Intended for healthcare professionals

General Practice

Collaborative survey of perinatal loss in planned and unplanned home births

BMJ 1996; 313 doi: (Published 23 November 1996) Cite this as: BMJ 1996;313:1306
  1. Northern Region Perinatal Mortality Survey Coordinating Group
  1. Members of the coordinating group who have served over the past 15 years were: S L Barron, P Blain, C H W Bullough, T Carney, R Gomersall, J Herve, E N Hey, A Irani, F S Johnson, W Lamb, M F Lowry, J B Lawson, R Layton, D Morris, P Morrell, L Parsons, W Reid, A Ryall, R Thomson, M Ward Platt, R G Welch, C Wright, S West, J Wyllie, and G Young. This paper was prepared for the coordinating group by Dr E N Hey.
  1. Correspondence to: The Survey Coordinator, Maternity Survey Office, 25 Claremont Place, Newcastle upon Tyne NE2 4AA.
  • Accepted 8 October 1996


Objective: To document the outcome of planned and unplanned births outside hospital.

Design: Confidential review of every pregnancy ending in stillbirth or neonatal death in which plans had been made for home delivery, irrespective of where delivery eventually occurred. The review was part of a sustained collaborative survey of all perinatal deaths.

Setting: Northern Regional Health Authority area.

Subjects: All 558 691 registered births to women normally resident in the former Northern Regional Health Authority area during 1981–94.

Main outcome measure: Perinatal death.

Results: The estimated perinatal mortality during 1981–94 among women booked for a home birth was 14 deaths in 2888 births. This was less than half that among all women in the region. Only three of the 14 women delivered outside hospital. Independent review suggested that two of the 14 deaths might have been averted by different management. Both births occurred in hospital, and in only one was management before admission of the mother judged inappropriate. Perinatal loss to the 64 women who booked for hospital delivery but delivered outside and to the 67 women who delivered outside hospital without ever making arrangements to receive professional care during labour accounted for the high perinatal mortality (134 deaths in 3466 deliveries) among all births outside hospital.

Conclusions: The perinatal hazard associated with planned home birth in the few women who exercised this option (<1%) was low and mostly unavoidable. Health authorities purchasing maternity care need to address the much greater hazard associated with unplanned delivery outside hospital.

Key messages

  • Only three of 134 deaths were associated with planned home birth

  • Over three quarters of the perinatal deaths asso- ciated with planned home birth occurred in hospi- tal

  • The hazards associated with planned home birth are quantifiable only when death is classified according to the original planned site of delivery

  • Perinatal mortality in the few (<1%) pregnancies in which home birth had been planned was less than half the average for all births, and few of these deaths were associated with substandard care


Home birth is uncommon in the United Kingdom and uncertainty exists about its safety.1 2 Almost all mortality figures available nationally1 provide merely a single global figure for planned and unplanned home births, though the constituent rates differ greatly.3 The only recent figures for planned home birth in England and Wales relating to 19794 and 19935 provide an inaccurately low estimate of risk because it was not possible to account for those mothers who originally booked to have a home delivery but ended up delivering in hospital. This report records the outcome of planned and unplanned births outside hospital to residents in the former Northern Regional Health Authority area between 1981 and 1994.


Records have been kept of every stillbirth and neonatal death to a woman normally resident in the Northern region, irrespective of where delivery took place, since clinicians in the area served by the former Northern Regional Health Authority launched their collaborative maternity survey in the second half of 1980.6 Information was collected on where every woman had initially booked for delivery as well as where delivery took place. Notifications were cross validated against birth and death registration data compiled by the Office of Population Censuses and Surveys (now the Office for National Statistics) and 70 perinatal deaths identified between 1981 and 1994 that did not seem to have been registered as such by local registrars of births, marriages, and deaths; eight were concealed births to women who were never traced. This report uses the pre-1993 definition of stillbirth throughout and is concerned with the pre-1994 regional health authority boundary.

