By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Dr M Habiba, Reproductive Sciences Section, Department of Cancer Studies and Molecular Medicine, University of Leicester, Robert Kilpatrick Building, Leicester Royal Infirmary, PO Box 65, Leicester LE2 7LX, UK. Email firstname.lastname@example.org
Objective To explore the attitudes of obstetricians to performe a caesarean section on maternal request in the absence of medical indication.
Design Cluster sampling cross-sectional survey.
Setting Neonatal Intensive Care Unit (NICU) associated maternity units in eight European countries.
Population Obstetricians with at least 6 months clinical experience.
Methods NICU-associated maternity units were chosen by census in Luxembourg, Netherlands and Sweden and by geographically stratified random sampling in France, Germany, Italy, Spain and UK. An anonymous, self-administered questionnaire was used for data collection.
Main outcome measures Obstetricians’ willingness to perform a caesarean section on maternal request.
Results One hundred and five units and 1530 obstetricians participated in the study (response rates of 70 and 77%, respectively). Compliance with a hypothetical woman’s request for elective caesarean section simply because it was ‘her choice’ was lowest in Spain (15%), France (19%) and Netherlands (22%); highest in Germany (75%) and UK (79%) and intermediate in the remaining countries. Using weighted multivariate logistic regression, country of practice (P < 0.001), fear of litigation (P= 0.004) and working in a university-affiliated hospital (P= 0.001) were associated with physicians’ likelihood to agree to patient’s request. The subset of female doctors with children was less likely to agree (OR 0.29, 95% CI 0.20–0.42).
Conclusions The differences in obstetricians’ attitudes are not founded on concrete medical evidence. Cultural factors, legal liability and variables linked to the specific perinatal care organisation of the various countries play a role. Greater emphasis should be placed on understanding the motivation, values and fears underlying a woman’s request for elective caesarean delivery.
Caesarean section performed at patients’ request in the absence of a medical indication is a focus of considerable attention both for clinical and ethical reasons.1–5 The reported rates range from 2.6% in Flanders6 to 26.8% in Western Australia.7 Different sampling frameworks and the inconsistent use of the term ‘on demand’ may, at least partially, explain such wide variation.8,9 Nevertheless, there are indications that maternal request is becoming increasingly relevant in situations where medical justification may not—in itself—be sufficient to recommend a caesarean delivery.10–12 The caesarean section rate is rising in most developed countries, and evidence from North Carolina suggests that the rising primary caesarean section rate is not explained by changes in patients’ characteristics.13 A growing awareness of consumer preferences is said to play a major role.14 However, physicians’ attitudes can significantly influence or motivate patients’ choice,15,16 a point which acquires prominence in the light of the evidence that a significant proportion of obstetricians in the USA (46%) and of female obstetricians in London (31%) would favour a caesarean section for themselves or for their partners in an uncomplicated pregnancy.17,18 Yet, this view was shared by only 16% of Scottish female obstetricians,19 15% of female UK trainees,20 7% of obstetricians in the Republic of Ireland,21 2% of Norwegian22 and Flemish6 obstetricians and by only 1.4% of those in the Netherlands.23
That a proportion of obstetricians would prefer a caesarean section for themselves or for their partner may reflect their perception of its overall safety or other advantages compared with vaginal delivery.24 There is also an evidence of an increased willingness of obstetricians to accept their patients’ request for a caesarean delivery in the absence of ‘mitigating’ circumstances. When surveyed, 69% of consultant obstetricians in England and Wales indicated that they would perform an elective caesarean birth on maternal request due to fear of litigation and pressure from the patients,25 and approximately 50% of obstetricians in Israel were willing to perform a caesarean section on request in support of patient’s autonomy.26 That this be the case despite the generalised concerns about the rising caesarean section rates represents a significant shift from the position held by most obstetricians two decades ago.27 The exact reasons for this change remain to be explored.
Data that allow direct comparison between countries and provide insight about the factors influencing clinical decision making are scarce. A European multicenter study project (EUROBS) on ethical issues in pre- and perinatal care offered the opportunity to describe the attitudes of a large representative sample of obstetricians in eight European countries towards a request for caesarean delivery in the absence of compelling medical reasons and to explore the possible underlying factors.
