Intended for healthcare professionals

Rapid response to:

Papers

The UK accelerated immunisation programme and sudden unexpected death in infancy: case-control study

BMJ 2001; 322 doi: https://doi.org/10.1136/bmj.322.7290.822 (Published 07 April 2001) Cite this as: BMJ 2001;322:822

Rapid Response:

Counfounding

Editor

I note the conclusion of Fleming et al is that “the accelerated
immunisation programme in the UK is not associated with sudden unexpected
death in infancy, whether the death is explained or not explained” and
further “the standard primary course of immunisation may also have a non-
specific protective effect on the risk of death in infancy” or that
“failure to begin the course may be a marker of family organisation where
SIDS may be more frequent”.

The study is an excellent advert for mass immunisation, but I think
the team are ignoring the obvious, as gained from an alternative
perspective: -

One must acknowledge that they have tried vary hard to reduce,
through confounders, the probability of mistakes in statistical analysis
of known facts; yet some confounders must have been missed, and there is
no guarantee that the calculations applied for the known potential
confounders is accurate. Their application of ‘highly significant risk
factors in the infants’ sleeping environment for the last or reference
sleep” made the difference between SIDS children and Controls
insignificant, they remark that “the difference in immunisation uptake
was not significant but was still in the direction of immunisation being
protective”, hardly conclusive evidence for their conclusion.

Looking at their chart of confounders there is a clear difference in
all areas between the SIDS children and ‘matched’ controls; these are
indicative of strong variations in selected control children compared to
SIDS children studied – not really the ‘close matched grouping’ of first
sight. In every group there are glaring, thus potentially confounding,
differences – which may be the effects that the team have accounted for in
their confounding corrections, but did they do enough (Fine PE, Chen RT
1992) and isn’t it of fundamental significance that such differences
between SIDS and the Controls used do exist despite ‘matching’– especially
when in most cases one would expect the difference to be in favour of an
earlier death for a SIDS child than a control with or without (an
innocuous) vaccination?

It appears that, for the SIDS children, by percentage, compared to
Controls: -

1. Twice as many are generally older than controls (>120days) at
death.

2. Three times as many come from group V & unemployed parents

3. Four times as many moved house more than twice before death

4. Almost twice as many had younger mothers (<_25 years="years" old="old" br="br"/>5. Twice as many had two or more siblings

6. Almost four times as many were <_2500gms at="at" birth="birth" br="br"/>7. Four times as many had <_37 weeks="weeks" gestation="gestation" br="br"/>8. Almost four times as many had been admitted to special baby care units
prior to death

9. Almost twice as many had been admitted to hospital prior to death

10. Almost three times as many had Apgar scores of <_8 at="at" birth="birth" br="br"/>11. Almost five times as many had apparent life threatening events prior
to death.

In every case the data shows SIDS children have been vaccinated less
than the controls. Let us not forget that more SIDS children died “despite
less vaccine uptake” than controls - from which the assumption is made
that vaccines must protect children, rather than that in general SIDS
children already had greater potential for early death whatever they did.

I suspect that vaccination may help some ‘SIDS children’ (ie. Those
with the extraordinary predisposition to die earlier seen in the
confounder chart) to die earlier, and that this will be born out once
recognisable physiological and biochemical markers for cause of death by
vaccination are established. This hypothesis seems to be born out by
parental experience of time of death occurring in very close proximity to
immunisation in children - some within 2 or 3 hours of been given a
vaccine and still suffering ADRs to the vaccine. Other SIDS cases will
have a number of recognised predispositions, as shown above, prior to
death – should they have been immunised at all? Then there are children
with no apparent predisposition to die, whose parents report having a
healthy child until immunisation took place; if vaccination kills these
children, how?

We must also bear in mind that previous research has shown

1. An unimmunised (with DTP) SIDS child had a higher chance of dying
after seeing a doctor prior to death. (Barraff LJ et al 1983)

2. A DTP immunised child was more likely to die than a non-DTP immunised
child. (Barraff et al 1983)

3. Twin boys simultaneously died of SIDS within 2 to 3 hours after DTP
(Roberts SC 1987)

4. Post vaccine symptoms were more common with Plain DTP, as opposed to
Adsorbed DTP (Pollock et al 1984).

5. Pyrexia was notably more frequent after Plain than Adsorbed DTP (Waight
PA 1983)

6. There was a considerable increase in local reactions after the 3rd dose
of Adsorbed DTP than earlier doses (Pollock et al 1984)

7. The team must assume the SIDS diagnoses are correct yet many SIDS case
may be misdiagnosed; for example why should asphyxia be classed as SIDS
(Dudley Bell e-BMJ, e-letter /319/7223/1457, Feb 2000)? How many Shaken
Baby Syndrome cases are actually undiagnosed SIDS?

8. They looked at a subgroup whose birth weight was >2500 gms, yet this
group has already been shown to have either no, or negligible
predisposition to SIDS from vaccination with DTP (Walker AM et al 1987)

9. Low weight gain predisposes a baby to SIDS, despite normal birth weight
(Blair et al 2000)

10. Recent studies are pointing a finger at thimerosal poisoning in many
children, through vaccination, who may reach their toxic overload at
different parts of the immunisation schedules.

Yet the team did not confound for vaccine Brand, or vaccine Type, or
how many vaccines had been used on any child pre death. Adjuvants were not
confounded for, especially important with the news of thimerosal toxicity
in many vaccinated children; was this additional confounder considered at
post-mortems prior to SIDS diagnosis? The SIDS children who suffered 'low
weight gain despite normal birth weight' do not appear to have been
accounted for. More SIDS children saw medical personnel prior to death,
and this has been linked as a potential cause of SIDS in unimmunised
children – a further reason why SIDS cases may appear without vaccination
as compared to controls with or without vaccination.

Rather than looking confoundedly at SIDS data surely the best method
of detection of cause has to be improved post-mortem procedures; looking
for vaccine virus particles, relevant DNA testing etc. should be a
prerequisite when a child dies within a few weeks of immunisation – is it
really impossible to identify such a cause? Or is the will to prove that
the system that autopsied is the same one that killed the child, in some
cases, not there?

Regards

John H

References

1. Fine PE, Chen RT, Am J Epidemiol 1994 jan 15; 139(2): 229-30

2. Baraff LJ et al, Pediatrt Infect Dis 1983 Jan-Feb; 2(1): 7-11

3. Roberts SC, Arch Dis Child 1987 Jul; 62(7): 754-9

4. Pollock TM et al, Lancet 1984 Jul 21; 2(8395): 146-9

5. Waight PA et al, Arch Dis Child 1983 Nov; 58(11): 921-3

6. Walker AM et al, Am J Pub Health 1987 Aug; 77(8): 945-51

7. Blair et al, Arch Dis Child 2000; 82: 462-469 June

Competing interests: No competing interests

08 April 2001
John P Heptonstall
Director of The Morley Acupuncture Clinic and Complementary Therapy Centre
West Yorkshire