This section is based on the 2019 HSPA report and has not been updated since.
Giving every child and every family a good start in life still remains a challenge for our health system. In Belgium, the risk of death in newborns is low, but caution remains necessary as this risk could begin to increase as a result of the greater number of high-risk pregnancies, in particular due to the increasingly late age of future mothers, or the difficult economic situation of part of the population. ‘Avoidable’ risk factors in the mother (for example, obesity or tobacco use) are also associated with more complications for the mother and her child.
The importance of good maternal and neonatal health extends far beyond the period around childbirth, as it has been established that the origin of many pathologies (even in adults) can sometimes be linked to intra-uterine incidents. Protecting mothers and newborns therefore requires a holistic approach to population health, aimed at an overall, trans-generational improvement of public health.
The indicators that we have selected to assess mother and newborn care are focused on:
- Effectiveness of care, based on the neonatal mortality rate (MN-1) and the proportion of newborns who have a good Apgar score (MN-2);
- Appropriateness of care, using the proportion of Caesarean sections (MN-3), induced deliveries (labour inductions) (MN-4), episiotomies (MN-5), vaginal deliveries after a Caesarean section for a previous pregnancy (MN-6), as well as the number of very premature births in maternity wards without a neonatal intensive care unit (NICU) (MN-7), and excessively frequent screening for toxoplasmosis during pregnancy (MN-8);
- Efficiency of care, based on the average length of stay in maternity units (MN-9) and the number of antenatal consultations (MN-10).
Neonatal mortality (MN-1)
The mortality rate during the neonatal period (i.e. until 28 days after birth) is a key measurement in the area of mother and child care, as it is a good indicator of the health system’s effectiveness and quality as well as of population health. In Western countries, the main causes for neonatal death are congenital abnormalities and complications related to very preterm births. The risk of neonatal death is also higher in case of multiple pregnancy.
Another indicator, which is less often used because it is more subject to caution, is the perinatal mortality rate, defined as the number of deaths occurring between the end of the 22nd week of gestation and the end of the first week of life. This indicator has also been decreasing since 2009. For details on this secondary indicator, please refer to the technical datasheet.
Fortunately, neonatal mortality is very low in Belgium. It should be noted that the ‘up-and-down’ appearance of the diagram is due to variations that are particularly amplified by the fact that the number of neonatal deaths is very low, as a single death more or less during one year can cause a marked difference.
RESULTS
- The neonatal mortality rate in Belgium is 2.16 per 1,000 live births (2015); this figure has been decreasing throughout Belgium since 1998 (it was 3.39 per 1,000 live births at that time) (Figure 1).
- Brussels is the region where the reduction is the strongest, after two peaks observed in 2003 and 2008. Currently (2015), Brussels is showing the lowest neonatal mortality rate in the country.
- This neonatal mortality rate places Belgium 7th among the EU-15 countries, just below the European average (2.3 neonatal deaths per 1,000 live births - Figure 2). However, there are some variations between countries with regard to recording, in particular of preterm births; these international comparisons should therefore be considered with some caution.
Data source: Statbel
Data source: OECD health data 2018
Link to technical datasheet and detailed results
Number of newborns with an Apgar score <7 at 5 minutes (MN-2)
The Apgar score is a measurement which is systematically performed on all newborns at the time of birth. By noting certain characteristics of the baby (heart rate, respiratory rate, muscle tone, reflexes, and skin colour) after 1 min, 5 min and 10 min of life, a score from 1 to 10 is determined. A score of 10 indicates a fully healthy baby; a score below 7 is the sign of a serious problem. The lower the Apgar score, the higher the risk of death and/or disability.
This indicator assesses the proportion of newborns whose Apgar score after 5 minutes is lower than 7, which is considered as a reflection of care quality, and more particularly of the effectiveness of care to the mother and newborn. The proportion of newborns with an Apgar score lower than 4 is also shown (as a sub-indicator); this is the sign of emergency distress requiring immediate resuscitation procedures.
RESULTS
- In 2015, approximately 18 newborns per 1,000 live births had an Apgar score lower than 7 at five minutes, and approximately 3 per 1,000 live births had a score lower than 4.
