Today, Pediatrics published a retrospective study with the above title, which purportedly provides the "strongest clinical evidence to date that on-time..." and that these results offer reassuring information" etc etc... I received an alert, and downloaded the study. And started reading. Most of the tables you can't make much of, since they simply tell you end-points and you have no way of knowing whether those end points are even accurate or not. I burst out laughing in the first paragraph of the results though when I read this: "A total of 491 (47%) of 1047 children met the study definition for timely receipt. An additional 235 (23%) received all recommended vaccines during the study period but not on time. The remaining 311 (20%) did not receive all recommended vaccines during the study period."
Now, anyone with even the most basic of maths, knows that 311 of 1047 can't be 20%. Okay, maybe it was a typo. So I worked out the percentages, which come to 100% when you round them up, but when you work them out exactly are missing almost 1 %. Which over 1047 adds up to 10 children. So then I added up those numbers, and guess what? 491+ 235+ 311 doesn't add up to 1047. It adds up to 1037. And all the accurate percentages, add up to 99.03, so yup, there's 10 kids missing.
But try as I might, I couldn't work out where the 10 missing kids are. Perhaps I'm dumb. We know that of the 1047 kids, 9 didn't receive any vaccines at all, but that can't explain the 10 missing kids. I could see no way in the data presented to work that out.
SO... I ploughed through the rest of the study, which really is meaningless, because it's endpoint "trust us" stuff. The data is supposedly publicly available (though there is no link to it) and it had been previously used in a thiomersal study before.
Ths study authors admitted that "This cohort did not have enough children who were fully unvaccinated in the first year of life to form robust estimates of neuropsychological outcomes as compared with children with other patterns of receipt. This is an inherent limitation of any VSD-based study given the generally high immunization rates of children within the member health maintenance organizations." But even more odd was this next sentence: "We did not attempt to control statistically for potential differences between completely unvaccinated children and those with later receipt."
Can we please stop calling "never vaccinated" children, "fully unvaccinated"? Why? In this country, the medical profession calls children "unvaccinated" when they have had all their vaccines bar one. So lets be scientifically accurate and consistent, and talk about fully vaccinated, partially vaccinated and never vaccinated.
What were these potential differences between fully vaccinated, partially vaccinated and never vaccinated children? They must have been observed, to say that there was no attempt to statistically control for them.
The authors also say, "...there may be alternative study designs that more accurately assess associations between time-dependent exposures and outcomes in retrospective studies."
Yes. There are far more accurate study designs, and one of the most accurate study design, would be to use the vaccine registers, which plainly were in existence in these health member organisations since 1993, to chronologically catalogue every child's health development, both acute and chronic, as those health effects are diagnosed.
Furthermore, if researchers are REALLY interested in a proper study, they should approach other organisations using similar data-bases who have a much higher percentage of never vaccinated children, who might be used as a scientifically valid control group.
My thoughts on this study.
Do I think this is a credible study? No. Basic math errors are a bad look. At the very least, a journal which is presumably peer reviewed, should have picked these mistakes up. If there were basic math errors in the text, what other analysis errors were made during the mining of the data? That the Journal didn't pick up the mistakes shows how much misplaced trust they have in "experts", and that peer review processes are a mockery. This is actually an opinion piece. And a biased one at that, with no serious attempt made to do a meaningful study with a meaningful control group.
Michael J Smith is a name that anyone who knows the name "Dr Paul Offit", should be familiar with. These two doctors have had a very close liaison for many years, and like Paul Offit, Michael J Smith has published his views on the media and MMR. Furthermore, like Dr Paul Offit, both the authors of this data-mining opinion piece, have disclosed that they have worked unfunded, or have received payment from many vaccine companies. Dr Michael J Smith has also authored a book called "The Complete Idiot's Guide to Vaccinations", and apparently has yet to gather the courage to do some serious research on the other side of the topic.
I will believe that Drs Michael J Smith and Charles R. Woods are interested in providing accurate scientific and mathematical answers, when they team up with Dr Mayer Eisenstein from Homefirst whose practice has a majority of children who are unvaccinated, and enlist an epidemiologist, to help them compare those children with a comparable practice from Chicago with fully vaccinated children.
And publish the results in a peer review journal.