Eight European countries took part in the EUROBS project (Developments of Perinatal Technology and Ethical Decision-Making during Pregnancy and Birth: the Obstetricians’ Perspective): France, Germany, Italy, Luxembourg, Netherlands, Spain, Sweden and UK. In every country, only maternity units associated with a third-level Neonatal Intensive Care Unit (NICU) were sampled. The cluster sampling strategy mirrored the one adopted in a previous European study on ethical issues in neonatal medicine (EURONIC).28,29 In Luxembourg, The Netherlands and Sweden, all the existing NICU-associated maternity units were recruited (census sampling); whereas in France, Germany, Italy, Spain and UK, random samples stratified by geographical area were selected. In every recruited unit, all physicians with at least 6 months experience in obstetrics were invited to participate in the study.
Data collection took place in 2001–02. Structured questionnaires were used to record information on unit organisation and policies and to survey the obstetricians’ practices and attitudes in six major areas: prenatal ultrasound examination, late termination of pregnancy, management of severe prematurity, situations of conflicting opinions between staff and the women, legal concerns in obstetric practice and maternal request for caesarean section. The head or the clinical director of each unit was contacted inviting the unit to participate and, in case of agreement, to identify a local study coordinator. The local study coordinator undertook the distribution of the staff questionnaires to eligible obstetricians. Completed questionnaires were returned in sealed envelopes to the coordinating centre in Italy either directly or via the local coordinator.
The staff questionnaire was anonymous and self-administered to protect confidentiality. Thus, nonresponders were not identified and no reminders were sent. It was developed in English and translated into national languages; a back-translation into English was performed to check the accuracy of the translation and ensure identical semantic content. A pilot study was carried out to assess the questionnaires’ readability and suitability to the different national contexts.
This study explores obstetricians’ attitudes to a woman’s request for caesarean delivery in case of uncomplicated term pregnancy as an area of potential conflict between staff and women. Respondents were asked to consider whether or not they would perform a caesarean section on a 25-year-old woman, with a singleton uncomplicated pregnancy in cephalic presentation at 39 weeks of gestation, who requests caesarean section despite the obstetrician’s recommendations for vaginal delivery and despite her being informed of the higher morbidity and mortality associated with surgical birth. Respondents were also asked to consider their response in case the patient’s preference was based on one of the following distinct scenarios: a) her personal choice, b) fear of vaginal delivery, c) previous caesarean section, d) previous traumatic vaginal delivery, e) previous intrapartum fetal death, f) that her first child was disabled and g) the patient were a colleague. Respondents who indicated compliance with maternal request based on any of these reasons were then asked to clarify their rationale by selecting one or more of the following nonexclusive options: a) out of respect for the woman’s autonomy, b) to avoid possible problems with noncompliance during delivery and c) to avoid possible legal consequences in case something goes wrong. Space was allowed to write in additional answers beyond the preset ones. Different sections of the questionnaire addressed doctors’ demographic and professional characteristics and their perception of the effect of litigation on their practice. The latter question aimed at establishing the number of obstetricians who felt that their personal medical practice was influenced ‘often’, ‘occasionally’ or ‘not at all’ by fear of litigation.
Questionnaire coding and data entry were performed at the coordinating centre in Italy. Statistical analysis was carried out with the STATA statistical package (version 8.0).30 Weights, computed as the inverse of the probability for a given maternity unit to be selected within a certain country and geographical stratum, were applied to take into account the different sampling fractions adopted in the participating countries.28,29,31 Standard errors were adjusted for intracluster correlation, that is the nonindependence of observations within the same maternity unit.31
Unless stated otherwise, results are presented as weighted proportions and 95% CI. Multivariate logistic regression analysis was used to explore factors associated with an obstetrician’s agreement with the request for caesarean delivery solely because it was the woman’s personal choice. The doctor-related variables considered as predictive factors included age, sex, having had children, current religion and religiousness, length of experience in obstetrics, involvement in private practice, involvement in research and self-reported feelings that fear of litigation influences one’s clinical practice. The unit-related variables included number of deliveries per year, number of obstetric beds, unit proportion of caesarean deliveries, whether or not the hospital and/or the unit was university affiliated and whether or not the unit was a referral centre for high-risk pregnancies. The final model retained the variables significantly associated with the outcome of interest at the 0.05 level. A statistically significant interaction (P= 0.01) was found between the obstetricians’ gender and having had children; therefore, the odds ratios for women versus men were reported separately according to whether or not they had children.