- Apgar scores <7 and <4 have slightly decreased in Belgium since 1998.
- Apgar scores <7 and <4 are fairly similar across the country’s three regions, although the situation has been slightly better in Wallonia in the past few years (15.06 for scores <7) compared to Brussels (18.40) and Flanders (18.70). (Figure 3)
Data source: Statbel
Link to technical datasheet and detailed results
Caesarean section rate (MN-3)
Childbirths by Caesarean section have been increasing throughout the world, approximately for the past thirty years. In 1985, the World Health Organisation (WHO) had recommended not to exceed a Caesarean section rate of 10 to 15%, but in 2010, some European countries were close to 35%. This increase has been attributed to various reasons, such as, among others, the reduction in surgical risk, the fear of professional responsibility issues in the event of a vaginal delivery with complications, or the possibility for physicians and patients to organise their schedules.
In theory, a Caesarean section is performed when a vaginal delivery is foreseen as involving a risk for the mother or the child (planned Caesarean section), or when a problem appears during delivery (emergency Caesarean section). In these conditions, a Caesarean section is clearly beneficial to both the mother and the child. By contrast, when this surgical procedure is decided without a precise medical reason, but rather to meet personal convenience needs, the balance between the benefits and risks is less evident. In addition, the average cost of a Caesarean section is at least twice as high as that of a normal delivery. Today, scientific associations of obstetricians and gynaecologists in countries such as Canada encourage the promotion of vaginal delivery; in France and in Belgium, it is recommended to inform the future mother about the increased risk of complications of a future pregnancy following a Caesarean section.
In 2015, the WHO reformulated is opinion: rather than setting an ‘optimal’ Caesarean section rate, it recommended performing a Caesarean section only when truly necessary.
The Caesarean section rate is therefore an indicator of appropriateness of care to the mother and newborn. An analysis of geographical variations can also provide a powerful screening tool for identifying areas where the use of Caesarean sections is inappropriate.
By sub-categorising this rate according to the Robson classification, it is also possible to analyse the Caesarean section rate depending on the characteristics of the pregnancy, in particular parity (number of previous pregnancies), child presentation (head, breech, transverse, etc.), and history of Caesarean sections in previous pregnancies.
RESULTS
- In 2016, the Caesarean section rate was 216 per 1,000 births (22%) for Belgium, with few variations between regions: 223 per 1,000 in Wallonia, 216 per 1,000 in Flanders and 206 per 1,000 in Brussels.
- Overall, the Caesarean section rate has increased by 11% compared to 2005 (195 per 1,000 births). Nevertheless, reduced rates are now being observed in Brussels (since 2013) and in Wallonia (since 2014). The trend is still increasing in Flanders (Figure 4).
- There is great geographical variability among practices, with the Caesarean section rate varying from 117 (Eeklo) to 303 (Arlon) per 1,000 births in 2016.
- Similarly, high variability can be observed among different hospitals, with the Caesarean section rate ranging between 155 and 352 per 1,000 births in 2016 (Figure 5).
- Women who have a transverse or oblique presentation (Robson cat. 9) most often have a Caesarean section, but because these cases are rare, they do not represent a major part of the total number of Caesarean sections.
- The largest part of Caesarean sections are performed in women who have already had a Caesarean section in a previous pregnancy (Robson cat. 5), but also for first deliveries (Robson cat. 1 and 2). For more details about the analysis according to Robson's classification, please see the detailed datasheet.
- Despite the increase in the Caesarean section rate over the years, Belgium remains below the European average (EU-13) (Figure 6).
Data source: SPF-FOD
Data source: SPF-FOD
Data source: OECD
Link to technical datasheet and detailed results
Induced delivery rate (MN-4)
Public opinion is becoming more and more mobilised against the unnecessary medicalisation of childbirth, and in particular against artificially induced deliveries. The term ‘induced delivery’ (or ‘labour induction’) means triggering labour by using a medicine or by artificially breaking the waters, or inducing contractions after a premature breaking of the waters in a pregnant woman showing no other signs of labour.
