Chamberlain, G, Wraight, A & Crowley, P (eds).
This is probably the most comprehensive study of home birth ever undertaken in the UK. It attempted to follow all women who booked a home birth in the UK in 1994. Midwives recorded outcomes for 5971 women who were booked for home births at 37 weeks' gestation, and they tried to find a matching woman from their practice who was as similar as possible to the home birth mother, but was booked to deliver in hospital. In some cases an appropriate matching hospital booking could not be found, so the hospital group only contained 4724 women.
The women were matched for age (within 5 yrs), number of previous children, where they lived, and past obstetric history. Thus, low-risk mothers were compared with other low-risk mothers, and the overall sample in the home birth group could be accurately compared to that in the matching group. The overall group profile was low-risk. For all outcomes, planned home births were compared with planned hospital births, so the data for planned home births include those births which occurred at home, and transfers to hospital. Separate data for transfers is also provided.
First-time mothers made up 16% of the group, whilst 84% were expecting second or subsequent children. Mothers who had a previous Caesarean made up 1% of the home birth group, and 3% of the hospital group.
It is impossible to remove all selection bias from a study of this type. Women planning home births tend to be more educated and more affluent than those in the hospital group, which would make them slightly lower-risk - although nobody knows why these factors make a difference amongst women already selected as low-risk.
There are probably other, unmeasurable subtle differences too. The home birth group had less incidence of high blood pressure between recruitment for the study, at 37 weeks, and the birth - which could be because midwife antenatal care was less stressful than hospital antenatal care, but could also be because women who developed high blood pressure before 37 weeks (and transferred their booking to the hospital) would not be included in the home birth group, but would possibly be included in the hospital group (although they would presumably still have to be considered low-risk, to be an appropriate matched pair).
Planning a home birth halved the chances of mothers in this study having assisted or caesarean births. The overall rate of these events was very low, because mainly only low-risk women booked for a home birth.
The authors note that
'The hypothesis that social and environmental factors can affect progress of labour and mode of delivery is strongly supported by experimental evidence. A systematic review of randomized controlled trials shows that continuous professional support during labour is associated with a reduced incidence of Caesarean section and instrumental delivery. In the light of this evidence it is easy to accept that some women may labour better in their own homes.'(1)
|Planned home births
|Planned hospital births|
|Spontaneous vaginal delivery||94.7%||90.2%|
|Assisted (forceps, ventouse)||2.4%||5.4%|
Home births, including transfers, were less likely to involve post-partum haemorrhage. Home birth mothers had fewer episiotomies but more first-degree perineal tears. Second-degree tears occurred at the same rate in both groups. All cervical tears occurred in the hospital group.
One set of twins were born at home, with no complications; one was cephalic and the other breech.
Of the women planning home birth who had given birth before, 9.7% had previous assisted deliveries and 1.4% (53 women) had previous caesareans. The transfer rate amongst the previous assisted delivery group was 15.6%, and among the previous caesarean group it was 28%; so 72% of these women had a home VBAC!
The study does not give data on the proportion of the women with previous caesareans who had another caesarean after transfer, but the overall caesarean rate for women transferred from home to hospital was 11.2% [p88].
16% of women booked for a home birth transferred to hospital. Dividing women into primigravidae (having first baby) and multigravidae (having second or subsequent babies), 60% of first-time mothers who had planned to deliver at home, did so, and 40% transferred. 90% of multigravidae who had planned to deliver at home, did so, and 10% transferred.
Some of these transfers occurred before labour actually started, whilst others occurred in labour. The single largest reason for transfer was slow or no progress, accounting for 37.2% of transfers. Premature rupture of membranes accounted for 24.8% of transfers, and most of these occurred before labour started. Foetal distress accounted for 14.8% of transfers. There was one cord prolapse - see "But what if?..." for more discussion.
The number of mothers and babies who transferred was 769; the authors say that 'the outcome was satisfactory for the baby in the vast majority of cases', although sadly this group included two stillbirths and two neonatal deaths.
The transferred home births had the worst outcome rate of all three categories (planned and born at home, planned and born in hospital, planned home but born hospital) for labour interventions and condition of the baby. Clearly, as transfer generally only occurs if there is a problem, this group would be the highest-risk component of the study. This does not mean that transferring from a home birth causes poor outcomes; it means that these were the women and babies who really needed extra monitoring or intervention, and whose outcomes would have been the same or worse if they had planned a hospital birth. This view is supported by the fact that the outcomes for planned home births plus transfers are still better than those for planned hospital births.
