The esteemed British medical journal, the Lancet, has gotten a black eye in the news lately for articles that were poorly screened and were biased for political/personal reasons.

The first one is the 1998 study linking MMR vaccine with Autism.

This study has had devastating effects on three fronts: One, the terrible guilt by parents who are convinced that their child’s autism was their fault because they allowed them to get the vaccine, and the second effect are the children made sick, brain damaged, or even dead because parents refused to let their children get the vaccine, and their children developed measles.

The third effect? Because parents of autistic children were so focused on the MMR link, they failed to press for research into other things that can cause autism or autistic like behavior, from searching for treatments for Fragile X syndrome to checking out lead levels in Washington DC water supplies.

So why would a physician do such a thing? Was it accidental? Inaccurate studies are frequently biased because the data base is limited, or because the disease is under or overdiagnosed. For example, if you study only people diagnosed in Hospital Specialty Clinics, you miss the mild cases that the GP’s either miss or figure are too mild to get referred to the clinic.

Or the numbers in the study might be too small to make a conclusion. Small studies may accidentally include a “cluster” where  the disease is more common (for example, a “cluster” in autism might include several families distantly related who carry the Fragile X Syndrome; if the study was done in another city/neighborhood, this cluster might be missing).

However, major medical journals are supposed to have “peer review” of the studies, so that this information is available and scrutinized, so such things don’t occur.

But in this case, more than bad statistics was involved. From the UKTimes:

The research was published in February 1998 in an article in The Lancet medical journal. It claimed that the families of eight out of 12 children attending a routine clinic at the hospital had blamed MMR for their autism, and said that problems came on within days of the jab. The team also claimed to have discovered a new inflammatory bowel disease underlying the children’s conditions.

The UKTimes then goes on to show the problems: The diagnosis didn’t agree with the GP and hospital findings,  there was no evidence of bowel disease, at least one of the children had autistic symptoms prior to the MMR shot, and the study was based on only 12—yes 12– children, making it completely worthless since the number is too small to draw a conclusion.

But the study made major headlines, with the obvious results that fewer kids got the shots, so ended up with measles. Again from the UKTimes:

After its publication, rates of inoculation fell from 92% to below 80%. Populations acquire “herd immunity” from measles when more than 95% of people have been vaccinated.

Last week official figures showed that 1,348 confirmed cases of measles in England and Wales were reported last year, compared with 56 in 1998. Two children have died of the disease.

A similar problem was found in the 1980’s, when whooping cough vaccine was blamed for brain damage. In that case, there was a small but definite statistical increase of problems after the older vaccine (estimated to be 1 out of 140,000 injections); even though the older pertussis vaccine was replaced by an acellular vaccine in the early 1990’s, there continues to be outbreaks of whooping cough due because parents don’t want their children to get these immunizations.

One of the “side effects” of the various anti vaccine hysteria is not just that well nourished children who rarely die of the disease don’t get vaccinated, it is that the rumors get magnified in the third world.

There have been several stories with mocking tones about imams in Nigeria and elsewhere who forbad children from getting Polio vaccine, for example. Yet these stories rarely mention the rumor comes directly from the UK, where a book called “The River” claimed HIV came from HIV contamination in early polio vaccines that were given in central Africa about the time HIV started devastating the west.

The rumor turned out to be nonsense: Blood tests showed HIV present in Africa long before the polio vaccine was given, and the spread from local outbreaks was probably due to the building of a road across central Africa, that allowed truckers to spread it to local sex workers and then into the general population.

The rumor’s kernel of truth was because one of the early injected polio vaccine was inadvertently contaminated with a monkey virus; but the vaccine was later proven not to contain the cells carrying the virus, and of course later vaccines did not contain the virus at all, including the oral vaccine.

But the rumor that polio vaccine caused HIV, or caused women to be sterile started to spread into those with little medical background, and the imams in Nigerian Mosques started warning their people not to get their children vaccinated.