KCE recommendations (2010) provide a very precise delineation of labour induction; it should only be performed in case of a major risk to the foetus or the mother, or in the event of spontaneous breaking of the waters, if labour does not begin spontaneously within 24 hours, or after 41 weeks of pregnancy (post-term). Any use outside of these limits may be considered as inappropriate. The induced delivery rate is therefore an indicator of appropriateness of care to the mother and newborn.
RESULTS
- The induction rate in Belgium has changed from 319 inductions per 1,000 births in 1998 to 267 inductions per 1,000 births in 2015, i.e. a reduction by 16.3% (Figure 7).
- However, major differences can be observed between regions. In 2016, Wallonia had the highest induction rate (308 per 1,000 births). In Flanders, the number was 238 per 1,000 births, and in Brussels, 284 (Figure 7).
- Significant variations can also be observed among maternity wards, with rates varying from 103 to 493 per 1,000 live births (2015) (Figure 8).
Data source: Statbel
Data source: Statbel
Link to technical datasheet and detailed results
Proportion of deliveries with an episiotomy (MN-5)
An episiotomy is an incision performed in the vulva during delivery in order to widen the passageway for the baby and to prevent tearing of the perineum. This was common practice – nearly systematic – in the 20th century, but it is considered today as a symbol of the excessive medicalisation of childbirth. Its systematic use was called into question by many women wishing to have a more ‘natural’ childbirth, and serious scientific studies tend to demonstrate that an episiotomy should only be used when it is truly necessary.
The episiotomy rate in Belgium is therefore an indicator of appropriateness of care to the mother and newborn. It is also interesting to analyse the variability of practices among different maternity wards.
RESULTS
- The proportion of women who had an episiotomy during a vaginal delivery is 40.8% (2015), which represents a reduction by 14.8% since 2010. (Figure 9)
- The episiotomy rate remains the highest in Flanders (46.6%), followed by Wallonia (35.9%), and Brussels (30.1%).
- Major variations can be observed among maternity wards, with episiotomy rates varying from 8 to 84% (2015). (Figure 10)
Data source: Statbel
Data source: Statbel
Link to technical datasheet and detailed results
Vaginal deliveries after a Caesarean section (MN-6)
Previously, a motto would say that any women having had a Caesarean section must always deliver by Caesarean section thereafter (‘Once a Caesarean, always a Caesarean’). This is no longer the case today, and it is even considered that a vaginal delivery in a woman having had a Caesarean section in the past causes less complications risks for the mother, both in the immediate future and in subsequent pregnancies. These new recommendations have led to an increase in vaginal deliveries among women who have had a Caesarean section in the past, which has helped diminish the total number of Caesarean sections. However, at the same time, an increase in the risk of uterine tearing (a complication that is potentially very serious for the mother and the newborn) has also been observed in women with a history of Caesarean section and in whom delivery had been induced. For this reason, it is considered today that vaginal delivery is a reasonable option which should be encouraged in women having had a Caesarean section for a previous pregnancy, but that any kind of labour induction should be avoided, particularly with prostaglandins.
This indicator is therefore an indicator of appropriateness of care to the mother and newborn.
RESULTS
- In 2015 the proportion of vaginal deliveries in women having previously had a Caesarean section (VBAC) was 31.88% in Belgium. The highest rate was in Brussels (38.68%), followed by Wallonia (31.24%) and Flanders (30.14%) (Figure 11).
- In Brussels and in Wallonia, this proportion has been increasing again after having fallen in 2012-2014; in Flanders, it continues to decrease slightly (Figure 11).
- Significant differences can be observed among hospitals, with the number of vaginal deliveries after a Caesarean section varying from 12 to 61% (2015 figures) (Figure 12).
Data source: Statbel
Data source: Statbel
Link to technical datasheet and detailed results
Proportion of very preterm births in maternity wards without a neonatal intensive care unit (MN-7)
Between 1 and 1.5% of all births occur very prematurely (between 22 and 31 weeks of pregnancy). These events are therefore rare, but they represent 30 to 50% of all perinatal deaths. When these births take place in maternity wards which have a neonatal intensive care unit (NICU) where the child can be treated immediately, the chances of survival are greater. For this reason, it is recommended that future mothers who have a risk of (very) premature delivery be transferred to this type of maternity wards before the child's birth (intra-uterine transfer). The proportion of very preterm births occurring in maternity wards without a NICU should therefore be as low as possible.