For more details of the study's findings on transferring from a home birth, see Studies on Home Birth Transfers.
Women who planned a home birth were much less likely to use drugs for pain relief, than women who planned a hospital birth. The drugs most commonly available at home births in the UK are Entonox (nitrous oxide and oxygen, 'gas and air'), and Pethidine (called Demerol in the USA). In some areas these substances are taken routinely to all home births, and in others they are available only on request.
This table shows the forms of pain relief used, as reported by midwives. Many women used more than one form, so the totals add up to more than 100%. It's also worth remembering that what one midwife might report as 'relaxation', another might report as 'no pain relief':
(born at home)
home to hospital
|Epidural||2.8% (all transfers)||11.3%||0||17.9%|
Babies planned for home birth were less likely to be in bad condition (low APGAR score) at birth, even including those transferred to hospital.
At 1 minute, 5.2% of the planned home babies had APGAR scores below 7, compared to 9.3% of the planned hospital babies. At 5 minutes, 0.7% of both groups had scores below 7.
'Resuscitation' is a term used to refer both to simple suctioning of the nose and mouth, giving oxygen with a bag and mask, or intubation, where a tube is passed into the trachea and oxygen is given under pressure.
Babies planned for home birth were less likely to have any of these interventions. Again, the rate for transfers was slightly higher than the rate for planned hospital births, but the overall planned home birth rate was lower.
|Planned home||Planned hospital|
|Bag and mask||5.6%||9.1%|
The most common birth injury was bruising, which affected 0.6% of the planned home babies (incl. transfers) and 1.4% of the planned hospital babies.
The overall rate of stillbirths and neonatal deaths was very low for the whole study - 5 stillbirths and neonatal deaths in the 5,971 planned home births and 5 in the 4,724 planned hospital births, compared to the national rate of 8 per 1,000 births. The researchers concluded that the death rate in both groups was too low to allow meaningful comparison between the two groups.
Exclusive breastfeeding rates 48 hours after birth were 80% in the home birth group and 58.1% in the planned hospital group; six weeks later, the figures were 65% and 44%. The relaxed nature of a home delivery and the absence of any separation of mother and baby certainly should help to establish breastfeeding.
However, breastfeeding rates for the transferred home births were almost identical to the home deliveries in all categories, including the mother's assessment of whether the first feed went well. The transferred mothers and babies had the most difficult deliveries on average, so this suggests a high level of commitment from these women to breastfeed despite the circumstances. It is possibly another indicator of the hidden differences between the average women planning home births and hospital births, which make it hard to compare like with like.
More to come as I work through the full report!
Home Births - The report of the 1994 Confidential Enquiry
by the National Birthday Trust Fund
Edited by Geoffrey Chamberlain, Ann Wraight and Patricia Crowley
Parthenon Publishing, 1997. Hardback, 293 pages
This book contains not only the study data, but chapters on the history of home birth, an economic evaluation of home births, and more.
Available from ACE Graphics (order from the UK or Australia) (www.acegraphics.com.au)
A summary of the report was published in 'Midwives', May 1997: 'Home Births - a report of the 1994 Confidential Enquiry by the National Birthday Trust' by Ann Wraight, Project Co-Ordinator.
The following paper is a report of the NBTF findings:
Chamberlain G, Wraight A, Crowley P
Pract Midwife 1999 Jul-Aug;2(7):35-9
Recently the National Birthday Trust performed a confidential survey of home births in the United Kingdom. A good response rate was obtained from midwives, who recruited two groups of women prospectively; those planned and accepted as suitable for a home delivery at 37 weeks and a matched group of similar women who were booked for hospital by 37 weeks. Some 16% of such women were transferred to hospital in late pregnancy (4%) or in labour (12%). This figure rose to 40% among the primiparous women in the survey. The survey report presents an analysis of 4,500 home births and 3,300 hospital controls. Outcomes could therefore be presented by the woman's intent or by what actually happened.
"In essence it seems that a woman who is appropriately selected and screened for a home birth is putting herself and her baby at no greater risk than a mother of a similar low-risk profile who is hospital booked and delivered. Home births will probably increase to 4-5% of all maternities in UK during the next decade and this needs preparatory planning."
Read the abstract on Medline
Back to index
On to Page Two of the Research Summaries
Home Birth Reference Page