Via Science and Sensibility, this new homebirth study out of British Columbia should be required reading for our Health Department and policy-makers: “Outcomes of planned home birth with registered midwife versus planned hospital birth with midwife or physician”.
This high-quality study compared equivalent-risk homebirths (HB), midwife-attended hospital births (Midwife-in-Hospital, MiH), and doctor-attended hospital births (Doctor-in-Hospital, DiH). In British Columbia, registered midwives are obliged to offer homebirth if the woman meets the eligibility criteria, so the study is not confounded for type of caregiver. Further methodological elaboration is below the cut.
The homebirth cohort included all births in British Columbia from 2000-2004 that were planned to take place at the woman’s home at the onset of labour. In-labour transfers were not excluded, so this was a true intention-to-treat analysis. 79% of the women who planned a homebirth had one, and 97% of the women who planned a MiH birth had one.
No Deaths Occurred At Home
The results are striking. Perinatal deaths in the planned-homebirth group and hospital groups were statistically similar: 0.35 per 1000 (95% confidence interval 0-1.03) for homebirth; MiH 0.57 (0-1.43); and the DiH 0.64 (0-1.56). Apgar scores were similar, and there was no difference in diagnoses of asphyxia at birth, seizures, or the need for assisted ventilation beyond 24 hours of life. None of the perinatal deaths occurred in births that actually took place at home.
Birth Injuries Lower at Home for Both Mothers and Babies
Both maternal and neonatal injuries were lowest in the homebirth group. Severe perineal tearing and postpartum haemorrhage were significantly lower in homebirth groups, as were neonatal complications such as birth trauma, the need for resuscitation at birth, and the need for oxygen therapy beyond 24 hours. Infections were lower in the HB group, but the difference was not statistically significant.
Neonatal birth trauma such as brain haemorrhage, skull and clavicle fractures, and nerve injuries was reduced by 75% in the homebirth group (values are rounded; see the study for full details). The need for resuscitation at birth was reduced by around 75%, and the need for prolonged oxygen therapy was decreased around 60%. Meconium aspiration was also less likely at home (remember, transfers to hospital were included in the analysis).
The only adverse outcome measure that was slightly higher in the HB group was admission to hospital (or readmission if born in hospital). This is hardly surprising, as the majority of neonatal admissions are for jaundice, which is typically recognised and treated before first discharge in hospital births.
Interventions Less Likely at Home
The rates of interventions in the HB group was also much lower across the board. Women planning homebirths had a 40% drop in augmentation of labour, 70% drop in narcotic analgesia, 60% drop in epidural analgesia, 60% drop in assisted vaginal delivery (vacuum or forceps), 25% drop in C section, and 50% drop in episiotomy (values are rounded; see the study for full details). All of these differences were statistically significant. Note that the total C section rate in Canada is around 26% and the section rate for DiH births for the lowish-risk women in this study was only 11% (!); C section rates in Australia are well above 30% (>34% in WA), so reductions with similarly supported homebirth in the Australian setting could well be greater, perhaps dramatically so.
These reductions in intervention are quoted from the midwife-at-home vs midwife-at-hospital model. The differences between midwife-at-home and doctor-in-hospital interventions were even larger across the board, with particularly large differences in rates of episiotomy, electronic fetal monitoring, and assisted vaginal delivery.
Key methodological points:
* The number of births in each scenario was in the thousands (2899 homebirths, 4752 midwife/hospital births, 5331 doctor/hospital births.)
* Only hospital births where the woman met the criteria for homebirth were included.
* The hospital cohorts were matched with the homebirth cohort for age, parity, marital status, and hospital where the midwife had privileges; and data was further collected for age, height, prepregnancy weight, BMI, income, drug/alcohol/tobacco use, gestations age at first visit, number of antenatal visits, and history of ultrasonography in early pregnancy.
* Unmeasured characteristics that may be related to self-selection remain uncontrolled-for. (Further comment in the CMAJ by McLachlan and Forster addresses the issue of randomisation in homebirth/hospital birth comparisons, and alludes to the current Australian political situation).
* The aim was to compare perinatal mortality, perinatal morbidity, and obstetric interventions.
* Homebirth qualification critera: gestational age 36-41 weeks, no more than one previous C section, spontaneous labour or outpatient induction (with intravaginal prosta- glandins or amniotomy). Excluded from homebirth are women who have certain disqualifying conditions: certain heart and kidney diseases, type I diabetes, pre-eclampsia, antepartum haemorrhage, active genital herpes, and placenta previa or abruption.
* Note particularly that vaginal homebirth after cesarean section is supported under the BC system, an idea that gives Australian OB/GYNs the vapours, and the study also included births where the fetus was found to be breech after the onset of labour. Additional subgroup analyses excluded women having VBAC and women having outpatient inductions, but this did not materially alter any of the conclusions.
Are We Looking at Risk the Right Way Around?
One of the interesting meta-things about this study is that hospital birth was considered the norm for the relative risk analyses, the “1.00″ to which homebirth was compared. Flipping the numbers and considering homebirth to be the norm would mean that we would have to talk about the relative increase in risk for neonatal and maternal injuries, and that in many cases the risks would be more than double. We would have to talk about how babies were three times as likely to get brain haemorrhages and broken bones in hospital births, how they were three times as likely to need resuscitation. How women were twice as likely to have their perineum cut open in hospital for no improvement in outcomes, how they were twice as likely to have an electronic monitor applied with no benefit, how they were more than twice as likely to have their baby removed with a mechanical device, how they were more than twice as likely to get a severe perineal tear.
I wonder how that would look, how that would be interpreted differently by the scientific and mainstream media? Would it have a different impact on you?
 Outcomes of planned home birth with registered midwife versus planned hospital birth with midwife or physician
Patricia A. Janssen PhD, Lee Saxell MA, Lesley A. Page PhD, Michael C. Klein MD, Robert M. Liston MD, Shoo K. Lee MBBS PhD
Canadian Medical Association Journal (CMAJ) 181(6-7), September 15 2009 (early release)