RESULTS
- The percentage of very preterm babies born in maternity wards which do not have a NICU has been decreasing throughout Belgium between 2011 and 2015: from 8.6% to 3.7% in Brussels, from 22% to 19% in Wallonia, and from 21.8% (2010) to 17.9% in Flanders (Figure 13).
- Brussels has the lowest number of very preterm infants born in maternity wards without a NICU (3.7% in 2015). Indeed, in this region, most maternity wards are attached to a NICU on the same site.
delivered in an hospital without a NICU, by region, 2011-2015
Data source: CEpiP (BRU, WAL)- SPE (FLA)
Link to technical datasheet and detailed results
Repeated screening for toxoplasmosis during pregnancy (MN-8)
Ideally, a pregnant woman should be monitored at regular intervals during her pregnancy in order to identify and treat any potential risk to the mother and the future baby. In 2015, KCE published a series of clinical recommendations defining which examinations should be performed at what stage of pregnancy (for pregnancies without any special risk). Some examinations do not need to be performed systematically and/or repetitively in all pregnant women as they bring no benefit, and can even lead to additional examinations which are not harmless and may give rise to unnecessary concerns. This is the case, in particular, of screening tests:
- For toxoplasmosis (taken as a primary indicator in this report): a single test at the beginning of the pregnancy (or before) is sufficient, if it helps encourage future mothers who are not immunised to take preventive measures and reassure those who are already immunised.
- For cytomegalovirus infection (secondary indicator): a single test at the beginning of the pregnancy (or before) may be considered if it helps encourage future mothers who are not immunised to take some precautions. If a screening test is done at the beginning of the pregnancy, the future parents should be clearly informed about the possible consequences.
- For hepatitis C (secondary indicator): this test should be reserved for women who have a special history, such as intravenous drug use or tattoos and piercings performed by non-professionals.
In June 2017, the INAMI nomenclature was updated: the new nomenclature codes for toxoplasmosis screening test cannot be billed more than twice during pregnancy, and cytomegalovirus infection tests are only reimbursed when acute cytomegalovirus infection is suspected. As the data for 2017 are not yet available, it is not possible to assess the consequences of the modification at this time.
These are therefore indicators of appropriateness of care to the mother and newborn.
RESULTS
- Toxoplasmosis: 74.16% of pregnant women were screened at least twice during their pregnancy in 2016, which represents an increase by 4.9 percentage points since 2010 (69.29%).
- The per-region figures are 77.7% for Wallonia, 73.73% for Flanders and 68.55% for Brussels (in 2016) (Figure 14). Major variations can be observed in screening test practices among regions (Figure 15).
- Cytomegalovirus (CMV) infection: 48.62% of pregnant women were screened at least twice for CMV infection in 2016, which represents an increase by 3.6 percentage points since 2010 (45%). The per-region figures are 56.25% for Wallonia, 49.21% for Flanders and 29.52% for Brussels (in 2016).
- Hepatitis C: 70.83% pregnant women were screened for hepatitis C in 2016, which represents an increase by 8.05 percentage points since 2010 (62.78%). The per-region figures are 82.34% for Brussels, 79.82% for Wallonia and 63.11% for Flanders (in 2016).
Data source: IMA-AIM
Data source: IMA-AIM
Link to technical datasheet and detailed results
Hospital length of stay for a normal delivery (MN-9)
The great majority of deliveries take place without any special problems: these are called ‘low-risk’ deliveries, for which the type of care is almost always the same. The length of stay is therefore practically identical for all mothers. In many countries bordering Belgium, the trend is clearly toward shortening the hospital length of stay and adding more care services at home performed by midwives and/or nurses. According to the Organisation for Economic Cooperation and Development (OECD), this tendency is accompanied by a reduction in costs, since hospital care is more expensive than home care.
In Belgium, policies aimed toward shortening maternity stays, with a shift to postnatal care services at home. According to a KCE report, length of stay under 72 hours are perfectly conceivable in Belgium provided that preparations are already made during the pregnancy and that great attention is paid to continuity of care as soon as the mother returns home. Several pilot projects have been launched in order to test the feasibility of various proposals for strengthening home care services.
