New Cot Death Research Invalid
31 January 2008
MEDIA RELEASE FOR PUBLICATION
New Cot Death Research Invalidated By Biased Sampling and Incorrect Statistical Analysis
New cot death research by Professor E A Mitchell published in the February 2008 issue of the European Journal of Pediatrics (1) is invalidated by biased sampling and incorrect statistical analysis, says cot death prevention expert Dr Jim Sprott.
The research reports a 2005 postal survey of 400 Auckland mothers regarding practices adopted for cot death prevention. (2) On the basis of 280 responses, Dr Mitchell reported the incidence of mattress-wrapping for cot death prevention as being 21.7%; and concluded that even if mattress-wrapping were 100% effective in preventing cot death, the practice could have reduced the cot death rate by only 22% from 1994 to 2004 (as compared with an actual reduction of 63%).
"Dr Mitchell's research is invalid," stated Dr Sprott.
"First, the research sample was biased. The mothers were recruited from the birth list at Auckland City Hospital, which is a Ministry of Health hospital; and the Ministry's policy on mattress-wrapping is that there is no evidence of a link between mattresses and cot death risk. Obviously, sourcing survey participants from Auckland City Hospital (many of whom would have attended antenatal classes at which the Ministry's policy on mattress-wrapping was presented) will have resulted in a highly unreliable statistic for the incidence of mattress-wrapping in the general community. It is likely that the incidence of mattress-wrapping has been under-reported."
In order to survey the incidence of mattress-wrapping in the community, survey participants must be sourced from records of the Registrar of Births, not records of a Ministry of Health hospital, said Dr Sprott.
"Secondly, Dr Mitchell's statistical analysis is nonsense," stated Dr Sprott. "In any group of medical interventions aimed at the same result (for example, prevention of a particular disease), it would be entirely possible for one intervention practised by (say) 40% of a population group to result in (say) 70% of the reduction of the incidence of the disease. As a matter of statistics, the efficacy of a particular intervention is not by definition linked to the percentage number of the group who practise that intervention.
"Dr Mitchell's claim that if 22% of babies sleep on wrapped mattresses, that can account for only 22% of the reduction in the cot death rate, is simply incorrect."
Since 1995 an estimated 165,000 New Zealand babies have slept on mattresses wrapped for cot death prevention. There has been no reported cot death among those babies, and since the commencement of mattress-wrapping the nationwide cot death rate has fallen by 67%. (3)
Dr Sprott noted that Dr Mitchell's research in the European Journal of Pediatrics may be publicised at the 10th SIDS International Conference, to be held in Britain in June 2008. Accordingly, Dr Sprott will be notifying cot death researchers internationally regarding the biased sampling and invalid statistical analysis in Dr Mitchell's research.
1. Mitchell, E A, Wrapping a cot mattress in plastic does not explain the continuing fall in SIDS mortality, European Journal of Pediatrics 2008; 167(2): 251-252 (Short Report).
2. "Survey of infant care practices that have been recommended for SIDS prevention", Auckland Ethics Committee reference AKY/04/08/217. Researchers: E Mitchell, L Hutchison, A Stewart, D Tipene-Leach, R Haretuku, M Battin.
3. Mattress-wrapping statistic based on two research studies which reported the incidence of mattress-wrapping in New Zealand: NZ Med J 2000; 113: 8-10; NZ Med J 2000; 113: 326-327. Source of cot death statistics: New Zealand Ministry of Health.