Ministry Ignores Published Matress Wrapping Proof
MEDIA RELEASE FOR PUBLICATION:
MINISTRY OF HEALTH IGNORES PUBLISHED STATISTICAL PROOF THAT MATTRESS-WRAPPING PREVENTS COT DEATH
On 18 October 2006 scientist Dr Jim Sprott OBE sent the submission below to the Minister of Health. Despite the statistical proof that mattress-wrapping prevents cot death (as published in two peer-reviewed medical journals), the Ministry of Health has stated that its new pamphlet on cot death prevention will contain no information about mattress-wrapping.
Honourable P Hodgson
Minister of Health
Re: Cot death (SIDS) prevention - Mattress-wrapping
I refer to your letter dated 13 October 2006, in which you state:
"The Ministry of Health is unaware of any published studies in peer reviewed journals which demonstrate the safety and efficacy of mattress wrapping in the prevention of sudden infant death syndrome (SIDS). Accordingly, as the Ministry's policy recommendations are expected to be evidence based, they are unable to recommend mattress wrapping as a method for preventing SIDS."
1. EFFICACY OF MATTRESS-WRAPPING: PEER-REVIEWED PUBLISHED STUDIES
I enclose herewith the following peer-reviewed study regarding the efficacy of mattress-wrapping, which was published in the Journal of Nutritional & Environmental Medicine (Taylor & Francis Ltd, UK) in September 2004: Sprott, T J, Cot Death - Cause and Prevention - Experiences in New Zealand 1995-2004, Journal of Nutritional & Environmental Medicine 2004;14(3):221-228.
At page 224 of that article, under the heading "Proof of toxic gas explanation for cot death", is set out the statistical "p" factor for the mattress-wrapping intervention: p = less than 1.9 x 10(exp -22), as calculated by German researcher of environmental medicine, Dr H Kapuste (in collaboration with statisticians at the University of Munich) and published in a German peer-reviewed journal of environmental medicine in 2002: Kapuste, H, et al, Giftige Gase im Kinderbett (Toxic Gases in Infant Beds), Zeitschrift fuer Umweltmedizin 2002;44:18-22.
It is usual in medical circles to regard a "p" factor of less than 0.01 = 10(exp -2) as sound proof of a scientific proposition; and if the "p" is less than 0.001 = 10(exp -3), that is regarded as virtually certain proof.
1.9 x 10(exp -22) (the "p" factor
for mattress-wrapping) can be written
2. EXPLANATION OF STATISTICAL CALCULATION BY REFERENCE TO NEW ZEALAND COT DEATH STATISTICS
The statistical explanation set out below has been compiled using methodology provided by a highly qualified UK pathologist who is expert in statistics and also a cot death researcher.
As calculated by Kapuste, the "p" factor for mattress-wrapping is 1.9 x 10(exp -22). Based on the fact that 1.9 x 10(exp -22) equals e(exp -50), the statistic is that there are zero cot deaths on wrapped mattresses when the expectation on the null hypothesis (that mattress-wrapping did not prevent cot death) would be 50 cot deaths. The probability of zero when the expectation is 50 is e(exp -50) (Poisson statistics).
Mattress-wrapping for cot death prevention commenced in New Zealand in late December 1994, with the first major nationwide publicity occurring in 1996 at the time of publication of "The Cot Death Cover-up?" (Penguin Books, New Zealand 1996; UK 1997).
There has been no reported cot death on a mattress wrapped inaccordance with the mattress-wrapping protocol.
Two separate epidemiological studies, conducted in (respectively) 1997 and December 1997-February 1998, and reported in the New Zealand Medical Journal, both found the incidence of babies sleeping on wrapped mattresses to be 23%: Ford, R, Schluter, P, Cowan, S, Changes to infant sleep practices in Canterbury, NZ Med J 2000;113:8-10; Mitchell, E A, Subramaniam, K, Blackburn, J, Shooter, D, Plastic wrapping of cot mattresses: results from a pilot study, NZ Med J 2000:113:326-327.
(Taken historically overall, however, the figure of 23% is probably conservative, owing to significant upsurges in mattress-wrapping at various points in the history of the mattress-wrapping intervention.)
Thus Kapuste's paper states the figure of at least 100,000 babies having slept on wrapped mattresses, i.e. from the commencement of the mattress-wrapping intervention to time of publication.
The expected number of cot deaths if mattress-wrapping did not prevent cot death can be calculated from the numbers of cot deaths in the three years immediately prior to the 1996 publicity, because:
(a) Face-up sleeping was introduced in New Zealand in September 1989;
(b) Having reduced by 49%, in 1993 the cot death rate reached the plateau which accompanies all face-up sleeping campaigns, and beyond which no further significant reduction in cot death is achieved by that intervention: in New Zealand 2.1 deaths per 1000 live births (125 deaths);
(c) The plateau continued through 1994 (2.1/1000; 121 deaths) and through 1995 (2.1/1000; 121 deaths);
(d) Mattress-wrapping received prominent nationwide publicity during 1996, and furthermore was the only new item of cot death prevention advice publicised since 1992 (i.e. before the plateau occurred). However, all items of orthodox cot death prevention advice which had been publicised prior to the plateau continued to be publicised throughout the years of the plateau.
Since, therefore -
(a) The plateau is indisputable (1993: 125 deaths; 1994: 121 deaths; 1995: 121 deaths); and
(b) All items of orthodox cot death prevention advice publicised prior to the plateau continued to be publicised throughout the plateau and have been since -
it can be calculated that if mattress-wrapping (being the only new item of prevention advice) did not prevent cot death, the expected number of cot deaths during the 11 years from 1995 to 2005 (inclusive) (throughout which time mattress-wrapping has been practised) would be:
121 x 11 = 1331 (say, 1330)
During those eleven years around 750 cot deaths occurred.