One hundred and forty-nine units were invited to participate in the study and 105 agreed, corresponding to a unit response rate of 70% (Table 1). Completed questionnaires were returned from 1530 obstetricians, with an overall staff response rate of 77% (ranging from 63% in Germany to 93% in France). The number and main socio-demographic and professional characteristics of the participating doctors are shown in Table 2. In every country except Italy, France and Luxembourg, about half of the respondents were women. In Italy and Sweden, 40% of respondents were aged 50 years or more, compared with only 7% in the UK. In all countries except Luxembourg, most physicians were working full time in the hospital; in Italy, Spain and Luxembourg, a large proportion of respondents also had a private practice either within the same hospital or outside it.
Table 1. Sample size and response rate
Response rate (%)
Response rate (%)
Table 2. Socio-demographic and professional characteristics of responding obstetricians (unweighted proportions)*
Italy, n (%)
Spain, n (%)
France, n (%)
Germany, n (%)
Netherlands, n (%)
Luxembourg, n (%)
UK, n (%)
Sweden, n (%)
Within countries, discrepancies between totals and number of responding obstetricians as presented in Table 1 are due to missing values.
Younger than 30
50 and more
Having had children
Length of experience in obstetrics (in years)
Less than 1
More than 5
Full-time work in the hospital
Private practice activity
Table 3 shows the proportion of obstetricians who, in each country, would perform a caesarean section at term for a 25-year-old healthy woman in the absence of strict medical indications. Compliance with this hypothetical woman’s request simply because this was ‘her choice’ was lowest in Spain (15%), France (19%) and The Netherlands (22%); highest in Germany (75%) and UK (79%) and intermediate in the remaining countries. Being a colleague did not significantly alter physicians’ willingness to perform a caesarean section; the same holds for fear of vaginal delivery, with the exception of Sweden where the proportion of respondents accepting the woman’s request raised from 49 to 79%. In every country, doctors said that they were more likely to comply with the woman’s request in case of a previous caesarean section, traumatic vaginal delivery, intrapartum death or if the first child was disabled. Although a shift in opinion in favour of caesarean delivery was observed in all countries given these mitigations, the differences between countries remained. Only in Spain and France, a sizeable proportion of physicians (33 and 16%, respectively) would in any case reject the woman’s request for a caesarean delivery.
Table 3. Respondents’ attitudes towards a request for caesarean delivery for an uncomplicated term pregnancy (weighted proportions)*
Italy, % (95% CI)
Spain, % (95% CI)
France, % (95% CI)
Germany, % (95% CI)
Netherlands, % (95% CI)
Luxembourg, % (95% CI)
UK, % (95% CI)
Sweden, % (95% CI)
A 25-year-old pregnant woman starts labour at 39 completed weeks. The fetus was apparently normally formed, healthy and in cephalic presentation. Despite being informed that a vaginal delivery is indicated, and of the higher morbidity and mortality associated with caesarean delivery, the woman insists on a caesarean section.
Proportion of physicians who would comply with this woman’s request for a caesarean delivery for each of the following reasons
1. This is her choice
2. Fear of vaginal delivery
3. Previous caesarean section
4. Previous traumatic vaginal delivery
5. Previous intrapartum death
6. Her first child is disabled
7. This patient is a colleague
Proportion of physicians who would in any case refuse to comply with this woman’s request for a caesarean delivery
Respect for the woman’s autonomy was the most frequently quoted reason for performing a caesarean section simply because it was the patient’s choice (Table 4). However, prevention of possible legal consequences linked to complications of vaginal birth was also mentioned by more than 50% of respondents in every country except The Netherlands and Sweden.
Table 4. Reasons for supporting this woman’s choice for a caesarean delivery independently from other medical or nonmedical indications (weighted proportions)
Italy, % (95% CI)
Spain, % (95% CI)
France, % (95% CI)
Germany, % (95% CI)
Netherlands, % (95% CI)
Luxembourg, % (95% CI)
UK, % (95% CI)
Sweden, % (95% CI)
Proportions computed on physicians who would perform a caesarean section simply because it was the woman’s choice. Answers were not mutually exclusive.