A total of 134 perinatal deaths occurred to women delivering outside hospital between 1981 and 1994 and all were treated as “home” births, though five actually took place in an ambulance, three in another person's house, and two in a general practitioner's surgery; 13 others were to women who were never traced. Additional information was collected on each death, including details of antenatal, intranatal, and postnatal care and results of any necropsy. Every stillbirth or neonatal death to a woman booked for home delivery at any time during pregnancy (irrespective of where delivery actually occurred) was also subjected to independent confidential review by clinicians from a different health district with access to copies of all the relevant unanonymised case records. Using the same approach as currently used in the United Kingdom confidential enquiries into maternal deaths, panels decided whether any aspect of the woman's professional care was substandard and whether any avoidable factor was present (that is, whether the pregnancy might have had a different outcome if a different strategy had been adopted).


Whereas detailed, contemporaneously collected information was available on every death, denominator data were harder to assemble. Information on the total number of births outside hospital was available each year from the Office of Population Censuses and Surveys but it was not known how many of these were planned home births.

Information had been collected retrospectively on a random sample of 100 women delivered outside hospital in 1983 and on all women delivered outside hospital in the region in 1988.7 Contemporaneous data were also collected on every delivery outside hospital during 1993.8 In these three studies only 53%, 55% (132/240), and 44% (142/324) of women delivered outside hospital were actually booked for a home birth when labour began. Women who had not received any medical or midwifery antenatal care and who had made no arrangements for professional care during delivery accounted for 15%, 13% (31/240), and 10% (34/324) of all births outside hospital. The Office of Population Censuses and Surveys recorded a near static proportion of all the region's births as occurring outside hospital in the years covered by the study (average 0.62% compared with 1.2% for the whole of England and Wales).

Estimates of the numbers of women booked for home birth but delivering in hospital were even more difficult to obtain because hospital records do not always specify this information accurately and no national estimate exists.1 4 Data collected in this region in 1983 suggested that 35% of these women changed to hospital based care either before or during labour, and a more detailed prospective study of all planned home births in 1993 found a total transfer rate of 43%.8 Women were classified as having booked for a home birth when a community midwife had accepted a woman for home delivery and had this arrangement accepted by her manager and supervisor of midwives at any stage in pregnancy, irrespective of any later change of plan. Reverse transfers (women arranging to have a home birth after initially making confirmed plans for hospital delivery) were uncommon. The one transfer of this nature associated with perinatal death was grouped with the other booked home births. Perinatal mortality associated with planned home births was calculated with an assumed predelivery transfer rate of 40%. Ford et al found a transfer rate of 19% in one inner city practice that provided substantial medical support for women requesting home birth between 1977 and 1989.9

Spot checks on data provided by the Office of Population Censuses and Surveys showed that official figures underestimated the number of live births and stillbirths outside hospital. Whereas all women who had planned a home birth registered that event as a home delivery, 14% of women who had booked a hospital birth but delivered at home, or before admission, in 1993 registered the birth as occurring in the hospital to which they were admitted after delivery. This happened to some women who had made no delivery plans. No correction was made for this or for other transcribing errors8 (whose net effect would be to lower the estimate of mortality for all women delivered outside hospital in table 1 by 6%) because it was not clear whether a similar degree of underascertainment operated throughout the 14 year study period.

Table 1

Perinatal mortality among 3466 women delivering outside hospital in Northern region, 1981–94

View this table:

Babies classified as born “elsewhere” by the Office of Population Censuses and Surveys were included in the totals recorded as born outside hospital. Most of the mothers had delivered on their way to hospital, and a few had delivered at the house of a friend or relative. None had delivered in a private maternity unit or in a psychiatric, remand, or penal institution.


Between 1981 and 1994 inclusive 3466 births and 134 perinatal deaths were reported as having occurred outside hospital. Mortality was four times as high as for all registered births (38.7 v 9.7 deaths/1000 births). An estimate of the contribution made by unbooked and hospital booked births confirmed that these were much more hazardous than planned home births (table 1).

Table 2 shows the estimated perinatal mortality for the women booked for home delivery. Reviews that ignore antepartum and intrapartum transfer seriously underestimate the risk of perinatal loss. Such an analysis does not, however, establish how many of these deaths could relate to the management of delivery itself.5 10 For this the analysis shown in table 3 is more appropriate. Over the whole 14 years the risk of death during delivery or in the first four weeks of life in a baby of normal birth weight and without a lethal malformation was higher in those born to the small group of women who had booked for home delivery. However, during the last 10 years of that period, when the midwife was always the community lead professional, mortality in this subgroup was lower in those booking for home delivery (1/1890 v 410/370 722). Neither difference was statistically significant.

