Video: https://www.tiktok.com/@distilledscience/video/7465039948178001183
Transcript
This study has been scaring a lot of people. You may have seen it being shared by people like this. And yet it's easy to glance at it, see these words, and immediately either think, Ha, I told them so, or to assume there must be a problem and discount it. But if you're on Team Science, that means you critically evaluate all new evidence with an open mind. So, let's put on our critical thinking hats and dive in. It was an observational study that used Medicaid claims billing data. It looked at 47,155 children enrolled in Florida Medicaid ages 0 through 9. neurodevelopmental disorders like autism based on the presence or absence of those billing codes. It concluded that vaccinated children were between 2.7 and 6.8 times more likely to be diagnosed with these disorders. It then compared the number of vaccination visits with the odds of an autism diagnosis. And it found that those with a single vaccination visit were 1.7 times more likely to be diagnosed with autism all the way up to 4.4 times the risk in those with 11 or more vaccination visits. That last 4.4 number is what he used to say that nearly 80% of all autism caused by vaccines. This all sounds quite scary. But let's break down four key flaws. The fourth alone being enough to make the results highly questionable. First, this was an observational study which can never be used to determine causation. If crocodiles cause autism and every autism clinic was next to a crocodile swamp, this study would still observe the vaccine correlation. Think about that. Second, using Florida Medicaid billing codes as the sole data source is highly unreliable, especially because people could have gotten vaccines from other providers or other states. they chose to only count autism diagnoses that were diagnosed between the ages of 5 and 9, when it can be reliably diagnosed by age 2, and on average, it's diagnosed by age 4.9. It could be that younger numbers didn't support their desired conclusion, so they drew the line there, which could be called cherry-picking dad. Fourth, the clincher. They didn't adjust for confounding factors like these. The key one being healthcare-seeking behavior. If a family doesn't trust doctors, they're a lot less likely to take their kid to a specialist to get the diagnosis of autism, just as they're less likely to be able to be able to get them vaccinated. But that case would still show up in this study as an unvaccinated child without autism. And to close it off, the study was published by an independent website made to look like an academic journal, but is owned by the same organization that funded the study, a big no-no. And most of the journal's editorial staff work for RFK Jr.'s anti-vaccine advocacy group, Children's Health Defense. Interesting timing given that RFK's confirmation hearing is this week. And it benefits him to make his past anti-vaccine campaigning and profiting, and profiting, to seem more legitimate to those unfamiliar with reading scientific research. So spread the word for science.
Additional notes
⚠️ Other Problems: 1. Poor statistical methods (no randomization, sensitivity analysis or proper adjustments) 2. Didn’t account for differences in how developmental disorders were diagnosed over time, especially after the publication of an updated DSM-IV-TR in 2000. (This shift is why population-level prevalence of ASD changed from ~1.16% in 2007 to 2.00% in 2011) 3. Didn’t track the development of the autism diagnoses over time or consider timing of vaccinations relative to diagnoses 4. They claim to track individual-level data while simultaneously admitting they cannot track specific vaccines 5. They don’t make their dataset publicly available I started writing up a full critique of the study, but part-way through I found two excellent ones already written by Dr. Jess Steier from @Unbiased Science Podcast and @Dr. Andrea Love | Immunologist –some of their points then influenced this video. I’d highly recommend checking both of them out! #science #vaccines #autism #misinformation #who
References
- Population-level prevalence of ASD source linked in caption: https://www.cdc.gov/nchs/data/nhsr/nhsr065.pdf
- Study discussed in video: title, DOI, and PMID not listed in workbook.
- Critiques mentioned in caption: Dr. Jess Steier / Unbiased Science Podcast and Dr. Andrea Love; source URLs not listed in workbook.