Response to Comments on My Last Post (Yikes This Can Go on Forever!)

Hello “Schwartz,”

Thanks for your comments. I’m going to try to quickly respond. Sorry that the sources are messily scattered throughout. I'm aiming to get this done before 6 a.m. Urology clinic.

Adina,Your method of debate is a refreshing change from the usual vitriol. Thank you.I have a couple of initial questions I'll get out of the way first:
“1) Do have experience in statistical analysis or epidemiology?”

I freely admit that I am not nearly an expert in the subjects of statistical analysis or study design. However, when discussing these topics, I hope you will focus on the content of my arguments, rather than on my degrees or qualifications (please note that I do believe degrees are important for some things). In other words, I don’t care if you graduated from clown college. If you make good points, I will afford them the same respect that I would to those of a statistician. As an aside, I have taken non-graduate-level courses in biostatistics, epidemiology, economics, and math. Additionally, I worked for a clinical research study (for only one year).

Did you read the details or any critical reviews of the epidemiology you broadly reference?”

I made sure to read the major components of every paper I referenced. In terms of critiques of the study design, I have read many thorough (unconvincing) rebuttals to a number of the studies I presented at http://www.generationrescue.org/autism/17-studies-against-vaccine-autism.htm and http://www.safeminds.org/research/commentary.html

There are some much better-reasoned review articles available, which address the drawbacks of various studies. For example, this article http://pediatrics.aappublications.org/cgi/reprint/115/1/200.pdf compares a number of epidemiological studies, addressing the purported link between autism and vaccination. They find that Heron’s study fulfills 5/8 epidemiological criteria, and the research of both Hviid and Verstraeten each fulfill 6/8 criteria. In other words, they are strong, but not flawless studies (there are extraordinarily few epidemiological studies about anything that would fulfill all 8/8 criteria). The Geiers’ study, the only one that found a link between vaccination and autism, fulfilled 0/10 criteria. The methodology used in this article review article appears quite sound.

(Parenthetically the third world application of Rota vaccine is relevant but so is the fact that it is currently recommended as a childhood vaccine in the US. The US recommendation is justified as a reduction in cost due to reduced hospitalization, not death. This is far more relevant to your readers when considering the risks.)

CDC: “The number of intussusception cases reported to date after RotaTeq administration does not exceed the number expected based on background rates of 18-43 per 100,000 per year for an unvaccinated population of children ages 6 to 35 weeks.”
http://www.mass.gov/Eeohhs2/docs/dph/cdc/immunization/alerts_rotavirus_vaccine_rotateq_intussusception_cdc_qa.rtf

"However, these claims have been investigated (Ten studies to show MMR doesn't cause autism. Six that show thimerosal doesn't cause autism, etc.)"
In order to support the notion that the claims of damage have been vaccinated, you would have to produce case studies or at least detailed follow-up for the affected group OR at least you have to illustrate that enough study of children with similar risk profiles were studied en masse. Neither you or Dr. Offit have provided them so right off the bat, the definitive statement that the topic has been studied is not justified in my opinion. In general, you will find that in the cases of vaccine damage very little documentation is available and there certainly is no large cumulative study of affected children. That alone should raise red flags in a system that mandates a medical intervention.

I’m not sure what you mean by similar risk profiles. Do you mean studying patients in a case-controlled study, rather than the current practice of conducting longitudinal cohort studies? (In other words, finding kids who are already known to have autism, and then finding out about their exposures, rather than recruiting kids with known exposures, and then following them to determine if they were later diagnosed with autism).

I’m not 100% sure what kind of study you’re referring to, but if it is a case-controlled study, then I feel that such a study would have little value. A case-controlled study, in this situation, is susceptible to information bias (Differential reporting of past exposure information based on disease status), Selection Bias (Controls and/or cases chosen in such a way that they are systematically more (or less) likely to be exposed than the population from which they were drawn), Confounding. (Failure to measure and adjust for all potentially confounding factors can lead to invalid associations), and Temporal Relationship Problems. (Can’t be certain the exposure preceded the onset of the disease). (Sources of bias copied from Epidemiology Lecture notes created by Dr. Christopher Haiman).