Proportions of physicians indicating the following reasons*
Out of respect for the woman’s autonomy
To avoid possible problems of noncompliance during delivery
To avoid possible legal consequences if something goes wrong
Obstetricians were asked in a different section of the questionnaire whether they felt that their personal medical practice was influenced by fear of litigation, and their answers are shown in Figure 1. In every country, more than 50% of respondents answered that ‘occasionally’ this was the case. Fear of litigation was quoted as ‘often’ influencing decision making more frequently in Italy and Spain, while the opposite was true in Sweden and The Netherlands.
Table 5 shows the results of the multivariate logistic analysis exploring the factors associated with obstetricians’ likelihood to comply with the woman’s request for a caesarean section purely because it was her choice. Differences between countries remained statistically significant after potential confounders were controlled for. Caesarean section on demand appeared more likely to be accommodated in the UK and Germany and less likely in Spain, France and The Netherlands. Physicians from university-affiliated units were twice as likely to perform a caesarean section on demand. In contrast, being a female doctor decreased the probability of accepting a patient’s request for a caesarean section, but the effect was statistically significant only among physicians who had children themselves (OR 0.29, 95% CI 0.20–0.42). A consistent, statistically significant trend emerged between obstetricians’ self-reported feeling that their clinical practice was influenced, occasionally or often, by fear of litigation and the willingness to perform a caesarean delivery at the patient’s request. Factors not associated with the likelihood of agreeing to a woman’s request for a caesarean section were the doctor’s age, current religion and religiousness, length of experience in obstetrics, involvement in private practice or research, the unit’s annual delivery rate, the number of obstetric beds, the unit’s caesarean rates and whether the unit was a referral centre for high-risk pregnancies (data not shown).
Table 5. Predictors of obstetricians’ reported decision to perform caesarean section because it is the woman’s choice
Odds ratio are adjusted for all the variables listed herein.
P values refer to the overall statistical significance of the association between the explanatory variable and the outcome.
Feeling that fear of litigation influences one’s clinical practice
This study presents the findings of a large representative sample of obstetricians from NICU-associated maternity units in eight European countries. The census or the random sampling strategy that was adopted and the overall high physicians’ response rate within the recruited units (77%) support the validity of our data, although in some countries, a lower unit recruitment fraction might have impaired the representativeness of the results at national level.
When confronted with a woman’s request for a caesarean delivery in the absence of clinical indication, the attitude of European obstetricians varied within as well as between countries. Obstetricians from Spain, France and The Netherlands were the least likely and those from UK and Germany the most likely to accept a request for a caesarean section based exclusively on patient’s choice. International variability was not explained by differences in physicians’ demographic and professional characteristics or by characteristics of the units. Thus, it is likely that cultural factors, as well as elements related to the national organisation of perinatal care, play a role. Differences in physicians’ attitudes across countries have been reported also by other studies. Doctors from Greece, Spain and Italy appeared less likely and those from the Netherlands most likely to give complete information in relation to intensive care interventions.32 Neonatologists were less likely to involve parents in decisions concerning their preterm babies in the NICUs of France, Italy and Spain, compared with the northern countries such as Sweden and UK.28 Medical paternalism is said to be a more dominant tradition in France, while in the Anglo-Saxon cultures, patient’s consent is given more prominence.33 These patterns may partially explain our findings in Spain and France on the one hand and the UK on the other hand. In the Netherlands, a very low national caesarean section rate34 is coupled with a policy of prenatal care delivered, in case of low-risk pregnancy, by GPs and midwives rather by obstetricians:23: a pattern consistent with the lower agreement with caesarean on demand documented in our study as well as in others.23,35
Obstetricians’ attitudes, however, are also dependent on the clinical question asked. According to our data, in most countries, a maternal request for a caesarean delivery appeared more likely to be accepted when backed up by a medical or quasi-medical justification, even when questionable from an evidence-based perspective (e.g. previous caesarean section, traumatic delivery, intrapartum death or disabled child). Physicians’ acceptance of such reasoning seemed to be high in every country except Spain. In the Netherlands, the resulting change in attitudes was quite remarkable, suggesting that a shift from the ‘normal pregnancy’ paradigm moved the woman’s request to a different level of consideration. In Spain, however, despite the overall high caesarean section rate,34 none of the mitigating quasi-medical indications was sufficient to persuade a significant proportion of doctors to accept a caesarean delivery on request. The marked contrast between Spain and other countries such as UK and Germany is unlikely to be explained by differences in the interpretation of medical evidence; rather, it seems to emphasise the value attributed to patient’s choice.