As a result of the much repeated rumor, there was an outbreak of polio in Africa that  paralyzed several hundred children, including unvaccinated children who “caught” the disease via children receiving the vaccine (a rare side effect, where the child receiving the vaccine containing weakened polio virus gets immunity but secretes a mutated stronger polio virus that can be caught by others who never got the polio vaccine…this is rare, but it is the reason most western countries went back to polio shots).

The dirty little secret is that when one talks about vaccines, one is playing the odds.

The chances of harm from vaccines are much, much lower than the chances of harm from the disease itself. But how many lay people have seen children choking from whooping cough, or dying from measles, or paralyzed by polio? Even many of today’s physicians have not seen cases of these dread diseases, unless they are over age 60, or have worked in third world countries like myself.

But there is a more disturbing problem in all of this.

Follow the money.

People who write books like “The River” are of course aiming at publicity to sell more books.  Perhaps he was too dumb to realize the possible problems of such a rumor.

But the NYTimes reports a more disturbing rumor about the author of the Lancet study:

The study came under fierce criticism last month when the editor of the Lancet said that the lead author of the report, Dr. Andrew Wakefield had failed to reveal that he had a conflict of interest when he conducted the research. At the time, the journal editors said, Dr. Wakefield was also gathering information for lawyers representing parents who suspected their children had developed autism because of the vaccine.

The author’s supporters deny this, saying his work with the lawyers came after he did the study.

But again it comes down to peer review.

Most physicians are aware of “clusters” which may mean something, or may be just a statistical coincidence. Often such data is published, as an anecdote or an observation, without drawing conclusions. In this case, the press who picked up the study as proof bears much of the blame in the fiasco.

But the reports are coming at a difficult time for the Lancet, who are also reeling under the scandal that their “Iraq Casualty Study report” article, published to influence the 2006 American election, also had major problems in methodology that cannot be resolved, and because the author refuses to release the data on how the study was done.

AAPOR’s standards committee chair, Mary E. Losch, said the association, acting on a member’s complaint, had formally requested from Burnham “basic information about his survey, including, for example, the wording of questions he used, instructions and explanations that were provided to respondents, and a summary of the outcomes for all households selected as potential participants in the survey.”

Without the author releasing his research data and collection methods data, how can one judge if the data was accurate?

There is a major question on the data, which shows a death rate much higher than any other study. Sources of accidental bias could be the choice of those living in areas with higher casualties, for example.

As I noted in an earlier report, one of the most blatant “clues” that the data was suspicious was not the death count from gunshot wounds, which after all might have been injured insurgents who later died without the authorities noticing the deaths, but the blatantly high number of deaths from “car bombs”, which do get noted in the media and therefore counted by other watchdog groups.

Was the study tainted by a “cluster” reporting error?

An example of this would be to take the death rate of inner cities in the US and use the statistic as if the statistic was valid for all areas.

In this case, the Iraqi doctor in charge of the data collecting had worked for Saddam’s Health ministry; I presume he was honest and had no reason to tilt the data. However, this suggests he is Sunni, and could have chosen his researcher data collectors from the Sunni population, which has suffered more casualties than the Sunni and Kurdish population, but used the numbers as if they were typical of all areas of Iraq.

Another bias would be that only those families who had suffered deaths could have been willing to talk to those collecting the data. Then there is the “coffee klatch” problem: it would take time for an interviewer to greet a strange family, talk to them long enough to gain their trust, and then wait while they found the death certificates. If one calculates the short time for each interview, it comes to one half an hour per interview. This either means selection bias to those already known to the research gatherers, or (dare one suggest it?) that  some of the data might have been made up by researchers who were in a hurry.

Like the MMR study, these issues should have been detected by peer review, and either corrected before publication, or the articles could have been published as anecdotal data, with the caveat that the data might not be accurate.

So in both these cases, the Lancet’s editors are the ones who are to blame, and the mainstream media share the blame for spreading the rumor without asking an expert to analyze the reports before spreading the stories.

——————–

Nancy Reyes is a retired physician living in the rural Philippines. Her website is Finest Kind Clinic and Fishmarket, and she writes medical essays at HeyDoc Xanga Blog.

 

 

 

 

Be Sociable, Share!