It shouldn't be a surprise that Hannah Poling's case was actually investigated in detail likely only because her father was a neurologist with both the power and wherewithal to ensure her case was studied in detail"In terms of the Vaccine Court, the fact that those girls got money from the Vaccine Compensation Fund is not particularly relevant, because the cases are not judged based on scientific evidence" “

I might have missed it, but I don't see anywhere they state that the decisions are not based on scientific evidence. Medical records and expert assessment were provided in this case, and certainly expert assessment qualifies as scientific evidence. Your definitive statement here is not justified.”

I’m having trouble with this argument. Do you think that if we went to court to resolve a dispute about evolution vs. creationism, then the findings of the court would be scientifically relevant, so long as the judges heard the testimony of experts? I am not sure if you are saying that courts ought to be recognized as legitimate arbiters of a scientific debate, but if so, I must disagree. Courts might invite scientists to testify about their research, but the court itself has no meaningful role in determining of the scientific merit of such testimonies. Because courts must make legal decisions, the public hopes that the courts will do their best to be objective and weigh already discovered evidence appropriately, as well as figure out which witnesses are most credible. Ultimately, however, the legitimate value of scientific findings is determined by the scientific community through the scholarly writing and peer review.

Additionally, I am pretty confident the percentage of “experts” who testified that vaccines cause autism was dramatically larger than the percentage of scientists, in the general population, who agree with that notion. Court cases are misleading, because disproportionate representation leads to an assumption that there is widespread debate about a given issue, when opposition might actually exist among only a small fringe group.

"A team of lawyers (called "special masters") with no medical background, rather than a judge, jury, scientists, or medical professionals, preside over the cases." The Poling case was conceded after a review by the HHS scientists. It never went to a hearing before the Special Masters. The HHS scientists examined the scientific evidence presented and determined that a hearing was not required, as they concluded the act of vaccination contributed to her brain damage (encephalitus) leading ultimately to a diagnosis of Autism. Your own statements don't match the facts and display the bias of Dr. Offit's opinion pieces (where it is difficult to evaluate the level of bias).

The only document relating the Poling case that is provided on the Vaccine Court's website is at http://www.uscfc.uscourts.gov/sites/default/files/CAMPBELL-SMITH.POLING041008.pdf
It states,
“The undersigned [special masters] directed respondent to file a status report after reviewing Dr. Zimmerman’s expert report that addressed respondent’s position regarding petitioners’ claim that Hannah’s seizure disorder was vaccine-related. Petitioners filed the expert report from Dr. Zimmerman after the status conference. Respondent’s review of Dr. Zimmerman’s expert report…On February 21, 2008, respondent filed a Supplemental Rule 4 Report addressing. The respondent stated that “[h]aving reviewed this additional evidence, medical personnel at the Division of Vaccine Injury Compensation, Department of Health and Human Services (DVIC)] now recommend compensation for Hannah’s seizure disorder as sequela of her vaccine-injury in accordance with 42 U.S.C. § 300aa-11(c)(1)(C)(ii).” Id. at 2. Based on respondent’s concession, a damages determination is now underway in this case.”

Please note that the report states that a single respondent, rather than a team of DHHS experts, was asked to review the case (using the files provided by Hannah’s doctor0. The document does not mention the identity of this respondent. He or she could be a member of ARI or Generation Rescue, for all I know. Then the special masters made the final decision. Your account of what happens does not correspond to the account written in the document (See page 2). However, there may additional documents that I cannot locate, so I will look at those, if you know of any.

