The study compares ‘like with like’ by focusing on ‘low-risk’
women, i.e. only those considered most suitable for midwifery-led,
community-based care for pregnancy and birth. Any other
differences among the groups of mothers in terms of age,
socio-economic background, education, etc. which might
affect outcomes are accounted for statistically with
appropriate weighting. There was some additional
analysis to further correct for differences in complicating
factors identified at the start of labour care. The substantially
higher odds of interventions being seen in obstetric units remained.

The Birthplace study compared outcomes by where women
planned
to give birth, not by where they actually gave birth, which
makes sure that the obstetric unit group were not disproportionately
made up of those with complications. Outcomes for women and babies
transferred to hospital from home or midwifery units during labour
(and immediately afterwards) were analysed in the ‘planned home birth’
and ‘planned midwifery unit’ groups. So the number of interventions
in midwifery units and home births do not represent births
that started out as being clinically more straightforward.