview hundreds or thousands of subjects are needed. You can even look at it yourself, as one can find the survey here and here. Its principal investigator is Anthony R. Mawson, M.A., Dr.P.H.. That name sounded familiar to me, and it didn’t take much Googling before it came to me.
It turns out that Dr. Mawson is a vocal supporter of—surprise! surprise!—Andrew Wakefield. For instance he wrote to a blogger telling him that he disagreed with Prof. Trisha Greenhalgh’s critical analysis of the Wakefield et al. study that was published in The Lancet in 1998 in which she characterized it as seriously flawed. Dr. Mawson even went so far as to say that the paper is “excellent” and “a superb case study that will join the ranks of other famous case studies, such as the link between rubella infection and congenital rubella syndrome (Gregg 1941) and between exposure to thalidomide and embryopathy (McBride 1956),” concluding:
The paper, once understood in this light, as case series analysis, is truly remarkable, well written and brilliantly documented. It concluded by stating the hypothesis, based on parents’ reports, that the children’s’ signs and symptoms were temporally connected to MMR vaccination. Subsequent studies may not have substantiated the hypothesis; but that does not detract from or invalidate the merits of the paper as a case series and as, essentially, a hypothesis paper.
So, to summarize, what do we have here? We have a fundraising project by the antivaccine organization Generation Rescue funneled right to Anthony Mawson. Then we have a “study” (really a questionable survey) sold as the definitive “vaxed versus unvaxed” study.This study is, of course, a steaming, stinking pile of fetid dingo’s kidneys, which leads to its abstract being retracted from a low rent open access journal from a predatory publisher.
In other words, when it comes to antivaccine “science,” it’s the same as it ever was.