Intended for healthcare professionals

Observations On the Contrary

Throwing the baby back into the bathwater

BMJ 2010; 341 doi: (Published 11 August 2010) Cite this as: BMJ 2010;341:c4292
  1. Tony Delamothe, deputy editor, BMJ
  1. tdelamothe{at}

    What explains the fall in planned home births in developed countries?

    As it is with individuals, so it is with companies, governments, and nations. Some thrive; others merely bump along. A few will even spectacularly crash and burn. The possible causes of these divergent life trajectories are endlessly fascinating—and not just to historians, investors, and readers of celebrity magazines.

    The fortunes of healthcare practices seem to wax and wane as well. You’d like to think that unfolding scientific evidence was the main explanation for this fluctuation. But even a cursory glance at a topic in the news of late—planned home births—doesn’t support such a simple correspondence. In fact, it hardly supports any correspondence at all.

    Since the second world war numbers of planned home births have fallen precipitously in the developed world, mostly to rates of less than 5% of all births. Yet over that time no compelling evidence has emerged that hospital delivery—the main alternative—routinely produces better outcomes in uncomplicated pregnancies. That has not stopped the issue becoming highly contentious, with official bodies lining up on opposite sides of the argument. For example, the United Kingdom’s Royal College of Obstetricians and Gynaecologists and the Royal College of Midwives think that planned home births are fine. The American College of Obstetrics and Gynecology and the Australian Medical Association do not.

    The consensus is that the best way of settling this dispute—a series of suitably powered, randomised controlled trials—is never going to happen, so interested parties have to depend on the results of less good studies.

    Two scientific studies purporting to add to the debate have been published this year, and both provoked the furores that routinely engulf such research. The first (Medical Journal of Australia 2010;192:76-80), comparing outcomes of planned home and hospital births in South Australia between 1991 and 2006, found no significant difference in perinatal mortality (defined as the sum of stillbirths and neonatal deaths within 28 days of birth) (see BMJ 2010:340;c826, doi:10.1136/bmj.c826). What it did find was an adjusted odds ratio of 7.4 (95% confidence interval 1.5 to 35.9) for intrapartum deaths among planned home births, when compared with planned hospital births, and a ratio of 26.7 (8 to 88.8) for deaths attributed to intrapartum asphyxia. Given the gargantuan confidence intervals—and the fact that the odds ratios were derived from just two intrapartum deaths and three deaths attributed to intrapartum asphyxia in the planned home birth group—wise epidemiologists (not to mention midwives, obstetricians, and expectant mothers) will pass discreetly by on the other side of the road.

    As they might do with the second study, published last month in the American Journal of Obstetrics and Gynecology (doi:10.1016/j.ajog.2010.04.041; see BMJ 2010;341:c3551, doi:10.1136/bmj.c3551). This meta-analysis of 12 studies from “developed Western countries” compared outcomes of 342 056 planned home and 207 551 planned hospital deliveries. As in the Australian study, perinatal mortality (this time defined as the sum of stillbirths among fetuses of at least 20 weeks’ gestation or 500 g weight and deaths of liveborns within 28 days of birth) did not differ statistically between the two groups. So can home birthers now breathe a sigh of relief?

    Not so fast. For reasons unclear from the published article the authors switched their attention to neonatal mortality (perinatal mortality minus stillbirths) (BMJ 2010;341;c4033, doi:10.1136/bmj.c4033), despite having relevant data for these calculations on only 9% of their total sample. Neonatal deaths were twice as high among the planned home birth group and three times as high among infants without congenital abnormalities. Despite the reassuring finding on perinatal mortality, it was these differences in neonatal mortality that were seized on by the headline writers of medical journals and newspapers alike. (None of them thought to ask why stillbirths suddenly didn’t matter.)

    Help—although not the holy grail of a randomised controlled trial—is at hand. The University of Oxford’s National Perinatal Epidemiology Unit will soon be reporting the results of its prospective cohort study of 75 000 pregnancies, looking at planned place of birth and clinical outcomes among women and babies at low risk of complications at the beginning of labour. It should at least allow us to draw more confident conclusions about the safety of planned home births (at least in 21st century England), but that just highlights the unanswered question: if unfolding scientific evidence can’t explain the waning fortunes of planned home births since the second world war, then what does?

    The historian of childbirth Irvine Loudon says that with the discovery of sulphonamides and the tremendous decline in mortality from puerperal fever, starting in 1937, home deliveries tended to give way to hospital ones. But a decade later the ultimate ambitions of the Royal College of Obstetricians and Gynaecology was also a decisive factor. It wanted the whole of obstetrics as a hospital specialty from which general practitioners would be excluded.

    Midwives have managed to retain a presence, but by 2008 Loudon was saying that “the active role of the GP in obstetrics is—or probably soon will be—dead” (Journal of the Royal Society of Medicine 2008;101:531-5, doi:10.1258/jrsm.2008.080264). He quoted a GP enthusiast for obstetrics, Gavin Young, on the reasons why GPs stopped delivering babies. The new arrangements for working out of hours and the GP contract loomed large, but even before that “most general practitioners would have run a mile to avoid attending a woman in labour.”

    Last week the healthcare think tank the King’s Fund was lamenting not the withdrawal of GPs from providing intrapartum care but the disappearance of any involvement of GPs in maternity care (BMJ 2010;341:c4254, doi:10.1136/bmj.c4254). Special training and financial incentives might lure them back to providing antenatal care but not, I’d vouch, to planned home deliveries. The results of next year’s National Perinatal Epidemiology Unit study will tell us whether we should be happy with midwives undertaking them.


    Cite this as: BMJ 2010;341:c4292


    • Competing interest: TD’s wife is a consultant obstetrician and gynaecologist.

    Log in

    Log in through your institution


    * For online subscription