Estimating intrapartum-related perinatal mortality rates for booked home births: when the ‘best’ available data are not good enough


G Gyte, Research Associate, Cochrane Pregnancy and Childbirth Group and NCT Research Networker, Division of Perinatal and Reproductive Medicine, University of Liverpool, Liverpool Women’s NHS Foundation Trust, Crown Street, Liverpool L8 7SS, UK. Email


Objective  To critically appraise a recent study on the safety of home birth (Mori R, Dougherty M, Whittle M. BJOG 2008;115:554) and assess its contribution to the debate about risks and benefits of planned home birth for women at low risk of complications.

Design  Critical appraisal of a published paper.

Setting  England and Wales.

Population or Sample  Home births from 1994–2003 and all women giving birth in the same time period.

Methods  Six members of a multidisciplinary group appraised the paper independently. Comments were collated and synthesised.

Main outcome measures  Assessment of: overall methodology; assumptions used in estimating figures; methods used for calculations; conclusions drawn from the results and reliability and consistency of data.

Results  Although there were some positive aspects to the study, there were weaknesses in design and an inaccurate estimate of risk. Our evidence suggests that the conclusions drawn did not reflect the results and the methodological weaknesses found in the study rendered both the results and conclusions invalid.

Conclusions  On the basis of our critical appraisal, the study does not contribute to the existing evidence about the safety of home birth to inform decision-making or provision of care. The limitations could have been identified by the peer review process and the problems were compounded by an inaccurate press release. Great care needs to be taken by journals to ensure the accuracy of information before dissemination to the scientific community, clinicians and the public. These data should not have been used to inform national guidelines.