You seem to make more factual errors about the case of Andrew Wakefield: "Many of Wakefield's co-writers testified in court that he had falsified data, and 10/13 authors retracted their names from portions of Wakefield's most significant study, stating "We wish to make it clear that in this paper no causal link was established between the MMR vaccine and autism, as the data were insufficient.""I am interested in which trial of Dr. Wakefield anyone testified that data was falsified? I suspect you're incorrectly referring to the Omnibus hearing (which you yourself stated was unscientific) which did not involve Dr. Wakefield at all. Allegations against him were made despite the fact he was not a participant in the hearing, and thus he had no opportunity to rebutt the accusations. It would be good for you to clear up this allegation of yours, because as it stands, your statement does not match the facts that I am aware of. Given the diligence of most of your post, you seem to be making quite a few errors around the legal and personal aspects. I can only assume that you have been swayed by the many biased writings of these events without actually understanding the details of the events that unfolded. Wakefield is currently undergoing a GMC fitness hearing in the UK but no transcripts have been published and no verdict delivered so your conclusions appear a bit premature.

You are correct. I apologize for the error, and have placed a correction it in my previous post. Chadwick had testified at the Omnibus Hearing, not at a trial related to Andrew Wakefield. (I confused this case with the hearing for professional misconduct that Wakefield is currently undergoing at the General Medical Council in the U.K.) The extraordinarily detailed testimony against Wakefield, by the primary researcher who had carried out the experiments is quite damning, nonetheless. Please see part of the transcript at http://breathspakids.blogspot.com/2007/06/patrick-holford-and-dr-andrew_28.html
There were other testimonies of many major problems in Wakefield’s research. For example, according to the testimony of Dr. Stephen Bustin, the now defunct diagnostic center that Wakefield used, Unigenetics, claimed to have “detected measles gene using a type of PCR that could detect only DNA." (Offit 171)” The measles virus doesn’t contain DNA, only RNA. Please read about the findings of the 2004 investigation of Wakefield’s laboratory at http://www.spiked-online.com/index.php?/site/article/3562/.

You also make two further misleading statements: First, your wording infers that the retraction was significant despite the fact they only retracted the interpretation not for scientific reasons, but public relational ones as they noted in the section of their statement that you ommited. Second, you infer that the lack of a causal finding between MMR and autism was a new revelation, when it was already explicitly stated in the original paper (discussion section page 641). Your selective quoting makes a big deal out of very little just like the misleading press releases on the topic.

Wakefield was very smart in trying to have it both ways: Always saying that nothing can be known for sure (so that no one can pin anything on him), but then making frightening statements about vaccines to the public. Wakefield et. al. were careful to write “We did not prove an association between measles, mumps, and rubella vaccine and the syndrome described. Virological studies are underway that may help to resolve this issue… We have identified a chronic enterocolitis in children that may be related to neuropsychiatric dysfunction. In most cases, onset of symptoms was after measles, mumps, and rubella immunization. Further investigations are needed to examine this syndrome and its possible relation to this vaccine.”

Every researcher is aware that no medical study can “prove” anything, so it is silly to write “we did not prove,” in a paper. They seem to imply that they have significant evidence, but that this evidence just doesn’t qualify as “proof.” If you believe this argument is just a petty focus on semantics, then please note that Wakefield used every public opportunity, following the article’s publication, to imply that his findings were indeed significant. In an interview following publication of the article, Wakefield says:

“It is our suspicion that there may well be [a link between MMR and what he refers to as “this syndrome”] but that is far from being a causal association that is proven beyond doubt… Again, this was very contentious and you would not get consensus from all members of the group on this, but that is my feeling, that the, the risk of this particular syndrome developing is related to the combined vaccine, the MMR, rather than the single vaccines…. Again, this was very contentious and you would not get consensus from all members of the group on this, but that is my feeling, that the, the risk of this particular syndrome developing is related to the combined vaccine, the MMR, rather than the single vaccines…Well, the interesting thing is that the damage, the behavioural or developmental change tends to occur quite soon after administration, and this is where, why parents or GPs or paediatricians have been able to make the link, the association with MMR.”