Although ‘country’ is an important explanatory variable for doctors’ attitudes towards maternal choice, responses within countries were not uniform, pointing to the role of unit policies or even practitioners as individuals. Physicians’ age as well as other personal and professional characteristics such as religion, length of clinical experience, private practice and involvement in research were not significant explanatory variables. Female gender was only associated with a lower likelihood of accepting the woman’s request for a caesarean delivery among female doctors who themselves had children: a finding which might explain the inconsistencies that emerged in previous studies between the obstetrician’s gender and the preferences for caesarean delivery for themselves or for their partners.18,20,21
In this study, personal financial incentives did not seem to influence doctors’ decision with regards to caesarean section on demand. Obstetricians from the three countries with the highest percentage of respondents engaged in private practice (Italy, Luxembourg and Spain) had an average or low acceptance for caesarean section on request, whereas in countries with the highest acceptance rate for caesarean on demand, a smaller percentage of obstetricians engaged in private practice. Also, in multivariate analysis, private practice was not a significant predictive variable.
In contrast, a physician’s self-reported feeling that ‘fear of litigation influences his/her clinical practice’ emerged as the variable more strongly and consistently associated with willingness to perform a caesarean delivery on maternal request, with odds ratios raising from 1.84 for those reporting ‘occasional’ influence to 3.01 if it was ‘often’. Although respect for patient’s autonomy was the most widely quoted reason by physicians’ willing to accept the woman’s request, a relevant proportion (from about 30% in the Netherlands and Sweden to more than 80% in Spain, Luxembourg and France) also mentioned the wish to avoid legal consequences in case of complications during vaginal birth. Taken together, these findings indicate that the issue of respect for autonomy may be compounded by the practice of defensive medicine, which has been linked to higher caesarean section rates in some countries,36–38 although supportive evidence remain controversial.36,39,40
Fear of litigation appeared less relevant to physicians’ decision making in Sweden and The Netherlands, a finding consistent with the low medico-legal burden in these countries. Yet, there was an important difference between the two countries with regards to accepting the patient’s ‘choice’ of a caesarean section. Clearly, several factors interact to determine obstetricians’ attitudes towards caesarean section on demand in various countries; the relative weight of these factors seems to be modulated by the specific national or regional characteristics of medical practice and perinatal care organisation.
A study published in 1986 found that only 2% of US obstetricians would agree to carry out a caesarean section on a primigravid woman at term with no other medical problems and that a higher risk of litigation swayed 3% towards performing a caesarean section.27 A later study conducted in 2000 reported that most practitioners in the Portland, Oregon metropolitan area, would not perform a caesarean section based on patients’ request or in the absence of clear medical indication.41 In the study by Ghetti et al.,41 only 9% of female responders and 29.2% of male responders said that they would consider a caesarean section for fear of urinary incontinence after vaginal delivery. This, however, contrasts with the higher acceptance of maternal request reported by Kenton et al.,42 where 67% of recently trained obstetricians and gynaecologist were willing to perform a primary caesarean section specifically to prevent pelvic floor disorders. The US findings that age, years in practice and practice type were not associated with willingness to perform caesarean section on request41,42 are in agreement with the findings in our study.
Many publications have mentioned about the risk of further increase in caesarean section rate should caesarean section on demand become widespread. So far, however, the relationship between obstetricians’ willingness to perform a caesarean section based on patient’s request and caesarean section rates in the respective countries has been inconsistent. Spain, where our study has shown a low acceptance of maternal request, has the second highest caesarean section rate after Italy.34 The UK, which has the highest level of acceptance for maternal request, ranks fourth with regards to its caesarean section rate.34 On the other hand, The Netherlands features both a low caesarean section rate and a low acceptance of maternal request.4 Up to now, the proportion of pregnant women who actively ask for and are granted an elective caesarean section has probably been quite limited in Europe.16 Should such demand increase, it is hard to believe that we would not witness a rise of caesarean section rates. In that situation, issues of costs and resource allocation should also be considered.43
Traditionally, caesarean delivery has been associated with a higher maternal mortality and morbidity, both short and long term, and with neonatal respiratory distress.2,5,44 However, most of the published literature is based on women who required a caesarean section because of medical reasons,45 and it is difficult to disentangle the adverse effects of surgery from those of its underlying indications. Advances of medical care have certainly rendered caesarean delivery safer.37 At the same time, medicalisation of vaginal birth has been linked to adverse effects such as pelvic floor injury and its sequelae, which are now quoted as major reasons in favour of caesarean delivery.46 Reliable data allowing an unbiased comparison of maternal and fetal benefits and risks, both short and long term, of planned caesarean versus vaginal delivery are scarce.5,45,46 Thus, the elements for evidence-based decision making about the most appropriate way of delivery for the individual, healthy woman who request an elective caesarean section are lacking.
