For some time now, there has been a heated debate in Australia about home births. Home birth advocates argue that a woman should have the right to choose how to give birth to her baby and if she chooses to have a home birth, that choice should be supported by providing the her with adequate care. Others claim that providing that kind of care encourages home births which are inherently dangerous.
Part of the problem with the debate is that it is sometimes dominated by crazies on both sides. On one side, some have called for the unnecessarily oppressive act of banning home births while on the other side, you see a lot of hippy-talk about the “sacred” and “natural” act of giving birth (see hilarious videos below).
Amid all this rhetoric, it was refreshing to read health reporter Melissa Sweet analyse the numbers of the relative risks of home versus hospital births as reported in a recent study published in the MJA. She points out that newspaper headlines reporting that “Babies are seven times more likely to die during home births” are exaggerating the degree of confidence the researchers have in the numbers.
The take-home message of Sweet’s article is that “there is quite a lot of good news for home-birth advocates in this study.” But reading her own interpretation of the figures, I don’t think this is so clear. She writes:
Where the media generally reported home-birth babies being seven times more likely to die during delivery, the estimate ranges from them being anywhere between 1.5 and 36 times more likely to have this happen.
So even the most generous interpretation of the results mean that if you choose to have a home birth, you are imposing a 50% increased risk of death on your child but as Sweet points out, it could be much worse.
Sweet continues:
A similar caution surrounds the widely reported finding that home-birth babies were 27 times more likely to die from lack of oxygen during delivery. Again, this finding had wide confidence intervals, with the estimate ranging from eight to 89 times greater — clearly, another one to take with caution.
Again, at the most generous interpretation, babies being eight times more likely to die from lack of oxygen seems like a very serious risk to impose on the child.
Certainly, as Sweet says, it is the right of mothers to choose how and where to give birth to their children but these significant risks ought to be communicated very clearly to parents. It seems to me the benefits of home birth are far outweighed by the apparent risks.
Kennare RM, Keirse MJ, Tucker GR, & Chan AC (2010). Planned home and hospital births in South Australia, 1991-2006: differences in outcomes. The Medical journal of Australia, 192 (2), 76-80 PMID: 20078406
Here’s some examples of the crazy end of the pro-home birth side.
The problem with the numbers is that they don’t come from a prospective randomised trial, but are retrospective observational data, with all the biases and variables that this entails.
My impression was that there were very few home births where the location of birth was a factor in an adverse outome. I’m not a statistician but it seemed that the small numbers of serious problems meant that any estimates had wide confidence intervals and therefore claiming that something was ‘27 times more likely’ to happen doesn’t appear to be very accurate.
Yeah – that sounds right. So perhaps the study doesn’t tell us much at all.
However, Sweet seems to think that if we take into account the wide confidence interval, babies are at least 1.5 times more likely to die during labour and possible 36 times. So, while the claim of “27 times more likely” is inaccurate, “at least 1.5 times more likely” might be right.
But I’m rough on my statistics too, so I’m just going on what Sweet says.
Here’s a really helpful follow up by Melissa Sweet on the Crikey blog “Croakey”. It includes some details about the numbers involved other problems with the study.
1.5 times what?
It would be interesting to see these figures in dots-on-paper natural frequencies.
There’s an interesting interview on converting unintelligible geewhiz probabilities into tractable medical options on the ABC Health Report that shows that probabilities aren’t always that useful.