These sound like pretty ominous warnings for an article whose authors are extraordinarily cautious and humble about the significance of their findings. Wakefield even held a press conference about this issue, which prompted segments of the British public into mass hysteria. Vaccination rates fell dramatically, and the incidence of deaths from measles rose.

We may disagree about the significance of the retraction. There were multiple flaws in the study, with or without the retraction. Because you feel that I was omitting information, here is the portion of the paper, to which you refer:
“We wish to make it clear that in this paper no causal link was established between MMR vaccine and autism as the data were insufficient. However, the possibility of such a link was raised and consequent events have had major implications for public health. In view of this, we consider now is the appropriate time that we should together formally retract the interpretation placed upon these findings in the paper, according to precedent.”

"However, these claims have been investigated (Ten studies to show MMR doesn't cause autism. Six that show thimerosal doesn't cause autism, etc.)"
Going back to this quote, I want to continue driving into the studies you reference through Dr. Offit. First, I'm surprised at your use of such a definitive statement (doesn't cause autism) that really can't be supported by any scientific studies of epidemiology. The conclusion that doesn't cause Autism isn't even that interesting really. We want to know if it contributes to damage including autism (or other problems). The case of Hannah Poling is a great illustration that vaccination alone did not cause the damage, but it certainly appears to have triggered and exacerbated it.

I only addressed the autism issue. If you feel that autism is only of many disorders caused by vaccination, and that autism is the least “interesting” among them, then I guess it’s the anti-vaccinationists' lucky day, because I have no energy to refute the link between vaccines and dozens of different disorders. Yes, you are correct. The studies did not (and cannot) conclude “thimerosal doesn’t cause autism.” I intended to say that we have an extremely low level of uncertainty that vaccines do not cause autism. Please read my last post, in which I spoke about how one can never reject a proposed null hypothesis. See my Xenu example.

This is particularly pertinent with Thimerosal and this leads to my original questions at the top. If you read the details of these studies, you'll find that the two founding components are data on the prevalence of Autism and the estimated cumulative exposure of Thimerosal during infancy. I hope you've read critical reviews of the studies quoted by Offit, and the data they're based on. If you aren't aware, there is wide disagreement on the reliability of the Prevalence data, and there are significant known issues with the Danish Registry, the UK's GPRD, and the US databases over the period of time studied. I recommend at least finding and reading some of the statistical arguments against the use of prevalance numbers in these studies. As you should know, the role (and bias') of the statistician is critical in influencing the outcome because numerous assumptions and adjustments must be made in these studies which are already subject to a high risk of false positives and negatives. When working with difficult, inconsistent, or incomplete data (sometimes withheld from independent public analysis) the knowledge of conflict of interest is quite important.

Yes, I read the critical review on the Danish study of 500,000 children (see links above). Just yesterday, Italy released a study stating that there was no link between autism and vaccination. I don’t believe that the statisticians in all of the different countries, for in all of the different studies are part of some vast conspiracy. Of course, in any discipline (perhaps excepting math), there can be biases introduced in a study. However, these studies generally feature good epidemiological study design. I don’t understand how people who oppose vaccination can be so demanding about the standards of these studies, when they rely primarily on case reports or research from faulty databases (Such as the Geiers’ use of the VAERS database) There is no flawless epidemiological study that shows smoking causes lung cancer. However, I’m sure we can agree that the epidemiological data from around the world have added up to demonstrate clearly that smoking causes lung cancer. Now all of the studies are adding up, and they refute a vaccination/autism link.