Table 2

Perinatal mortality among mothers booked for home confinement in Northern Region, 1981–94

View this table:
Table 3

Rates for intrapartum stillbirth and neonatal death in babies of 2500 g or more without lethal malformation in Northern region, 1981–94

View this table:

No such analysis can establish how many of these deaths were potentially avoidable. Details of the deaths are summarised in table 4. Confidential inquiry identified three deaths in which issues of substandard care were raised and two in which a different line of management might have produced a different outcome. In one case this entailed aspects of care before hospital admission. There were also two home births in which more regular intrapartum monitoring might have shown some heart rate abnormality in the absence of any other sign of trouble. Both deliveries occurred before 1984. Delays occasioned by the need to arrange and effect transfer probably contributed to only one death. No fault was found with midwifery care in any death.

Table 4

Stillbirths and neonatal deaths to mothers booked for home birth in Northern region, 1981–94

View this table:


The debate about planned home birth centres not on whether it is safe but on whether it is less safe than hospital birth. That issue is now almost unaddressable in Britain because those women booking for home birth are not comparable with those booking for hospital delivery. Any matching process is fraught with uncertainty2 11 and any formal comparative trial impracticable.12 Even if a trial was done it could not give a generalisable outcome—that is, the result would not necessarily be the same in a different setting. All we can say with certainty is that, of the 1890 women who were estimated to have booked for home delivery in this region in the last 10 years of the study period, only five lost a baby and intrapartum events were implicated in only one of those deaths.

During that time the death rate in labour or the neonatal period in non-malformed babies of normal birth weight born to women booked for a home delivery (those deaths most capable of reduction by high quality care during labour) was as low as the regional figure for all other such losses (0.05% v 0.11%). This contrasts with the outcome of a national analysis of such births,5 in which some women resisted intrapartum transfer when problems arose (but in which some deaths may have gone unanalysed). Studies in Australia,13 Canada,14 and the United States15 16 have concluded that in some settings midwife managed home birth can be associated with as low a perinatal mortality as hospital birth for low risk women, reviving the debate over the need to allow women genuine choice.17 18

Perinatal loss is only one issue that needs to be taken into account when considering home birth, and the fact that very few babies died does not of itself show that arrangements for home birth were necessarily safe. Nevertheless, women wanting a home birth will take heart from these figures. Such results were achieved only by vigilance, ready access to hospital services, appropriate and timely transfer when problems arose during either pregnancy or labour, and by the readiness of both midwives and mothers to contemplate transfer promptly once problems were identified.8

That half the women delivered outside hospital in this region between 1981 and 1994 had not booked to have a home delivery underlines the importance of accepting that maternity services have to be planned on the assumption that some women will deliver in the community whether we (or they) like it or not. A service geared to cope with these unplanned events ought to be able to deal with a proportion of planned low risk deliveries. The estimates in table 1 are a sobering reflection of the perinatal hazards that these women face,3 4 even if the exact rates have been exaggerated by some underas-certainment of the relevant official denominator figure for all births outside hospital and are subject to uncertainty because sampling methods had to be used to aportion the overall figure.

The number of home births in the Northern region is currently very low (0.9%). The rate is much the same as in Scotland but lower than in any other area of England and Wales. Numbers could well increase, however (as they already have in parts of southern England),8 once women start to exercise the “choice in childbirth” envisaged by the government's endorsement19 of the Cumberlege report.20 More women could almost certainly be delivered outside hospital with equal safety (given that the obstetric “profile” of many women booking for hospital delivery was no different from that of those initially booking for delivery at home), but whether the community midwifery service could at the moment and within its current budget respond to any rapid rise in the number of women wanting a home birth is less certain. A study of 1005 United Kingdom mothers for the Department of Health in 1993 indicated that 22% would have liked the opportunity to consider home birth,21 and several studies suggest that 10% of women might request such an option were it available and considered safe (a proportion that does not seem to have varied appreciably over the past 20 years).22 23 24 25 26 27 28 29 30

The group acknowledges the crucial contribution the many district convenors and their colleagues have made to the survey voluntarily over the years and are grateful to Mr R A McNay (regional statistician) and Mrs M Renwick (survey coordinator) for administrative help.


  • Funding Survey office funding came from the Northern Regional Health Authority.

  • Conflict of interest None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.