Thus the expected number of cot deaths on wrapped mattresses if mattress-wrapping did not prevent cot death is: 1330 minus 750 = 580. This equates an average of 580/11 annually = 53 (rounded down to 50 for statistical analysis).
Thus the annual expectation of cot deaths on wrapped mattresses if mattress-wrapping did not prevent cot death is 50. Observed cot deaths on wrapped mattresses: 0
The above prime data results in p = e(exp -50); in other words, the same statistical outcome as that published by Kapuste: p = less than 1.9 x 10(exp -22).
I refer you to the typical standards of proof applied by the medical profession to scientific propositions [as set out in section 1 above]. The statistical proof that mattress-wrapping prevents cot death (as reported by Kapuste) is - in mathematical terms - one billion billion times the level of proof which medical researchers generally regard as constituting virtually certain proof of a scientific proposition.
OF MATTRESS-WRAPPING: ELEVEN-YEAR INTERVENTION IN NEW
As stated above, mattress-wrapping for cot death prevention commenced in late December 1994.
A conservative estimate of the number of babies who have slept on wrapped mattresses to date is 150,000. There has been no reported cot death among those babies.
By contrast, since the commencement of mattress-wrapping around 800 cot deaths have occurred on unwrapped mattresses (or parallel bedding situations, e.g. in adults' beds).
In July 1996, Dr P G Tuohy (then National Paediatrician for the Royal New Zealand Plunket Society; now Chief Adviser on Child Health at the Ministry of Health) wrote to Dr G Durham (then Director of Public Health) as follows: "... I believe we should capitalise on this massive uncontrolled experiment by monitoring the uptake of mattress wrapping in NZ. According to the Sprott theory, mattress wrapping should totally eliminate SIDS. Therefore one SIDS case on a properly wrapped mattress disproves the [toxic gas] theory [for cot death]."
It is, of course, the case that the converse also applies: nil cot deaths on wrapped mattresses proves the toxic gas theory (of which mattress-wrapping is the corollary intervention).
That situation has now been demonstrated in New Zealand, and furthermore over a statistical sample of massive size.
IRRELEVANCE OF EPIDEMIOLOGICAL (ODDS RATIO) ANALYSIS
It has been claimed by some researchers that mattress-wrapping for cot death prevention cannot be officially endorsed in the absence of an epidemiological case-control study into mattress-wrapping.
The claim that such a study is required is incorrect as a matter of mathematical logic.
Certainly, if it were the case that some babies had died of cot death on wrapped mattresses, and some babies had died of cot death on unwrapped mattresses, such a study would be valid, as calculations of odds ratios would be relevant.
However, the absence of cot deaths on wrapped mattresses over an extended period of time (11 years) and across a massive sample has the following mathematical result:
Regardless of whether the proportion of babies sleeping on wrapped mattresses is 20% or 40% or 60%, the success ratio of mattress-wrapping is the number of deaths on unwrapped mattresses divided by zero (being the number of deaths on wrapped mattresses), i.e. "infinity".
Since, therefore, it is known that a large and statistically significant number of babies have slept on wrapped mattresses, it is valid to state that the success of the mattress-wrapping intervention is 100%.
5. SAFETY OF MATTRESS-WRAPPING FOR COT DEATH
The proprietary product which is used for mattress-wrapping for cot death prevention is the BabeSafe cot mattress cover.
In 1999 the Ministry of Consumer Affairs stated that BabeSafe covers do not raise any consumer safety issues.
In 2003 Dr P G Tuohy, Chief Adviser on Child Health at the Ministry of Health, stated that BabeSafe covers are safe for babies.
In 1999-2000 the United States Consumer Products Safety Commission examined BabeSafe covers and cleared them for use on babies' mattresses.
The statistical evidence that mattress-wrapping prevents cot death is indisputable. As demonstrated at section 2 above, the level of proof that mattress-wrapping prevents cot death (as now reported in two peer-reviewed environmental medicine publications) is greater than the level of proof afforded by DNA testing.
Accordingly, irrespective of whether or not cot-death-related research continues in New Zealand (for example, the Ministry of Health's proposed three-year SUDI Case-Control Study 2008-2010), it is incumbent on the Ministry of Health not to await the outcomes of that research but to immediately advise parents to wrap babies' mattresses for cot death prevention.
As demonstrated at Section 4 above, the claim that such a case-control study is required to assess the efficacy of mattress-wrapping is fallacious as a matter of mathematical logic.
Failure to advise the public to wrap babies' mattresses for cot death prevention lays the Ministry of Health (and other providers of babycare advice) open to legal challenge.
This is because - regardless of the stance towards mattress-wrapping taken in other legal jurisdictions - it has been statistically demonstrated from New Zealand experience that mattress-wrapping prevents cot death; and that the level of proof thereof far exceeds the evidence which formed the basis for the introduction of items of cot death prevention advice currently publicised by the Ministry of Health and other organisations (for example, face-up sleeping).
The safety of the mattress-wrapping protocol is unimpeachable.
Your letter dated 13 October 2006
"... as the Ministry's policy recommendations are expected to be evidence based, they are unable to recommend mattress wrapping as a method for preventing SIDS."
Given the evidential matters set out above, the Ministry of Health's position in not recommending mattress wrapping for cot death prevention is demonstrably illogical and legally indefensible.
T J Sprott
T J Sprott OBE MSc PhD FNZIC
Consulting chemist & forensic scientist