In principle, the existence of clinical uncertainty about alternative treatment strategies makes a good case for allowing the patient’s preferences to prevail.39 However, compliance with patients’ request for caesarean delivery has been criticised as the easy short cut, which exempts physicians from dealing with the anxiety associated with childbirth and providing proper counselling and support.47 Ultimately, the ‘choice’ in favour of caesarean section might be disempowering to women.48 Against this background, it is interesting that in both our study and that by Ghetti et al.,41 female obstetricians were less likely to agree with the woman’s request, particularly when they had children themselves.
Apart from financial incentives, obstetricians may see other advantages in elective caesarean section, such as better control of the birth process and timing of delivery.15,49 The relationship between the physicians’ attitudes towards caesarean on demand and the fear of litigation, highlighted in ours as well as in previous studies,25,35 represents an additional warning. So far, litigation has largely centred on fetal damage and on withholding rather than performing medical procedures.47–49 Thus, elective caesarean birth seems less risky than labour from a liability standpoint.49
One potential criticism of our study is the choice of the cutoff point for inclusion, which allowed the inclusion of obstetricians who may not themselves be finally responsible for making such controversial decisions. However, as the questions posed addressed values and beliefs, we were interested in the views of all respondents (of whom only 71 had between 6 and 12 months of experience in obstetrics). The choice of 12 months experience as a cutoff point did not alter the overall results, and furthermore, neither the obstetricians’ age nor the length of experience were significant predictors in the multivariate analysis.
Despite the appeal of autonomy and ‘consumer’s ethics’,14 women are still highly dependent on the information provided by their physician. Rather than expressing a free, informed choice for caesarean delivery, they may become trapped between the obstetricians’ attitudes,15 lack of opportunities for nonmedicalised vaginal birth16 and media-publicised fashionable trends featuring vaginal delivery as unsafe, archaic, disfiguring and ultimately socially unacceptable.48 The current controversy surrounding this topic suggests the need for further research to better characterise the cost-benefit balance and risks of caesarean section in uncomplicated pregnancies compared with truly nonmedicalised spontaneous vaginal birth. Both short- and long-term outcomes should be considered.
Until better evidence becomes available, individual obstetricians faced with a request for elective caesarean delivery are charged with the delicate task of fostering their patient’s autonomy and freedom of choice by exploring the motivation and fears underlying such request5,45,46 and ultimately act according to what they believe to better promote the health and welfare of mother and fetus.50
We acknowledge the contribution of Michael Hills as statistical advisor and of the IRTEF Institute in data management. We are very grateful to our colleagues who answered our questionnaire.
Conflict of interest
The results presented in this study are part of the European Concerted Action project EUROBS on ‘Developments of perinatal technology and ethical decision-making during pregnancy and birth: the obstetricians’ perspective’, funded by the European Commission (Contract no. BMH4-CT98-3376, Project coordinator: Marina Cuttini, IRCCS Burlo Garofolo, Trieste). The views expressed are those of the authors and the sponsor has played no part in study design, execution, analysis, preparing or reviewing the manuscript. The decision to submit for publication it that of the authors and the corresponding author. Dr M.C acts as custodian for the data.
Other members of the EUROBS study group
C Arnaud and M Garel (France); P Benciolini, C Viafora and R Saracci (Italy), Manuel Marín Gómez (Spain), G Lingman and T Nilstun (Sweden), I de Beaufort (Netherlands).
List of participating obstetrics units
• France Hôpital Maison Blanche, Reims (R Gabriel, C Quereux); Centre Hospitalier Universitaire, Amiens (J Gondry, JC Boulanger); Centre Hospitalier Universitaire Jean Bernard, Poitiers (G Magnin); Centre Hospitalier, Saint Brieuc (B Cloup, A Renaud-Giono); Centre Hospitalier Universitaire, Caen (M Dreyfus, M Herlicoviez); Centre Hospitalier Universitaire, Rouen (L Marpeau); Hôpital Antoine Béclère, Clamart (F Audibert, R Frydman); Hôpital Robert Debré, Paris (P Blot); Hôpital Nord, Saint Etienne (MN Varlet, P Seffert); Hôpital François Mitterand, Pau (C Belcikowski, M Chevalier); Hôpital Bretonneau, Tours (J Lansac); Maternité Port-Royal, Paris (D Cabrol); Hôpital Jeanne de Flandre, Lille (F Puech).
