Obstetrics Healthcare Atlas
The main findings are summarised below, which also contain a brief description of the patient samples. In addition, information about the population’s use of services in geographical areas (hospital referral areas) are provided. The use is measured as the number of events per 1,000 population or as the proportion of patients that receive a certain service. Brief comments on the variation in use between geographical areas are also include.
About the Obstetrics Healthcare Atlas
The Obstetrics Healthcare Atlas maps geographical variation in the use of health services relating to pregnancy, childbirth and the postnatal period during the period 2015–2017. The Obstetric Healthcare Atlas is the second of two healthcare atlases in the area of gynaecology and obstetrics produced by the Centre for Clinical Documentation and Evaluation (SKDE) at the initiative of the Norwegian Society of Gynecology and Obstetrics. The first one was the Gynaecology Healthcare Atlas, which was published on 8 January 2019.
The Gynaecology Healthcare Atlas was characterised by high, and sometimes very high, geographical variation. Overall, the Obstetric Healthcare Atlas has identified less variation, but we have found several cases of high and unwarranted variation in this field as well. As for the Gynaecology Healthcare Atlas, it is worth noting that an association-based discipline guide with clear guidelines and long-standing traditions does not necessarily result in low variation.
Births in Norway 2015–2017
The number of births per year differs greatly between hospital referral areas. The hospital referral areas in the Oslo region are somewhat different from other areas in that women are older when they give birth for the first time, and these areas also have a high proportion of primiparous women and the lowest proportion of overweight women. The proportion of women giving birth who were themselves born abroad was also high for the hospital referral areas in the Oslo region. Risk factors were relatively evenly distributed across the other hospital referral areas.
About the births
During the period 2015–2017, the Medical Birth Registry of Norway registered an average of 58,198 births per year, of which 57,572 (98.4%) are included in the Obstetric Healthcare Atlas. The highest number of births was found in the hospital referral areas of Bergen and Akershus (approx. 5,400 per year), while the areas of Helgeland and Finnmark had the lowest number of births (approx. 750 per year).
The numbers are highest in the period from April to August, with the fewest births around the New Year. The average for July is 173 births per day, while for December the average is 131. However, there was considerable variation from day to day.
In Norway, 1.6% of births result in two or more babies. This proportion was more or less the same in all the hospital referral areas.
The baby was in the breech position in 4.0% of all births in Norway. The proportion varied from 3.0% in Nord-Trøndelag hospital referral area to 5.2% in the Diakonhjemmet area. The proportion of breech presentations seems to be somewhat higher among women in the Oslo region than among women in other parts of Norway.
For Norway as a whole, 4.5 % of births took place before the 37th week of pregnancy. The proportion varied between hospital referral areas, from 4.5 to 6.7%, and there was some variation from one year to the next within each hospital referral area. This indicates that the proportion of births before the 37th week of pregnancy is more or less the same throughout Norway.
About the mothers
Age when giving birth for the first time
The average age of first-time mothers varied from 26.7 years to 31.9 years between hospital referral areas. Women resident in the Oslo region were older when giving birth for the first time than women in other parts of Norway.
At the start of the pregnancy, 10.3% of primiparous women and 13.1% of multiparous women had a body mass index (BMI) of 30 kg/m2 or more, which WHO defines as obesity. Information about BMI before pregnancy was lacking for a relatively high percentage of the women. Therefore, the proportions of overweight or obese women in the different hospital referral areas are not necessarily comparable. Nonetheless, it may seem that the proportion who are overweight or obese is somewhat lower for women resident in the Oslo region than for women in the rest of Norway.
Mother’s country of birth
For Norway as a whole, 28.9% of primiparous and 29.6% of multiparous women were born abroad. The proportion of primiparous women born abroad varied from about 20% in Nord-Trøndelag hospital referral area to nearly 40% in the Akershus area. More than 40% of multiparous women giving birth in Lovisenberg, Akershus and OUS hospital referral areas were born abroad.
The purpose of antenatal care is to prevent serious complications and illness for both mother and baby. The primary healthcare service has primary responsibility for antenatal care, and the pregnant women choose whether they want to have their antenatal appointments with a GP or a midwife. The specialist health service is responsible for ultrasound scans, follow-up of complicated pregnancies and check-ups after the 40th week of pregnancy.
