Skip to content

Safety of home birth

April 15, 2015

Important, and reassuring, new evidence from the Netherlands

Evaluating the relative safety of home and hospital birth is tricky; most home births are low-risk and many hospital ones high risk, so comparing outcomes for babies born in either place is hopelessly biased. And it’s difficult to adjust for risk status after the event, so until recently most researchers have ended up concluding that home birth might be safe, or might be a bit more dangerous, but they can’t be sure. This has left the field clear for partisans to shout at each other across the barricades.

To answer the question properly we need to identify, before labour starts, low-risk women who plan to deliver at home and low-risk women who plan to deliver in hospital, and compare outcomes by the planned, not the actual, place of birth. Unfortunately such data are rarely recorded.

Until 2011, when the UK Birthplace Study (click here), registered 17,000 women planning to deliver at home, and 20,000 planning to deliver in hospital, recorded their risk status before they went into labour and compared outcomes by planned place of birth. For the low-risk women who had already had a baby, home birth was as safe as hospital. But for low-risk women giving birth for the first time, “there were 9.3 adverse perinatal outcome events per 1000 planned home births compared with 5.3 per 1000 births for births planned in obstetric units, and this finding was statistically significant”. Hence current UK advice that hospital is slightly safer for first births.

However, some Birthplace “adverse perinatal outcomes”, like encephalopathy and meconium aspiration, while undoubtedly serious, are things from which most babies eventually recover, and their diagnosis could also be influenced by knowledge of the intended place of birth. Perinatal death is a harder outcome, but rare. Among low-risk women giving birth for the first time in Birthplace there were only six deaths out of 4,500 deliveries in the planned home birth group and five out of 10,000 in the planned hospital group. These raw numbers favour hospital, but they are hardly conclusive. We need larger numbers, and this month the Dutch have provided them.


Professor Ank de Jong (above) from the Department of Midwifery Science at the Free University in Amsterdam, and her obstetric colleagues, combined three Dutch registries to do a Birthplace type analysis; namely one based on planned place of birth among women judged at low risk before the onset of labour. The 750,000 women planning home birth dwarf all previous similar studies, so her conclusions matter. The paper is in this month’s BJOG (click here) or for those with access problems Jonge_et_al-2015.

There was no difference in perinatal death between planned home and planned hospital births among low-risk women. The lack of difference applied to both first and later births. For first births the rates were 1.02/1000 for planned home births v. 1.09/1000 for planned hospital births, (adjusted odds ratio 0.99, 95% confidence interval 0.79–1.24).

In summary, and in contrast to the UK Birthplace results, home appears to be safe for first births in Holland. This is important news.

The BJOG editor obviously realised the topic was controversial and commissioned not one, but two commentaries.

frank chervenak

The first (click here) (or Chervenak_et_al-2015) was led by Frank Chervenak (above) a respected New York obstetrician. (Full disclosure Dr Chervenak is a well-known opponent of home birth.  I’ve known him for years, and coincidentally I debated this topic with him at an obstetric conference a few weeks ago. I wasn’t aware of de Jonge at that time!)

Read it for yourself, but in my, perhaps biased, opinion he made five weak points in his commentary, and one good one.

  1. He cited a quite different, much smaller, and much criticised, single centre Dutch study (click here) to suggest that the data on intended place of delivery was inaccurate.
  2. He grumbled about the high rate of missing data on neonatal deaths after one week of age, while ignoring the fact that this was probably random. The fact that some neonatal units did not report any data at all to the national registry is a pity but hardly likely to introduce bias. The authors’ sensitivity analyses did not alter the conclusions.
  3. He suggested that home birth recording of Apgar scores may be inaccurate. This may be correct, but it has no bearing on risk of death. You can misrecord an Apgar, but dead is dead!
  4. NICU admissions were higher among babies of nulliparous women planning to deliver at home, as compared with multiparous women planning to deliver at home, but this is also irrelevant to the question at hand.
  5. Finally he found a Dutch language article in which one of the authors, professor Jan Nijhuis from Maastricht, had apparently recommended that all nulliparous women should deliver in hospital.  I don’t have access, and can’t read Dutch, but surely the fact that Nijhuis was not a reflex supporter of home birth, but happy to stand behind de Jong’s paper strengthens her conclusions!

However, Chervenak made one good point.

  1. Even if home birth is safe for low risk women in the Netherlands, a small country with well-trained midwives, and well-regulated systems in place, it may not be safe in the US where home births are often supervised by untrained self-styled experts.

He is surely right about that! Although whether this means US obstetricians should campaign for better midwifery support for home birth, or for everyone to deliver in hospital, is another matter!

Marian Knight

The other comment piece (click here) was by Marion Knight (above), one of the Birthplace researchers. She agreed that the Dutch data were reassuring, but worried about the same missing deaths which had concerned Chervenak; although they were probably random and probably wouldn’t have altered the conclusions, there were rather a lot of them.  Since Birthplace had more complete data, she suggested it may more accurately reflect the relative safety for low-risk women giving birth for the first time in the UK.

I agree with Dr Knight. For low-risk women in their first pregnancy I will continue to advise that in the UK hospital is probably safer. But if they wish to go ahead and deliver at home, these new Dutch data suggest there’s no need to make a fuss.

Of course none of this applies to women with significant risk factors (click here); they would be much safer in hospital.

Jim Thornton

3 Comments leave one →
  1. Sicco Scherjon permalink
    April 16, 2015 8:27 am

    Dear Jim

    I always read your ripe tomato, with so much pleasure, canoeing, poetry and a home delivery!!

    Thanks so much Sicco

    > Op 15 apr. 2015 om 19:20 heeft “” het volgende geschreven: > > >

  2. April 16, 2015 11:00 am

    Wow thanks Jim. NICE has said a similar thing in their recent updated intrapartum care guidance-
    In short evidence for homebirths and stand alone midwife led units, for low risk women, are safer , than hospital settings.Interesting arguments though…for and against!


  1. Home birth again |

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: