You are here

Severe postpartum haemorrhage after vaginal delivery: a statistical process control chart to report seven years of continuous quality improvement

European Journal of Obstetrics & Gynecology and Reproductive Biology, pages 169 - 175


Severe postpartum haemorrhage after vaginal delivery: a statistical process control chart to report seven years of continuous quality improvement


To use statistical process control charts to describe trends in the prevalence of severe postpartum haemorrhage after vaginal delivery. This assessment was performed 7 years after we initiated a continuous quality improvement programme that began with regular criteria-based audits

Study Design

Observational descriptive study, in a French maternity unit in the Rhône-Alpes region.

Intervention: Quarterly clinical audit meetings to analyse all cases of severe postpartum haemorrhage after vaginal delivery and provide feedback on quality of care with statistical process control tools. Main outcome measures: The primary outcomes were the prevalence of severe PPH after vaginal delivery and its quarterly monitoring with a control chart. The secondary outcomes included the global quality of care for women with severe postpartum haemorrhage, including the performance rate of each recommended procedure. Differences in these variables between 2005 and 2012 were tested.


From 2005 to 2012, the prevalence of severe postpartum haemorrhage declined significantly, from 1.2% to 0.6% of vaginal deliveries (p < 0.001). Since 2010, the quarterly rate of severe PPH has not exceeded the upper control limits, that is, been out of statistical control. The proportion of cases that were managed consistently with the guidelines increased for all of their main components.


Implementation of continuous quality improvement efforts began seven years ago and used, among other tools, statistical process control charts. During this period, the prevalence of severe postpartum haemorrhage after vaginal delivery has been reduced by 50%.

Keywords: Severe postpartum haemorrhage, Quality improvement programme, Statistical process control chart.


The rate of postpartum haemorrhages (PPH) has increased in recent years in many developed countries, including the United States [1] and [2], Canada [3] , Australia [4] , Norway [5] , and Ireland [6] . In particular, PPH due to uterine atony has contributed to this rise, although the reasons for this remain unclear [3], [4], [7], [8], and [9]. Neither a change in the frequency of risk factors associated with patient characteristics nor trends in risk factors associated with practices explain this increase [9] . In France, severe PPH is the leading cause of maternal mortality and, according to the national committee of experts, around 80% of these deaths may be avoidable [10] .

Different quality improvement programmes (QIPs) aimed at improving the medical care of women giving birth have been shown to be effective in obstetrics. These include audits with feedback to the staff, reminders at the moment of prescriptions, and multifaceted intervention programmes [11] . Similar initiatives in the area of PPH have been described and reported to improve the process of care and PPH rates in several maternity units [12] . Other initiatives have been reported or in a single facility [13] or have improved only the process but not the PPH rates [14] . These programmes have been applied for periods ranging from 12 to 36 months, with follow-up and assessment either immediately afterwards or at most a year later. Nonetheless, it is sometimes difficult to make the effectiveness of these multifaceted intervention programmes last over time [15] . The concepts of sustainability applied by various authors differ in their definition, methods of measurement (qualitative or quantitative), quantification, and expression of results [16] and [17]. To our knowledge, no study has assessed the long-term effects of such programmes in obstetrics and, more precisely, for the management of PPH.

Studying the variations of clinical practices and outcomes over time improves our understanding of the factors involved. This in turn makes it possible to determine the most useful actions to implement.

Statistical process control (SPC) is based on a graphical method that can analyse the behaviour of one or more process indicators as part of a continuous QIP [18] . This method is widely applied in healthcare and has recently been applied in the field of obstetrics [19] and [20] but not to the PPH rate. In 2010, our level 3 maternity ward integrated SPC into the QIP. Since 2005, we began to monitor, analyse, and understand the variations in the rates of severe PPH [21] . The rate of severe PPH fell between 2005 and 2008 [21] .

The objective of this study was to describe the course of the rate of severe PPH after vaginal deliveries, seven years after the QIP began, with graphical statistical process control charts.



The study took place from 2005 through 2012 in the level 3 maternity ward of La Croix Rousse Hospital (Lyon), which currently has approximately 4000 deliveries per year. In 2012, the professional staff included 74 full-time equivalent midwives, 19 obstetricians (6 hospital staff physicians, 3 chief residents (senior specialist registrars), 10 associated physicians), and 8 interns.

The source population included all women who gave birth in the maternity ward between 2005 and 2012. The frequency of each of the following risk factors for PPH was measured annually for all women: past caesarean (or other uterine scar), multiple pregnancy, placenta praevia, mode of delivery (spontaneous vaginal delivery, instrumental vaginal delivery, that is, by vacuum or forceps, or caesarean), macrosomia (>4000 g), and episiotomy.

The study population comprised all women with a vaginal delivery (n = 21,822). In 2005, the hospital implemented systematic prospective identification of all cases of severe PPH after vaginal delivery.


Severe PPH was defined by the presence of at least one of the following criteria: blood loss greater than 1500 mL, or transfusion with concentrated red cells, or treatment of haemorrhage by radiologic embolization, or conservative surgical treatment, or hysterectomy, or transfer to the critical care department, or intrapartum haemoglobin loss of 4 g/dL or more, or death due to PPH. The rate of severe PPH after vaginal delivery is the ratio of the number of patients with a severe PPH to the number with a vaginal delivery during the same period.

Blood collector bags were systematically used to measure postpartum bleeding

The continuous QIP began in 2005 and took place in three consecutive stages. It began by combining a clinical audit with quarterly audits of morbidity and mortality from severe PPH [21] . The second stage of the continuous QIP started in 2008, when we began applying special data collection procedures to these cases: summary forms are completed by the obstetric team during daily staff meetings (see Appendix 1). During the quarterly audits, the team defines the quality of the care provided to each as optimal, suboptimal, or not optimal [21] , in accordance with the French clinical practice guidelines for PPH [22] . It was considered optimal if four key steps were taken in the following time periods after diagnosis of a haemorrhage: call to the senior physician <10 min, performance of a manual uterine examination or manual removal of the placenta <15 min, administration of oxytocin as a first-line treatment and of sulprostone in the 30 min after the PPH diagnosis if atony persisted. It was considered not optimal if at least one of these stages was not performed. It was considered suboptimal if at least one of these stages was not performed within recommended time period or if a minor element was omitted, such as verification of lower genital tract or rapid suturing of soft-tissue wounds.

The third stage of the continuous QIP began in 2010, when we added to it a quarterly monitoring of the severe PPH rate, inspired by the specific methods developed by the French Health Authority (Haute Autorité de Santé, HAS) for application of SPC in French healthcare facilities [18] .

Graphical SPC chart

The SPC tool applied here to monitor the behaviour of the indicator is composed of a graphical control chart together with a tracking log. This chart has a central line (CL) and two limits, a lower control limit (LCL) and an upper control limit (UCL), plotted at a given number of standard deviations from the central line. These 3 lines are calculated according to the validated mathematical formula currently applied by the HAS [18] . The indicator values are reported chronologically in this chart. The continuous QIP endeavours to reduce the variability of the process results. A system or process is said to be in statistical control when the value of the indicator changes only within the control limits. Crossing a limit expresses the presence of a cause that must be identified, corrected and used for future improvement.

For this study, based on prospectively collected data about the severe PPHs occurring after vaginal delivery, a control chart was established retrospectively for the 2005–2009 period and prospectively beginning in 2010 [18] . A p-chart (for proportion/percentage) was used to monitor this qualitative binary variable (presence or absence of severe PPH) with variously sized populations of women with vaginal deliveries, depending on the quarter. In view of the frequency of the cases, this monitoring took place quarterly. According to a consensus of local experts, obstetricians, anaesthetists and midwives, and in the absence of available data in the literature, the central line was set at 0.5% based on the mean rate observed from 2008 through 2010. A single standard deviation around the central line was tolerated for the UCL and LCL, in view of the severity of the cases. These limits were calculated for each quarter (as detailed in Appendix 2). The process of care was considered in statistical control if the severe PPH rate remained between the UCL and the LCL.

Statistical analyses

Between 2005 and 2012, the annual frequency of the characteristics of all women in the source population and of those with severe PPH, as well as the compliance of care practices, were compared by a chi-2 test for trend. The threshold of significance was set at 5%. The statistical analyses were performed with EPIINFO.

A segmented linear regression of the number of severe PPHs according to time was performed to demonstrate a change in the general trend between the period from 2005 (when we began this QIP) to the end of 2007 and the period from 2008 (when we began applying special data collection procedures) through 2012. The model’s assumptions were verified.

Ethics approval

No specific ethics approval for this study was required because outcome data were routinely collected at maternity units and analysed in an aggregate format.


Among the 21,822 women who gave vaginal birth in our facility from 2005 to 2012, 140 cases of severe PPH were identified.

Population characteristics

In the source population, patient characteristics and mode of delivery were stable between 2005 and 2012, except for the percentage of women with previous caesareans, for which the trend increased significantly. That trend was no longer significant if the 2005 value was excluded from the analysis (p = 0.9) ( Table 1 ).

Table 1 Characteristics of women with a vaginal delivery (n = 21,822).

  2005 2006 2007 2008 2009 2010 2011 2012  
  n % n % n % n % n % n % n % n % p for trend
Number of women 2919 100 3113 100 3058 100 3213 100 3539 100 3966 100 4019 100 4085 100  
Previous caesarean 219 7.5 342 11.0 354 11.5 353 11.0 366 10.34 427 10.8 481 11.97 432 10.6 0.001
Multiple pregnancy 97 3.3 93 3.0 112 3.7 99 3.1 108 3.4 125 3.2 127 3.2 129 3.2 0.17
Placenta praevia 31 1.1 38 1.2 33 1.1 35 1.1 25 0.71 28 0.7 24 0.6 45 1.1 0.06
Fetal macrosomia 210 7.2 221 6.9 215 6.8 218 6.6 235 6.43 270 6.6 261 6.5 257 6.3 0.07
Caesarean delivery 636 21.8 669 21.5 676 22.1 706 22.0 757 21.4 870 21.94 906 22.5 870 21.34 0.7
Number of vaginal delivery 2283 100 2444 100 2382 100 2507 100 2782 100 3096 100 3113 100 3215 100  
Episiotomy 552 24.2 572 23.4 665 27.9 653 26.1 715 25.7 794 25.7 694 22.3 766 23.8 0.14
Instrumental vaginal delivery 216 9.5 234 9.6 299 12.6 343 13.7 411 14.8 406 13.1 427 13.7 412 12.8 <0.001
Severe PPH at vaginal delivery 27 1.2 25 1.0 16 0.7 9 0.4 16 0.6 13 0.4 16 0.5 18 0.6 <0.001

Among the patients with vaginal deliveries, the frequency of operative vaginal (instrumental) deliveries tended to increase significantly ( Table 1 ). The frequency of episiotomies has been stable since 2005.

Severe postpartum haemorrhage

The incidence of severe PPH after vaginal deliveries was cut in half between 2005 and 2012 (1.20% versus 0.56%, p for trend <0.001)( Table 2 ). There was a statistically significant reduction in the mean number of severe PPHs before the first quarter of 2008 (beta coefficient = −0.0010, p = 0.003); the slope of the line did not differ from zero after this date (beta coefficient = 0.0002, p = 0.19). The difference between the slope before and after the first quarter of 2008 was statistically significant (beta coefficient = 0.0012, p = 0.007).

Table 2 Severe PPH after vaginal delivery (n = 140): quality of care.

  2005   2006   2007 2008 2009 2010 2011 2012  
  n % n % n % n % n % n % n % n % p for trend
Number of severe PPH 27 100 25 100 16 100 9 100 16 100 13 100 16 100 18 100  
Care provided                                  
Optimal 7 25.9 4 16.0 7 43.8 6 66.7 8 50 9 69.2 11 68.8 12 66.7 <0.001
Suboptimal care 10 37.0 17 68.0 8 50.0 2 22.2 4 25 3 23.1 4 25.0 3 16.7  
Non-optimal care 10 37.0 4 16.0 1 6.3 1 11.1 4 25 1 7.7 1 6.25 3 16.7  
Specific components of management                                  
Prophylactic administration of oxytocin 5 18.5 18 72 10 63 9 100 14 87.5 13 100 16 100 16 88.8 <0.001
Examination of the uterine cavity: 19 70.4 23 92 16 100 8 89 15 93.8 12 92.3 16 100 18 100 0.03
Within 15 min of PPH diagnosis 7 25.9 22 88 10 63 8 89 15 93.8 12 92.3 16 100 17 94.4 <0.001
Instrumental examination of vagina/cervix 11 40.7 18 72 9 56 8 89 15 93.8 13 100 15 95.8 18 100 <0.001
Severe PPH due to uterine atony 24 88.9 19 76.0 8 50 7 77.8 10 62.5 8 61.54 9 56.3 10 55.6  

24 100 19 100 8 100 7 100 10 100 8 100 9 100 10 100  
Intravenous administration of sulprostone 11 45.8 15 78.9 7 87.5 6 85.7 8 80.0 8 100.0 8 88.9 8 80.0 0.1
Within 30 min of PPH diagnosis 0   8 42.1 2 25.0 6 85.7 5 50.0 5 62.5 8 88.9 8 80.0 <0.001

The process was totally in control by the first quarter of 2010; and no quarterly severe PPH rate has exceeded the upper control limit since then ( Fig. 1 ).


Fig. 1 Quarterly monitoring with a control chart about proportion of severe PPH after vaginal delivery.

Since 2005, the overall quality of care has improved significantly with a 50% reduction in the percentage of cases involving non-optimal or suboptimal care. All the key stages of management after vaginal delivery have progressively tended to be performed more consistently. Since 2008, the level of the performance of the key stages has stabilised and has ranged from 89% for prophylactic oxytocin administration to 100% for the performance of the manual uterine examination. In cases of persistent uterine atony, sulprostone was administered more rapidly and more often within the 30 min following the PPH diagnosis (80.0% in 2012 versus 0% in 2005, p < 0.001).


From the beginning of the QIP in 2005 through 2012, the incidence of severe PPH after vaginal deliveries has fallen significantly. Since 2010, no quarterly severe PPH rate has been out of control. The quality of management has also improved significantly for each key stage of the process of care.

To our knowledge, this is the first single-centre study to report a longitudinal follow-up of the rate of severe PPH after vaginal delivery over 7 years of a continuous QIP for the process of care. The cases were clearly defined and easily identified clinically. During daily staff meetings, all files from the previous day were reviewed, and a summary form was completed for all women with severe PPH. The continuing fall of the severe PPH rate is mainly due to the regular quarterly audits, combined with the continuous mobilization of all of the professionals involved. Over so long a period, the Hawthorne effect alone cannot explain the improvement. Besides, the modification of practices under the effect of participants’ observations is part of the QIP and is transformed into a management tool, with mutual recognition by all participants of the efforts of the others [23] .

Nonetheless, we do not know the distribution of the PPH risk factors in the population of patients with vaginal deliveries. The source population characteristics have been stable since 2005 except for the rate of women with previous caesareans: their proportion increased, especially from 2005 to 2006 and it stabilised thereafter. Nonetheless, this probably had little impact on the population studied, since 63.5% of women with previous caesareans had caesarean deliveries [24] . Moreover, if this factor had had an impact, it would have been in the direction of an increase in the number of cases of severe PPH due to the increase in yet another risk factor—operative vaginal deliveries.

Choice of indicator

The rate of severe PPH appears to reflect the quality of care of obstetric patients, because prevention makes it possible to reduce the PPH rate [24] , and the frequency of severe cases depends on the time that it takes the team to begin to provide care for the haemorrhage [25] . For a maternity unit, this item could be one minimum component of the performance indicators intended to monitor the quality of care provided to women according to their mode of delivery. This indicator has also been proposed by the Australia and New Zealand College of Obstetricians [26] , to be separated by mode of delivery and with a narrower definition including only patients with transfusions. More recently it was also selected as a patient safety indicator in the United States [27] . The qualitative analysis of these sentinel events has also the advantage of offering teams a space for multidisciplinary dialogue and communication, essential to team cohesion [28] .

Graphical control chart

SPC, with its graphical methods allowing easy visualization, functions as a warning tool that can document periods when incidence exceeds a threshold value (UCL). It thus makes it possible to guide the topics covered at audit meetings toward understanding the reasons for these variations and taking corrective action [18] . It is simple to set up in an Excel file, and on-going data entry is also easy. This graphical control chart also represents an interesting communication tool especially for the maternity units with large numbers of professionals who cannot all attend weekly staff meetings or quarterly audits. It can be displayed in the delivery room and used as communication support when the audit minutes are sent to all team members.

At the introduction of the graphical control chart in 2010, the process of care had already been controlled in part since the last trimester of 2007 and has been under complete control since 2010. The quarterly monitoring of the severe PPH rate enables faster response to problems than annual monitoring. The frequency of this monitoring must be adapted to the frequency of cases and the facility’s volume of activity.

Local experts set the central line, based on the observed mean severe PPH rate in 2008–2010 in order to take the local context (level 3 maternity ward) into account. No data for selecting this value exist as far as we know, but healthcare teams can and should determine the relevant data and analyse the appropriate values. Consideration of the type of activity in setting target and alert values can be debated. This value must be reassessed over time. Too low a value might discourage teams, while too high a value could reduce their vigilance.

In view of the severity of the cases, the permissible range around the central value was set at only one standard deviation. This internal decision rule means that we are subject to a "false alarm" rate on the order of 50%, versus 27.7% with 2 standard deviations and 6.5% with 3 standard deviations [29] . But the analysis of the quality of care of these cases, even if excessive, provides reassurance when management is optimal, especially if the case is determined to be unavoidable.

Continuous quality improvement programme

Our study shows that our QIP has led to an overall improvement in the quality of care of women with severe PPH after vaginal delivery, mainly for the administration of prophylactic oxytocin and of sulprostone. Those elements have contributed in part to reducing the rate of severe PPH. The fourth step of our QIP is simulation training for the care of severe PPH after vaginal deliveries. It began in 2013 to ensure the sustainability of this programme. It has been shown to improve team communication [30] . The SPC can be used to measure the impact of this simulation training program. Its success lies in the combination, step by step, of different tools for measurement (data collection form in 2008), analysis, communication (graphic control chart in 2010), and team training (simulation). The latter element is primordial in a team with numerous participants and with interns who change every 6 months and play a role in the quality of care [31] .

This method could be extended to cases following caesareans, but clear and objective rules for assessing the quality of care and its compliance with guidelines appear more difficult to establish, as clinical practice guidelines are less specific and codified than for cases following vaginal deliveries [22] .

Disclosure of interests

The authors have nothing to disclose.

Contribution to authorship

CD, PO, CH, RCR and CDT participated in the design and the implementation of the study, the collection and the analysis of the data and the drafting and revision of the paper.

STo, PO, MHBC and AD participated in the revision of the paper.

RCR initiated the collaborative project, participated in the design and the implementation of the study, the management of the audit meetings, the analysis of the data and the drafting and revision of the paper.


The authors thank Mrs Anne Lafon for her help in data monitoring.

Appendix 1


Appendix 2

Calculation of the lower control limit (LCL) and the upper control limit (UCL) for building a p-control chart are given. The exact control limits can be calculated using the cumulative binomial distribution expressed statistically by:


stripin: si1.gif

Where stripin: si2.gif is the number of severe PPH in the sample stripin: si3.gif, stripin: si4.gifis the sample size and stripin: si5.gifis the proportion of severe PPH used for plotting the central line.

