Too early to question effectiveness of Dutch systemBMJ 2010; 341 doi: https://doi.org/10.1136/bmj.c7020 (Published 07 December 2010) Cite this as: BMJ 2010;341:c7020
- Ank de Jonge, senior midwife researcher1,
- Ben Willem Mol, professor of obstetrics and gynaecology2,
- Birgit Y van der Goes, midwife researcher2,
- Jan G Nijhuis, professor of obstetrics3,
- Joris A van der Post, professor of obstetrics2,
- Simone E Buitendijk, professor of maternal and child health4
- 1Midwifery Science, AVAG and the EMGO Institute for Health and Care Research, VU University Medical Centre, PO Box 7057, 1007 MB Amsterdam, Netherlands
- 2Department of Obstetrics and Gynaecology, Amsterdam Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, Netherlands
- 3Department of Obstetrics and Gynaecology, Maastricht University Medical Centre, P Debyelaan 25, 6229 HX Maastricht, Netherlands
- 4TNO Quality of Life, PO Box 2215, 2301 CE Leiden, Netherlands
The study by Evers et al is the first to show a higher mortality among births that started in primary care compared with secondary care.1 We have concerns about the methods used.
Firstly, although the title suggests that it is a prospective cohort study, the entire population at risk was defined retrospectively and based on postal codes of the catchment area of one university hospital. All intrapartum and neonatal deaths were included from hospitals and midwifery practices in this area, but potentially not all births. Midwives in practices at the periphery of the catchment area will also care for many women in neighbouring regions. These births have not been included in the study unless the baby died. This will artificially inflate mortality in midwifery practices.
Secondly, the study was conducted in only one region in the Netherlands. Intrapartum mortality and neonatal mortality were twice as high as in recent national studies among women in primary care at the onset of labour (1.39 v 0.65 and 0.52 per 1000).2 3 Although classification bias and under-reporting may have played a part in these retrospective studies, half of all deaths are unlikely to have been missed. In another prospective study of perinatal mortality only 3.5% additional cases were found compared with national registration data.4 This discrepancy suggests that the study sample may be rather different from the national population.
Previous audit studies did not find that features of the Dutch maternity care system were related to preventable perinatal deaths.4 5 Given the limitations of the study, the suggestion that the obstetric care system in the Netherlands may contribute to the high perinatal mortality cannot be made on the basis of these data alone.
Cite this as: BMJ 2010;341:c7020
Competing interests: None declared.