The other issue is that these studies only look at the cumulative exposure over a long period of time. We know that the timing toxic exposure to virus' (CRS is a great example) or toxins in fetus' or infants can be quite significant in determine the outcome of damage. We also know that very low levels of heavy metal exposure can also have a significant effect neurological outcomes (ex. lead: http://www.ehponline.org/docs/2005/7688/abstract.html). None of these factors are even remotely considered in these studies, yet you and others including Offit make definitive statements like "doesn't cause Autism". They only studied cumulative dose to be the defining factor, an assumption that does not hold up all the time on it's own. Another aspect that is often ignored (and Contrary to Dr. Offit's published opinion in peer-reviewed journals) the concomitant application of vaccines (MMR and Thimerosal containing) are not actually required study for pre-regulatory approval and generally remain unstudied from a safety perspective. (http://www.cdc.gov/vaccinesafety/00_pdf/draft_agenda_recommendations_080404.pdf)

No matter when the vaccination occurs, if vaccinations indeed caused autism, we would find a difference between the vaccinated group and the control group. Even if we miss a lot of diagnoses by failing to follow everyone to late adulthood, the cases missed should be equal in both groups (or perhaps even greater in the control group- parents of kids who have vaccines are more hypervigilant, due to their concerns, and would likely notice autism symptoms earlier), so this wouldn't skew the results toward the null.

Thus, I don't think that following vaccinated people for a very long time is necessary, unless you believe that vaccines are only responsible for autism diagnoses that become apparent after childhood. In that case, we would see a difference in autism rates, only if we followed children all the way to adulthood. Sounds far-fetched.

Or perhaps you're implying that, even if mercury doesn't cause autism, it does cause neurological disorders that are only apparent adulthood? These new claims about non-autism related neurological disorders are a whole different topic. I don't know why there is a new idea about vaccines every day! Anyway, the absence of a link between vaccination and (non-autism) neurological disorders have been addressed in at least three studies (see Offit 247)or H. Frankel. "Report finds no link between thimerosal and neurodevelopmental disorders." The Lancet, Volume 358, Issue 9288, Pages 1163-1163

In reference to the notion that metals and other hazards are all around us, I don’t see how that would influence the results. Both vaccinated and unvaccinated kids would be exposed to such metals, so we would still need to see an increased incidence of autism for vaccinated kids. Your premise would require that vaccinated kids are subject to fewer “heavy metals” in their daily lives, and that this is the reason why the incidence of autism is equal in the vaccinated and non-vaccinated groups (i.e. the non-vaccinated group got all of their “heavy metal poisoning” from the environment, but vaccinated kids got their's from vaccines, so now the incidence is equal). This sounds farfetched, and is practically irrelevant, because it doesn't matter what would "cause" one's autism, if the chances of "acquring it" are the same, no matter the ubiquitous exposure.

Honestly, I wasn’t aware of the brand new suspicion (i.e. it’s not MMR that causes autism, nor thimerosal that causes autism, but only using the two of them together). After a claim is debunked, it doesn’t follow that a study must be made to investigate an “updated” conjecture. As I said in my last post, whenever new evidence comes along, the proposed causes always seem to keep change.

In terms of environmental exposure, data from the Minamata Bay mercury disaster show that no increased number of children developed autism, which even Aposhian was forced to admit at the omnibous hearing. (He then said that this is because mercury toxicity isn’t dose dependent, and that “This is an ancient form of quotation that until recently we taught in medical schools…we no longer believe that the dose makes the poison”) See Offit 166 and http://www.huffingtonpost.com/arthur-allen/in-autismvaccine-case-r_b_52408.html.

From my perspective that type of language is unsubstantiated by the reality of the limitations of the studies. The fact that Dr. Offit regularly makes public statements of this type (in addition with factual error that require correction) significantly hurts his credibility and I argue exposes his bias. Overstating the conclusions of scientific study, especially ones based on weak data is a reccuring issue. As for MMR, I suggest you read the Cochrane systematic review of MMR studies from 2005/2006: http://www.cochrane.org/reviews/en/ab004407.html.
After reviewing an extensive list of "definitive" MMR studies performed over several decades, they concluded that the study of both efficacy and safety was inadequate. If you read the details, (I suspect you'll find several of Dr. Offit's references in there) you'll see findings of serious methodological issues with seemingly "credible" study also heralded historically by regulatory bodies as "definitive". From this perspective you might understand why I remain skeptical of definitive claims issued by the likes of Dr. Offit.

