Home birth again

Most people imagine hospital is the safest place to be born, and it is if your mother has obstructed labour, major bleeding, or eclampsia. But hospital also has dangers. Not just infection, and blood clots, but staff led astray by false positive tests, or intervening “to be on the safe side”. Hospitals can also harm.

Consider the low-risk woman, without problems so far, who’s lucky enough to live in a rich country with skilled midwives who can look after her and her baby at home, but transfer if necessary.  Could planning to birth at home be equally safe?  Randomised trials are impossible, and simply comparing women who birth in either place is hopelessly biased because the groups are so different. The best we can do is compare cohorts of women judged low-risk at the start of labour, by “planned” rather than “actual” place of birth.

Two recent studies have done just that. The first from Canada (click here) retrospectively identified women without a risk factor before the onset of labour, classified them into planned home, or planned hospital, and randomly matched 11,493 in each group by parity and by whether or not they had had a previous Caesarean; apparently the latter is not judged a contraindication to home birth in Canada. There is potential for bias in this sort of retrospective matching, but the authors did their best to avoid it. The risk of stillbirth, neonatal death or serious morbidity did not differ significantly by planned place of birth (relative risk [RR] 1.03, 95% confidence interval [CI] 0.68–1.55). These findings held true for both nulliparous (RR 1.04, 95% CI 0.62–1.73) and multiparous women (RR 1.0, 95% CI 0.49–2.1). There were also fewer interventions among planned home births.

Last week (click here) a similar analysis came from Australia. Again the authors did their retrospective best, identifying 8,212 planned home births among low-risk women, and comparing them with a much larger number of planned low-risk hospital births. There were only nine baby deaths in the planned home group, but this was a slightly higher rate than for the planned hospital group, RR 1.6 (0.65–3.7). When divided by parity the differences were more marked in primiparous women RR 2.1 (0.58–7.8) than multiparous women 1.3 (0.4–4.1). But as the confidence intervals show, all the differences could have occurred by chance.

If these were the only studies, few partisans from either side of the home birth debate would be convinced. They are too small for precise estimates and there’s too much scope for retrospective classification errors. But they’re not.

The 2011 UK Birthplace study (click here) included 17,000 low-risk women planning birth at home, and was much more rigorous. Classification of risk and planned place of birth was done prospectively, before labour started. This reduces bias. In Birthplace for low-risk women having a first baby, planned hospital birth was safer for the baby, 5.3 v 9.3/1000 adverse perinatal outcomes, despite nearly half transferring to hospital in labour.  For second or subsequent births, there was no difference in adverse baby outcomes, the transfer rate was only 10%, and home birth substantially reduced intervention.

A similar 2015 study from the Netherlands (click here) included 466,112 low-risk women who planned to have a home birth at the onset of labour and 276,958 who planned a hospital birth. For first births the RR of baby death was 0.94 (0.76–1.16) and for second and subsequent births 1.02 (0.78–1.33).

Taken together the data are reassuring for low-risk multiparous woman in a rich country like UK, Netherlands, Canada or Australia where midwives are registered and well trained. For anyone like me who’s worked in hospitals for 40 years and seen both wise and unwise interventions, these data are plausible. Planned home birth, with transfer to hospital if problems develop, is safe, and intervention rates are much lower. For first births the highest quality study from the UK, Birthplace, suggests hospital is a bit safer and transfer rates are rather high.

The NHS is right to offer all low-risk multiparous women a home birth.

Jim Thornton

Note. Before anyone asks, birth in a midwife-led unit, whether freestanding or alongside a consultant-led unit, is safe for low-risk mothers, whether having their first or subsequent births. But everyone knows that. It’s hardly controversial any more.