For a 1 SD control limit strategy, the exact LCL is the value of stripin: si6.gifthat satisfies:


stripin: si7.gif

For a 1 SD control limit strategy, the exact UCL is the value of x_ithat satisfies:


stripin: si8.gif

In practice, this can be done using the CRITBINOMa function of MS Excel:


stripin: si9.gif


stripin: si10.gif


  • [1] W.M. Callaghan, E.V. Kuklina, C.J. Berg. Trends in postpartum hemorrhage: United States,1994–2006. Am J Obstet Gynecol .. 2010;202(353):e351-e356
  • [2] M.S. Kramer, C. Berg, H. Abenhaim, et al. Incidence, risk factors and temporal trends in severe postpartum hemorrhage. Am J Obstet Gynecol. 2013;209(449):e1-e7
  • [3] K.S. Joseph, J. Rouleau, M.S. Kramer, D.C. Young, R.M. Liston, T.F. Baskett. Investigation of an increase in postpartum haemorrhage in Canada. BJOG. 2007;114:751-759 Crossref
  • [4] J.B. Ford, C.L. Roberts, J.M. Simpson, J. Vaughan, C.A. Cameron. Increased postpartum hemorrhage rates in Australia. Int J Gynecol Obstet. 2007;98:237-243 Crossref
  • [5] J. Rossen, I. Økland, O. Bjarte Nilsen, T.M. Eggebø. Is there an increase of postpartum hemorrhage, and is severe hemorrhage associated with more frequent use of obstetric interventions?. Obstet Gynecol Surv. 2011;66:18-20 Crossref
  • [6] J.E. Lutomski, B.M. Byrne, D. Devane, R.A. Greene. Increasing trends in atonic postpartum haemorrhage in Ireland: an 11-year population-based cohort study. BJOG. 2012;119:306-314 Crossref
  • [7] M.S. Kramer, M. Dahhou, D. Vallerand, R. Liston, K.S. Joseph. Risk factors for postpartum hemorrhage: can we explain the recent temporal increase?. J Obstet Gynaecol Can. 2011;33:810-819
  • [8] A. Mehrabadi, J.A. Hutcheon, L. Lee, R.M. Liston, K. Joseph. Trends in postpartum hemorrhage from 2000 to 2009: a population-based study. BMC Pregnancy Childbirth. 2012;12:108 Crossref
  • [9] A. Mehrabadi, J.A. Hutcheon, L. Lee, M.S. Kramer, R.M. Liston, K.S. Joseph. Epidemiological investigation of a temporal increase in atonic postpartum haemorrhage: a population-based retrospective cohort study. BJOG. 2013;120:853-862 Crossref
  • [10] M. Saucedo, C. Deneux-Tharaux, M.H. Bouvier-Colle. French National Experts Committee on maternal mortality. Ten years of confidential inquiries into maternal deaths in France, 1998–2007. Obstet Gynecol. 2013;122(4):752-760 Crossref
  • [11] N. Chaillet, E. Dube, M. Dugas, et al. Evidence-based strategies for implementing guidelines in obstetrics: a systematic review. Obstet Gynecol. 2006;108:1234-1245 Crossref
  • [12] F. Althabe, P. Buekens, E. Bergel, et al. Guidelines trial group. A behavioral intervention to improve obstetrical care. N Engl J Med. 2008;358:1929-1940 Crossref
  • [13] F. Rizvi, R. Mackey, T. Barrett, P. McKenna, M. Geary. Successful reduction of massive postpartum haemorrhage by use of guidelines and staff education. BJOG. 2004;111:495-498 Crossref
  • [14] C. Deneux-Tharaux, C. Dupont, C. Colin, et al. Educational multifaceted intervention to decrease the rate of severe postpartum haemorrhage – the Pithagore6 cluster-randomized controlled trial. BJOG. 2010;117:1278-1287 Crossref
  • [15] P. Pluye, L. Potvin, J.L. Denis. Making public health programs last: conceptualizing sustainability. Eval Program Plann. 2004;27:121-133 Crossref
  • [16] S. Wiltsey, J. Kimberly, N. Cook, A. Calloway, F. Castro, M. Charns. The sustainability of new programs and innovations: a review of the empirical literature and recommendations for future research. Implement Sci. 2012;7:17