You are correct that none of the 139 articles fulfilled all of Cochrane’s inclusion criteria. I don’t know how you think a better epidemiological study could be designed, which would adequately fulfill the exacting criteria. No study that is currently being proposed by anti-vaccinationists could possibly suffice, so we would be wasting money if conforming the Cochrane standards was their primary goal. Epidemiological studies have a lot of disadvantages, in comparison to controlled environments with drosophila insects or rats. The Cochrane study does recommend vaccination and says, “Exposure to MMR was unlikely to be associated with Crohn's disease, ulcerative colitis, autism or aseptic meningitis (mumps) (Jeryl-Lynn strain-containing MMR). We could not identify studies assessing the effectiveness of MMR that fulfilled our inclusion criteria even though the impact of mass immunization on the elimination of the diseases has been largely demonstrated… No credible evidence of an involvement of MMR with either autism or Crohn's disease was found.”

Overall, I think there is plenty of justification of continued study. None of these studies you referenced ever examined the long term effects of either MMR or Thimerosal exposure in any RCTs (or as close as we can get given ethical limitations). We know from recent history (HRT being a prime example) that large long term studies often reveal unexpected negative outcomes in treatments assumed and shown to be safe by previous regulatory studies.

There are lots of claims we haven’t studied. In a world of unlimited abundance, I'd say, sure let’s have continuous studies about everything. A million studies. Regardless, NIH-funded studies are currently underway to examine this issue, and many more are planned, so it is a mute point.

"As I wrote on my other post, learning about the experiences of my patients and their families will be crucial for my providing good care, but is not relevant to science-based medicine (excepting in the use of case studies, which have limited value, and which are generally only used to develop hypotheses)."
This is an interesting opinion which I would like to examine outside of the vaccine discussion. Expert opinion still plays a role in evidence based medicine (which is what I assume you meant when using the term "science based medicine"). Cumulative clinical experience forms the heart of expertise. I am concerned that you also dismiss the value of clinical experience in detecting problems. It is exactly the detection of patterns of anecdote that usually lead to further study of issues. The other important factor is that repeated follow-up into similar individual cases may reveal biological mechanisms which can (and should) also be used to form a hypothesis. The fact that disease reporting is mandated by law, yet pharmaceutical adverse event follow-up outside of research settings is voluntary are rife with error is a sad reflection on our priorities in my opinion.

You are incorrect that I dismiss the value of clinical experience in detecting problems. Gaining clinical skills in how to recognize signs and symptoms, in order to form a diagnosis is quite different than assuming to know the etiology of an insufficiently studied disease. A good doctor might develop the clinical skills to recognize autism, but this does not mean he has any knowledge what caused it. Recognizing this requires some humility on the part of the clinician. Yes, I’ve mentioned that case studies, or any other observations, can be used to form a hypothesis.

Again, thank you for bringing a refreshingly politeness to the discussion. I wish you success in your studies. My only suggestion is to drill into the details of the assumptions underlying the evidence you use to justify closing the door on further research. Given the significant challenges facing our newer generation of doctors, I think you should drill into the well documented issues (plenty of peer review studies on this topic) with industry funded peer-review studies, and into the details of the impact that bias can have on statistical analysis. You might find (like I did) that the conclusions you're quoting are based on a lot more assumptions and questionable source data than you would like. (the latter not the fault of the authors, but on the lack of interest/diligence by medical regulatory authorities).
Potential Bias: I have two healthy NT children, one partially vaccinated (older), one unvaccinated (younger)

Thank you “Schwartz.” If you respond to this, I’m going to have to let you have the last word. My adventures in autism blogging are putting a cramp my non-school-related life.

Take Care,
Adina