When this abstract came out I had hopes, vain hopes, that it wouldn’t get discussed much online. It’s perfectly reasonable to discuss issues surrounding conflicts of interest in vaccine safety research, but the author has a track record of less-than-excellent publications. Prof. DeLong wrote A positive association found between autism prevalence and childhood vaccination uptake across the U.S. population, which was discussed here at Left Brain/Right Brain, Neuroskeptic and The Biology Files (to name a few). Again, not an excellent study.
Here is the abstract of the new paper:
Conflicts of interest (COIs) cloud vaccine safety research. Sponsors of research have competing interests that may impede the objective study of vaccine side effects. Vaccine manufacturers, health officials, and medical journals may have financial and bureaucratic reasons for not wanting to acknowledge the risks of vaccines. Conversely, some advocacy groups may have legislative and financial reasons to sponsor research that finds risks in vaccines. Using the vaccine-autism debate as an illustration, this article details the conflicts of interest each of these groups faces, outlines the current state of vaccine safety research, and suggests remedies to address COIs. Minimizing COIs in vaccine safety research could reduce research bias and restore greater trust in the vaccine program.
The introduction to the paper starts with “How safe are vaccines?” This for a paper whose purported subject is “Conflicts of interest in vaccine safety research”.
I was going to put off reading the paper until I found this line in Orac’s article on it. Prof. DeLong, former board member of SafeMinds, wrote about “advocacy groups” (which includes SafeMinds, NVIC and the Autism Research Institute) thus:
While these organizations are not as well-staffed or well-funded as government agencies or vaccine manufacturers, their main task is to generate information to refute agency or industry claims.
We hear over and over how these groups have the focus to provide accurate information, to allow for “informed consent”. But here we find that they are focused on generating “..information to refute [government] agency or industry claims.”
Which begs the question: what if/when the information from the government or industry is correct? Prof. DeLong appears to be assuming that the information is always in need of refutation. Not exactly what I want for in an “advocacy group”.
Which brings us to another point: why are “advocacy groups” only those who are in the business of “refuting” government and industry information? What about advocacy groups such as the Sabin Vaccine Institute, the Bill and Melinda Gates Foundation, CHOP’s Vaccine Education Center, Every Child by Two, to name a few. I know many vaccine skeptics dismiss these groups as “front groups” for the government or industry, which is a particularly hard argument to make for the Gates Foundation.
Prof. DeLong makes a passing attempt to include conflicts of interest for the advocacy groups she does include. I say “passing attempt” because while much of the paper is devoted to methods whereby public health decisions should be made without influence of COI’s, she only makes these suggestions for people from industry or government. For example, she says that we should “manage the influence of vaccine manufacturers on medical journals”. No mention of, say, Andrew Wakefield serving on the editorial board of “Autism Insights” when a paper supporting him was published by one of his employees. Or for less severe questionable activities such as people funded by “advocacy groups” serving as referees or editors of journals. She also has a section on how we should “prohibit agencies that promote vaccines from overseeing vaccine safety”. But no prohibition for members of “advocacy groups” which often include, as in her case, people who have filed claims in the vaccine court and are, thus, facing a significant financial COI.
Much of the paper is basically a Trojan Horse. Under the title of “Conflicts of interest in vaccine safety research”, the paper spends considerable space discussing vaccine safety (remember that first sentence of the paper?). We are left with the premise that “trust in vaccine safety is low”. Those two paragraphs are a good example of selective quoting of statistics. For example, we are told that 77% of people surveyed have at least one concern about vaccine safety. Count me in. Am I an example of “low trust” in vaccine safety? “Low” is a relative word. A counter example to those given in the paper: vaccine uptake is still about 90% (and more) for most childhood vaccines. People watching these numbers are concerned with fractions of a percent drop. “Low” is “low enough to potentially endanger public health by allowing for outbreaks”. That’s my definition of low. Not the implied, most people have low trust.
The paper is about 25 pages long, in published format. That’s a big effort. It is, in the end, largely a very long blog post. If you’ve been reading the online discussions for the past few years, you’ve likely encountered much of what she has to say. It ends up more of an opinion piece than a scientific paper.