• Germany Marienkrankenhaus, Hamburg (U Blasshof, P Scheidel); Städt. Krankenhaus, Lüneburg (E Walbrodt, J Gille); Städt. Krankenhaus Köln-Holweide, Köln (U Schellenberger, F Wolff); Klinikum Frauenklinik, Bamberg (J Peisl, R v Hugo); Universitätsklinik, Göttingen (W Heyl, G Emons); Universitätsklinikum Virchow Klinikum, Berlin (U Büscher, JW Dudenhausen); Universitätsklinikum, Dresden (A Riehn, W Distler); Universitätsklinikum, Leipzig (B Viehweg, M Höckel); Universitäts-Frauenklinik, Münster (W Klockenbusch, L Kießl); Frauenklinik der Universität, Gießen (M Zygmunt, W Künzel); Frauenklinik der Universität, Mainz (P Brockerhoff, PG Knapstein); Frauenklinik der Universität, München (F Kainer, G Kindermann); Universitätsklinikum, Tübingen (B Schauff, D Wallwiener); Städt. Krankenhaus München-Schwabing, München (EM Grischke); Städt. Krankenanstalten, Krefeld (W Poleska, J Baltzer).
• Italy Ospedale Santa Croce, Moncalieri To (ME Renzetti, R Monti); Ospedale Niguarda Ca’ Granda, Milano (A Ragusa, S Garsia); Ospedale di Circolo e Fondazione Macchi, Varese (P Clerici, G Maffioli, D Balestreri, PF Bolis); Azienda Ospedaliera S Anna, Como (G Bonifacino, F Colombo); Ospedale Bolognini, Seriate Bg (G Palmerio, LD Moretti); Ospedale Policlinico G B Rossi, Verona (E Zardini, D Pecorari, V Silvestre, L Fedele); Università degli Studi di Padova, Padova (DM Paternoster, A Ambrosiani, F Lauri, S Mazzer, M Rondinelli, P Grella); Istituto per l’Infanzia Burlo Garofolo, Trieste (S Guaschino); Policlinico Universitario di Udine, Udine (F Petraglia); Presidio Ospedaliero di Gorizia, Gorizia (C Gigli); Policlinico S Orsola-Malpighi, Bologna (LF Orsini, D De Aloysio); Ospedale Maggiore, Bologna (M Lenzi, C Melega); Ospedale S Maria Annunziata, Firenze (C Campatelli, Gaggi); Ospedale Civile Spirito Santo, Pescara (V Palladoro, R Lotti); Presidio Ospedaliero di Belcolle, Viterbo (G Palla); Policlinico Umberto I, Roma (M Anceschi, EV Cosmi); Ospedale SS Annunziata, Napoli (C Picardi, R Arienzo); Azienda Ospedaliera S Giovanni di Dio e Ruggi d’Aragona, Salerno (C Lomiento, A Fasolino, De Angelis, R Quirino); Ospedale A Perrino, Brindisi (S Burlizzi, ER Poddi); Azienda Ospedaliera Vito Fazzi, Lecce (F Totaro Aprile, A Perrone, FG Tinelli); Ospedali Riuniti di Foggia, Foggia (Maruotti, F Pietropaolo, G Arciuolo, P Lauriola, C Napolitano, CM Troysi); Ospedali Riuniti Bianchi Melacrino Morelli, Reggio Calabria (N Bitto, PF Tropea); Presidio Ospedale Civico e Benfratelli, Palermo (P Bellipanni, C Giannola, C Vicari, V Giambanco); Università degli Studi di Cagliari, Ospedale San Giovanni di Dio, Cagliari (S Ajossa, GB Melis).
• Luxembourg Luxembourg Hospital Centre (P Gratia).