The average length of stay for a normal delivery is an indicator used by the OECD to compare the efficiency of health systems on an international level.
Sub-indicator: Screening tests done outside the required deadlines in newborns
In all infants born at full term, a few drops of blood are systematically taken from the heel (the ‘heel-stick’ test) in order to screen for a few rare metabolic diseases whose consequences can be limited if treatments are initiated immediately. However, it is very important that these tests be done between the 3rd and 5th day of life, otherwise the test results may be erroneous.
In the current context of reducing the length of stay in maternity wards, these blood samples have to be taken more and more often at home by midwives, which increases the risk of delay. It is therefore important to verify that this change of practice does not give rise to a higher number of tests performed outside the required deadlines.
There are currently no centralised results for Flanders; we are only able to present here the results for the Wallonia-Brussels Federation, centralised by the ONE (Office de la Naissance et de l’Enfance)..
RESULTS
Average length of stay for a normal delivery:
- The hospital length of stay for a normal delivery has changed from 5 days in 2000 to 3.1 days in 2016 (Figure 16).
- Slight differences can be observed between regions: Brussels has the lowest average length of stay (2.8 days) compared to Wallonia (3.1 days) and Flanders (3.2 days).
- After having remained far above the EU-13 average for a long time, Belgium has been coming gradually closer to this average in 2016 (2.8 days). Any comparison between countries should, however, be done with caution, as not all OECD countries use exactly the same calculation methods (Figures 17 & 18).
Data source: RHM-MZG
Data source: OECD, health statistics 2018
Data source: OECD, health statistics 2018
Screening tests done outside the required deadlines in newborns
- In the Wallonia-Brussels Federation (no centralised data available for Flanders), only 1.8% of neonatal screening tests were done after the 5th day, and 0.57% before the 3rd day, which means that most screening tests did actually take place between 3 and 5 days of life (2016 figures).
- The proportion of blood samples taken at the right time has been constantly increasing since 2013. A rise in the number of excessively early or late screening tests was observed in 2014, the year when the length of maternity ward stays was shortened, but these blood sampling delays have been largely cleared since then.
- However, the number of tests arriving at the laboratory 4 or more days after sampling has been consistently growing since 2013, which could be related to the fact that blood samples are more and more often taken outside of maternity clinics. This point deserves attention.
Data source: ONE
Link to technical datasheet and detailed results
Number of antenatal consultations (MN-10)
For pregnancies involving no special risk, the recommended number of antenatal consultations is 10 for women who have never had children before, and 7 for women who have had children before, regardless of which type of healthcare professional is monitoring the pregnancy (gynaecologist, midwife, or general practitioner). In the past decade, midwives are increasingly often called upon by pregnant women for pregnancy monitoring visits. This increase should, in theory, be compensated by a reduction in the number of antenatal consultations with gynaecologists, otherwise one might consider that there is a redundancy (over-consumption of care).
The number of antenatal consultations and the analysis of the types of professionals involved is therefore an indicator of efficiency of care to the mother and newborn.
RESULTS
- 85% of pregnant women have had more than 10 contacts with a gynaecologist, a midwife and/or a general practitioner during their (low-risk) pregnancy in 2016. The median number of prenatal visits is 15, increasing slightly over time in the country’s three regions.
- If one excludes contacts with general practitioners (which may be motivated by other reasons than the pregnancy), this proportion drops to 67%. The median number of prenatal visits then becomes 12.
- The number of antenatal consultations with midwives has increased over time, but this increase has not been compensated by a reduction in the number of antenatal consultations with gynaecologists (on the contrary, this number has increased as well).
- In 2016, 3.7% of (low-risk) pregnant women had less than 7 contacts with a gynaecologist, a midwife and/or a general practitioner, which is insufficient. This proportion is higher in Brussels (6%), but has been decreasing.
- If one excludes visits to general practitioners, the proportion reaches 6.29% at the national level and 7.84% in Brussels.
Data source: IMA-AIM
Data source: IMA-AIM
Data source: IMA-AIM