03
SIGINT
Paper detail
Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children
R 53 T 0 C 25 I 50
case-studyobservational
Key findings
In a case series of 12 children referred with both developmental regression and gastrointestinal symptoms, all showed intestinal abnormalities (chronic colitis in 11, ileal lymphoid hyperplasia in 7-10). Parents of 8 children reported onset of behavioural symptoms temporally associated with MMR vaccination. Urinary methylmalonic acid was elevated compared to controls (p=0.003). The authors proposed but did not demonstrate a link between MMR, intestinal inflammation, and developmental regression.
Claims (5)
moderateAll 12 children had intestinal abnormalities ranging from lymphoid nodular hyperplasia to aphthoid ulceration.
weakOnset of behavioural symptoms was associated by parents with MMR vaccination in 8 of 12 children.
moderateUrinary methylmalonic acid was significantly raised in the 8 children tested compared with age-matched controls.
unsupportedThe connection between intestinal and behavioural pathology reflects a unique disease process.
unsupportedThe described syndrome may be related to MMR vaccination.
Red flags (8)
Tiny uncontrolled sample: Case series of 12 self-referred children with no matched controls for the primary clinical outcomes (GI findings in children with developmental disorders). The only controlled comparison is for urinary methylmalonic acid (n=8 vs 14 controls).
Self-referral and recruitment bias: Children were referred by parents who already suspected a vaccine connection. The Acknowledgments thank 'the parents for providing the impetus for these studies.' This creates severe ascertainment bias - families who believe vaccines caused their child's regression are more likely to seek out a researcher investigating that hypothesis.
Causal framing from observational data: Despite stating 'We did not prove an association,' the paper's title, structure, discussion, and press conference heavily implied a causal link between MMR and autism. The study design (unblinded case series relying on parental recall) is incapable of supporting causal claims.
Reliance on parental recall for temporal association: The temporal link between MMR and symptom onset is based entirely on parental/physician recall, which is subject to recall bias. Parents seeking explanation for their child's regression may retrospectively anchor on a salient event like vaccination.
No blinding of histological assessment: The lead author (Wakefield, who had a prior hypothesis about measles and bowel disease) was one of the four pathologists assessing the histology. No blinding protocol is described for any clinical assessment.
Undisclosed conflicts of interest: The paper does not contain a competing interests declaration. It was later revealed that Wakefield had been secretly funded by lawyers seeking to sue vaccine manufacturers and had filed a patent for a competing single measles vaccine - conflicts that fundamentally undermine the study's credibility.
Claims outrun evidence: The paper identifies a 'unique disease process' and a 'syndrome' from 12 highly selected cases with no control group for the primary observations, no epidemiological data, and no mechanistic evidence beyond speculation.
Missing effect sizes: The only statistical test (methylmalonic acid comparison) reports p=0.003 but no means, standard deviations, or effect size - making independent evaluation of clinical significance impossible.
Games detected
Big Numbers No Error BarsOverclaimingMoving Goalpost
Dimension scores
Composite: 14.3(harmonic mean)
Checklist (16/35 passed)
Category scores
artifacts
0
statistical methodology
20
evaluation design
85.7
claims and evidence
50
setup transparency
0
limitations and scope
0
data integrity
25
conflicts of interest
50
human studies
80