Cosleeping risk, an engineering analysis

Provincial coroners are going on about the risks of cosleeping with one's baby. It's difficult to separate the hysteria from the facts, as "one kid died somewhere" gets widely reported but the tens of thousands of kids that safely benefit from it does not. I don't want to raise a kid leashed to the clothesline in the backyard with a helmet on, but I don't want to expose my son to unnecessary risk either. If the risk is remote, and/or able to be mitigated, there are advantages to be had.

I don't understand the bias of the medical community against cosleeping. I'm not militantly for it - we ended up there because it was the only way anyone would sleep at all in the early going, and we're on a track to get him into his own bed - but the rank bias of the coroner's report, the hysteria of the reporting, and the lack of hard facts really gets my back up. But pride and principle don't mean a damned thing if our kid dies, so we tried to establish the facts to determine if we had an imminent problem.

• In 2003 there were 130,927 live births to Canadian residents in Ontario (from Statscan). Numbers were similar in 2002. I am assuming numbers would be similar for more recent years. The Wikipedia entry for Ontario demographics indicates 671,250 children under age 5, which roughly correlates.

• The percentage of families that cosleep "always" or "almost always" was 25% between 1991 and 1999 according to a National Center For Health Statistics survey cited in the Dr. Sears book "The Baby Sleep Book". A further 42% slept with their baby "sometimes". Dr Sears is a major proponent of cosleeping so it's possible there is bias there but the number looks solid to me. At 25%, this would be 32,732 cosleeping babies in any given year.

• The 2008 Ontario Coroner's Report includes a section on infant deaths. They are mandated to investigate all deaths of children under the age of 5 (not all are full blown inquests but evidence from all are examined). They investigated 186 cases over two years; 77 involved "unsafe sleeping", of which 41 were "bedsharing" (their term for babies sleeping in an adult bed). Of the 41, 37 involved bedsharing with the baby and one or two adults; the other 4 also involved siblings (a recognized cosleeping problem due to various unpredictabilities of sleeping children). They did not break it down per year. It should be safe to assume half in each year - round up and say 19 bedsharing deaths in one year.
  

So that means 19 deaths per 32732 cosleeping children, or 1 in 1722, or 0.06%. I'd be more comfortable with another order of magnitude but that's less than one's lifetime odds of dying in a car accident (1 in 100!), by fire or smoke, or even by "falling down" according to one list.

But even that's not the whole story. The coroner's report states that of the 77 unsafe sleeping deaths, the sleeping positions were: 31 on stomach; 10 on back; 9 on side; 27 unknown. How those numbers map to the 41 bedsharing deaths is not stated. However, stomach sleeping is a known SIDS risk factor. Stomach sleeping deaths were 31/77 or 40% of the 77 unsafe sleeping deaths. If you exclude a similar ratio of deaths from the 19, we're down to 12 (rounded up). That would be 12 in 32732, or 1 in 2728, or 0.04%.

Furthermore, the coroner's report states a couple of other nebulous statistics: that 23 of the 77 unsafe sleeping deaths had been Children's Aid Society cases within the 12 months prior to their deaths. Also, drugs/alcohol were somehow involved in 14 cases. It's not clear how these map to the 41 cosleeping deaths, or the degree to which drugs/alcohol were involved. However, there is a definite implication there.

So if we put him to sleep on his back, keep neighbourhood kids out of the bed, don't drink or do drugs, and don't beat him (CAS), the odds of a problem look long to me. I'm open to the possibility that there is a problem with my numbers or assumptions, but from what I can tell, I'm comfortable that cosleeping with my kid is just fine. We're on a trajectory to get him out of our bed anyway.

I'm going to bed with my wife and kid.