The Netherlands always prided itself in its custom of delivery by midwives – widely regarded by the Dutch as more humane and natural. On 2nd of November 2010, BMJ published a paper showing that in an area of The Netherlands, perinatal mortality was twice as high in deliveries started under the supervision of midwives as in those started under the supervision of an obstetrician . On the 14th of December the Minister of Health announced sweeping changes to the organisation of health care during pregnancy and delivery in the Netherlands .
The findings in the BMJ paper rocked Dutch obstetric care to its foundations. Delivery by a midwife, either at home or as an outpatient, has long been the Dutch default option. Midwives select which deliveries they expect to be high risk (multiple pregnancy, previous caesarean, known congenital and placental problems, growth retardation, preeclampsia etc). Those pregnant women are referred to obstetricians and have a fully ‘medicalized’ delivery in-hospital. The remaining pregnancies are regarded as sufficiently low risk for home or outpatient delivery under supervision by a midwife.
The BMJ paper was criticized in an avalanche of letters to the editor, as well as in sharp comments in Dutch newspapers. All kinds of methodologic arguments were raised against the findings: the mortality in the low risk group in the new study would be different from that found in other studies, the study could not correct for confounding, the study was not ‘prospective,’ and the analysis was not prespecified in a protocol (!). All arguments and counterarguments – including my own entry – can be found in the Rapid Responses to the paper in BMJ . My personal view is that, when all the counterarguments are weighed, they do not detract from the results unless a gross calculation error had been made or there was some very strange play of chance.
The most interesting counterargument was that there was no control for confounding. Indeed, the researchers could not stratify the denominator according to delivery and perinatal risk factors. Still, it is obvious that any difference in risk indicators between the two groups would go against the results found: pregnant women with known higher risk are referred to obstetricians. If correction for baseline risk had been possible, the difference would even have been more dramatic.
The new data come against a particular background. Already in 1986, a paper had shown that The Netherlands had dropped in its rank order in European national perinatal statistics . Where once The Netherlands was one of the countries with lowest pregnancy-related baby deaths, by the early 1980s it was amongst the European countries with highest mortality. As in 2010, the 1986 paper was met with a barrage of methodologic criticisms. The main argument, next to arguments about ‘fishing expeditions’, was that registration of perinatal deaths was different and more complete in the Netherlands. How this could account for a shifting rank order always remained a puzzle to me: either the Netherlands would have steadily improved its perinatal death registration (and was originally also amongst the countries with high death rates), or registration of all other European countries would have become considerably sloppier. Anyway, nothing happened.
The poor rank order position of the Netherlands was confirmed in a pan-European survey (PERISTAT-I) in 2003, and swiftly dismissed by the Ministry of Health with the same old arguments. The position of the Netherlands was again confirmed in PERISTAT-II in 2008. By this time, The Netherlands had twice the mortality of leading European countries. In the wake of this confirmation, working groups were finally established. It appeared that about 50% of women who started labour for a first delivery under supervision of a midwife needed to be referred to an obstetrician during the delivery because of acute complications. It also appeared that these transitions were far from smooth. In the recent study the highest mortality and Neonatal Intensive Care admission rates were found in the babies of women who needed urgent referral during labour – despite the fact that these pregnancies had been regarded as low risk throughout the pregnancy and up to the initiation of labour.
These developments seem to conform to the law of ‘de remmende voorsprong’: the law of the ‘retarding lead.’ Coined by the Dutch historian Huizinga in 1937, this notion suggests that being ahead may slow you down in improving further. Decades ago, when medical interventions were crude, a solid force of midwives delivered the country in a more even and resourceful way than may have happened in other countries. But as the medicalization of pregnancy and delivery became more refined, this advantage was lost.
Besides methodologic arguments, current proponents of midwife delivery argue that we should not become preoccupied with the statistics of perinatal deaths, which are only a very few. They argue that generalized care of pregnancies by obstetricians will result in unwanted medicalization (such as having too many Caesareans) and that women will pay the price in subsequent pregnancies. The nightmarish example is, of course, the US: numbers are cited of 30% Caesarians, coupled to poor perinatal outcomes. So, they argue that ‘we should still be proud of home delivery’.
In January 2010, a committee of the Ministry of Health, installed in 2008, had issued a report, written by representatives of midwives (who are largely self-employed) and obstetricians (who have their practices affiliated to hospitals), urging the abolishment of barriers between midwife and obstetric care. In the wake of the publication of the latest study in BMJ  the recommendations of this report have now been urgently accepted by the current Minister of Health in a letter to parliament, signed on the 14th of December 2010 . The report argues for restructuring the process of care so that the central focus is the baby and mother rather than the type of care giver. The report asks for less than 15 mins transit time between midwife and obstetrician ‘for necessary interventions during delivery’. Barriers to cooperation exist not just during the acute delivery phase, but also during pregnancy, with poor referral practices in both directions. While the debate is not yet completely closed, the methodological arguments will slowly be replaced by investigations about practical solutions. The recent data (which, rumour has it, will be confirmed in other Dutch provinces), clearly increased the urgency of the recommendations. Still, the fact is that it has taken three decades for professional organisations to sit together to find solutions based on epidemiologic health care research.
If you like to comment, Email me directly at firstname.lastname@example.org or submt your comment via the journal which requires a password protected login. Unfortunately, comments are limited to 1000 characters.
 Evers AC, Brouwers HA, Hukkelhoven CW, Nikkels PG, Boon J, van Egmond-Linden A, Hillegersberg J, Snuif YS, Sterken-Hooisma S, Bruinse HW, Kwee A. Perinatal mortality and severe morbidity in low and high risk term pregnancies in the Netherlands: prospective cohort study. BMJ. 2010 Nov 2;341:c5639. doi: 10.1136/bmj.c5639.
 Croonen H. Dutch health minister announces plan to reduce country’s perinatal mortality. BMJ 2010 Dec 20;341:doi:10.1136/bmj.c7325
 Hoogendoorn D. Indrukwekkende en tegelijk teleurstellende daling van de perinatale sterfte in Nederland. [[Impressive but still disappointing decline in perinatal mortality in The Netherlands] [Article in Dutch] Ned Tijdschr Geneeskd 1986; 130: 1436-440.
© Jan P Vandenbroucke, 2011