  • [17] S.M. Ament, F. Gillissen, J.M. Maessen, C.D. Dirksen, T. Van der Weijden, M.F. Von Meyenfeldt. Sustainability of healthcare innovations (SUSHI): long term effects of two implemented surgical care programmes (protocol). BMC Health Serv Res. 2012;12:423 Crossref
  • [18] French Health Authority. Statistical process control. 2004. 92 pages available .
  • [19] R. Boulkedid, O. Sibony, C. Bossu-Salvador, J.F. Oury, C. Alberti. Monitoring healthcare quality in an obstetrics and gynaecology department using a CUSUM chart. BJOG. 2010;117:1225-1235 Crossref
  • [20] R. Boulkedid, C. Alberti, O. Sibony. Quality indicator development and implementation in maternity units. Best Pract Res Clin Obstet Gynaecol. 2013;27:609-619 Crossref
  • [21] C. Dupont, C. Deneux-Tharaux, S. Touzet, et al. Clinical audit: a useful tool to reduce severe postpartum haemorrhage?. Int J Qual Health Care. 2011;23:583-589 Crossref
  • [22] Postpartum haemorrhage: recommendations for clinical practice by the French College of Gynecologists and Obstetricians Paris December 2004. (College National des Gynécologues Obstétriciens Français– ).
  • [23] T.R. Lied, V.A. Kazandjian. A Hawthorne strategy: implications for performance measurement and improvement. Clin Perform Qual Health Care. 1998;6:201-204
  • [24] B. Blondel, N. Lelong, M. Kermarrec, F. Goffinet. National coordination group of the National Perinatal Surveys. Trends in perinatal health in France from 1995 to 2010. Results from the French National Perinatal Surveys. J Gynecol Obstet Biol Reprod. 2012;41:e1-e15 Crossref
  • [25] M. Driessen, M.H. Bouvier-Colle, C. Dupont, B. Khoshnood, R.C. Rudigoz, C. Deneux-Tharaux. Postpartum hemorrhage resulting from uterine atony after vaginal delivery: factors associated with severity. Obstet Gynecol. 2011;117:21-31 Crossref
  • [26] Retrospective data in full Australasian Clinical Indicator Report 2004–2011. Obstetrics version 7. The Australian Council on healthcare standards p 30–32. Available .
  • [27] K.D. Gregory, L.M. Korst, M.C. Lu, M. Fridman. AHRQ patient safety indicators: time to include hemorrhage and infection during childbirth. Jt Comm J Qual Patient Saf. 2013;39:114-122
  • [28] D. Siassakos, T. Draycott, J. Crofts, L. Hunt, C. Winter, R. Fox. More to teamwork than knowledge, skill and attitude. BJOG. 2010;117:1262-1269 Crossref
  • [29] J.C. Benneyan, R.C. Lloyd, P.E. Plsek. Statistical process control as a tool for research and healthcare improvement. Qual Saf Health Care. 2003;12:458-464 Crossref
  • [30] T.J. Draycott, T. Sibanda, L. Owen, et al. Does training in obstetric emergencies improve neonatal outcome?. BJOG. 2006;113:177-182 Crossref
  • [31] M. Dixon-Woods, R. Baker, K. Charles, et al. Culture and behaviour in the English National Health Service: overview of lessons from a large multimethod study. BMJ Qual Saf. 2014;23:106-115 Crossref


a Réseau périnatal Aurore – Hôpital de la Croix Rousse, Université Lyon 1, EAM 4128, laboratoire Santé, Individu, Société , faculté de médecine Laënnec, 69372 Lyon, France

b Hospices Civils de Lyon, Pôle Information médicale Evaluation et recherche; Université Lyon 1, Lyon, France

c Inserm U953 Recherche épidémiologique en Santé périnatale Santé des femmes et des enfants, Paris

d Réseau périnatal Aurore – Hôpital de la Croix Rousse, Université Lyon 1, INSERM U846, Stem Cell and Brain Research Institute, Lyon, France

lowast Corresponding author at: Hospices civils de Lyon, Laboratoire Reseau perinatal Aurore Hôpital de la Croix Rousse 69004 Lyon, France. Tel.: +33 68234876; fax: +33 472004163.