Lancet Embargo: not for use before 0001 Friday 30th
No cause is known for the majority of cases of SUDI – sudden death of an infant. A study at Great Ormond Street Hospital led by Dr Neil Sebire, Dr Weber and colleagues, has shown that infection may be more significant than previously thought.
Published in the Lancet, this is the largest retrospective study of post mortem data, benefiting from being a single centre study based on a common protocol. 546 cases were examined. These were divided into those known to have been caused by infection, those known to have been caused by non-infectious causes such as congenital heart defects or accidents, and those, the majority, where the cause was unknown.
In half of all cases of 'unexplained' SUDI, potentially pathogenic bacteria were cultivated, nearly as high a proportion as for the infection group. In addition, the nature of the pathogenic bacteria found in the 'unexplained' group was similar to those found in the 'infection' group, particularly S aureus and E coli. It did not match the distribution found more widely.
On the other hand the presence of pathogenic bacteria was shown not to be conclusive proof of infection: one quarter of cases which could be attributed to non infectious causes (heart disorders, accidents) proved to have such bacteria.
Therefore association between bacteria and SUDI has been shown, but any mechanism is speculative. In addition, one could not now 'diagnose' any particular unexplained case, since we know bacteria are often present even when unconnected to the death.
In the long term the aim is not just to understand the cause of these tragic deaths but to have better evidence on which to act to prevent them. This was a technically demanding study… the sort of study that can only be carried out at a large specialist centre like Great Ormond Street. It highlights the important role that specialist paediatric pathology services provide when dealing with such infant deaths.
FIGURES: the Lancet release quotes isolates. This statement quotes cases, ie individuals. Both sets of figures are correct and meaningful and demonstrate the same point, but we have used cases as being simplest.
One case may lead to one or more samples, and each sample may lead to one or more isolates. (If a sample cultures three different bacteria, that would be three isolates.)
We gratefully acknowledge funding from the Foundation for the Study of Infant Death.
GOSH-ICH Press Office: 020 7239 3125
For genuine and urgent out of hours call speak to switchboard on 020 7405 9200