• Spain Hospital Clínico Universitario de Santiago, Santiago de Campostela (R Ucieda Somoza, M Iglesias Díaz); Hospital Central de Asturias, Oviedo (A Escudero Gómis, S Villaverde Fernandez); Hospital de Basurto, Bilbao (T Martinez-Astorquiza, JM Usandizaga Pombo); Hospital Ntra. Sra. de Aránzazu, San Sebastián (JJ Urtiaga Unda, JJ Larraz Soravilla); Hospital Infantil Miguel Servet, Zaragoza (JJ Tobajas Homs, C González Batres); Casa de Maternitat, Barcelona (E Barrau Vernia, V Caradach); Hospital Ntr Sra de Candelaria, Tenerife (JA Cortell Olcina, F Martin Casañas); Hospital Universitario Virgen de la Arrixaca, Murcia (JL Delgado Martín, JJ Parrilla); Hospital Universitario de Canarias, Tenerife (J Parache); Hospital Son Dureta, Palma de Mallorca (A Marqués Bravo, M Usandizaga Calparsoro); Hospital Clínico S Carlos, Madrid (MA Herráiz, M Escudero Fernandez); Hospital Universitario Santa Cristina, Madrid (E Soto Sanchez, F Izquierdo Gonzalez); Hospital Materno-Infantil de Málaga Carlos Haya, Málaga (J Carrera Rodriguez, M Abehsera Bensabat); Hospital San Pedro de Alcántara, Cáceres (JI Moriñigo Yague, C Alcón Alcón); Hospital Clínico San Cecilio de Granada, Granada (A Caño Aguilar, M Dolz Romero); Hospital General Universitario de Alicante, Alicante (JC Martínez Escoriza).
• Sweden Centrallasarettet, Falun (I Westman, A-C Cachrimanidou); Centralsjukhuset, Karlstad (J Hareide, G Wadsten); Sundsvalls Sjukhus, Sundsvall (L Berglund); Centralsjukhuset, Kristianstad (H Ström, G Helm); Universitetssjukhuset MAS, Malmö (S Montan); Mälarsjukhuset, Eskilstuna (B Möller, M Rom); Kärnsjukhuset, Skövde (J Leyon, G Wallstersson); Centrallasarettet, Västerås (LOW Svensson); Länssjukhuset Ryhov, Jönköping (R Boij, R Lenrick); Centrallasarettet, Borås (T Solum); Sahlgrenska Universitetssjukhuset/Östra, Göteborg (U-B Wennerholm, M Wennergren); Akademiska Sjukhuset, Uppsala (U Hansson); Norrlands Universitetssjukhus, Umeå (P-Å Holmgren, I Sjöberg); Regionssjukhuset i Örebro, Örebro (G Falk, M Lood); Regionssjukhuset i Linköping, Linköping (A Jeppsson, G Berg, S Kjellberg); Universitetssjukhuset i Lund, Lund (G Lingman, K Marsal); Karolinska Sjukhuset, Stockholm (O Bakos, V Odlind).
• The Netherlands Maxima Medical Center, Veldhoven (MY Bongers); Isala Clinics, Zwolle (J van Eyck); Academic Medical Centre at the University of Amsterdam, Amsterdam (M Pel, OP Bleker); Free University Medical Centre, Amsterdam (JIP de Vries, HP van Geijn); Academic Hospital Groningen, Groningen (MP Heringa, JP Holm); Leiden Academic Medical Centre, Leiden (J van Roosmalen, HHH Kanhai); Academic Hospital Maastricht, Maastricht (LLH Peeters, JG Nijhuis); Academic Medical Centre St Radboud, Nijmegen (PP Van den Berg, DDM Braat); Academic Medical Centre Utrecht, Utrecht (HW Bruinse, G Visser); Erasmus Medical Centre, Rotterdam (EAP Steegers, JW Wladimiroff).
• The UK Derriford Hospital, Plymouth (I Montague, AD Falconer); Leicester Royal Infirmary, Leicester (A Akkad, C Stewart); King’s College Hospital, London (M Marsh); Southmead Hospital, Bristol (D Bisson); Nottingham City Hospital, Nottingham (H Hamoda, DT Liu); Birmingham Womens Hospital, Birmingham (MJ Whittle); St James University Hospital, Leeds (G Mason); Rosie Hospital, Cambridge (P Plumpton, C Lees); North Staffordshire Maternity Hospital, Stoke on Trent (GV Sunanda, G Masson); Royal Gwent Hospital, Newport (R Gonsalves); University Hospital of Wales, Cardiff (K Sidhu, N Amso).