Healthy pregnant women are offered a basic programme of eight appointments. If illness or risk factors are identified, closer follow-up is provided. The early antenatal appointments map factors such as complicating conditions, hereditary diseases, previous pregnancy and childbirth experience, use of medication, tobacco and alcohol, and psychosocial problems.
The purpose of later antenatal appointments is to monitor the health and development of the mother and baby. This includes blood pressure measurement, urine tests and foetal growth monitoring.
During the period 2015–2017, there were nearly 700 000 antenatal appointments per year with the primary healthcare service (approx. 60%) and the specialist health service (approx. 40%). For Norway as a whole, women had an average of approx. 12 antenatal appointments: around 3 with their RGP/the emergency primary healthcare service, 4 with a municipal midwife and 5 with the specialist health service. Sørlandet hospital referral area stood out with a total of 14 antenatal appointments per pregnancy. Otherwise, there was little geographical variation.
There was considerable geographical variation in the use of the gestational diabetes diagnosis, and Førde hospital referral area stood out with particularly high rates both for primiparous and multiparous women (figures for multiparous women shown in the report). More than four times as many per 1,000 primiparous women were diagnosed with gestational diabetes in Førde hospital referral area as in the UNN area. If we exclude Førde hospital referral area, the variation was more moderate, but still significant.
There was not much variation between hospital referral areas in the number of antenatal appointments. In principle, the specialist health service performs a routine ultrasound scan in week 17–19 and check-ups 7–9 days after the due date. Follow-up of certain pregnant women who need additional consultations come in addition to this. An average of 5 consultations with the specialist health service for all pregnant women indicates that the level of activity in the specialist health service is somewhat higher than expected.
There was striking variation in the use of the gestational diabetes diagnosis, and Førde hospital referral area stood out with a particularly high rate. The composition of the population in the different hospital referral areas cannot explain these differences. Possible explanations include differences in compliance with the national guidelines that applied at the time and variations in reporting to the Medical Birth Registry of Norway.
Induction of labour
In some cases it will be desirable to induce vaginal birth. This normally applies to women who are past their due date or if other medical factors makes it desirable for the child to be born before the spontaneous onset of labour. Starting the birth artificially is called induction of labour.
The most common reasons for inducing labour are pregnancies that continue for 7–9 days past the due date, twin pregnancies, pre-eclampsia, diabetes or the waters breaking without labour starting within 24 hours. Several methods can be used to try to induce labour. The choice of method depends on the maternity unit’s procedures for induction, how ripe the cervix is and the woman’s previous births. If attempts to induce labour are unsuccessful, caesarean section will be considered.
During the period 2015–2017, approx. 5,900 primiparous women and 6,600 multiparous women were induced each year. This corresponds to about 24% of primiparous women and 20% of multiparous women. The figures below show the number of induced births per 1,000 births for primiparous and multiparous women. The geographical variation was low, and the highest rate was about 40% higher than the lowest rate.
There was a clear correlation between the use of induction in primiparous and multiparous women. The hospital referral areas with the highest induction rates for primiparous women are mostly among the areas with the highest rates for multiparous women. For Norway as a whole, there was a definite increase in the use of induction during the period.
The Norwegian Society of Gynecology and Obstetrics’ 2014 obstetrics guide provides clear recommendations about induction of labour. Despite this, findings identify some systematic geographical variation that is more than random variation.
The reason for this could be that there is room for discretionary judgement in individual cases and that culture and practices vary between different maternity units. For Norway as a whole, the number of induced births has increased from 16% in 2009 to 22.5% in 2017. This could be due to changes in practice, but can also be explained by a growing need, for example the increasing incidence of gestational diabetes.
The findings indicate that the guidelines have mostly been observed. The geographical variation was so modest that there is no basis for characterising it as unwarranted.
Epidural pain relief
Many women will describe labour as the most intense pain they have ever experienced. Better pain relief during labour could be one way of reducing fear of childbirth and the number of pregnant women who want a caesarean section in their next pregnancy. Epidural pain relief is the most effective form of treatment for labour pain.
A fine plastic tube called an epidural catheter is inserted between two lumbar vertebrae by means of a needle and remains in place in the spinal canal until the birth is over. The woman will still feel the contractions and can participate actively in the birth, but epidural pain relief takes the edge off the pain. The midwife assesses the need for epidural anaesthesia in consultation with the woman in labour, and the obstetrician on duty is always consulted. The epidural catheter is inserted by an anaesthetist.
In some cases, epidural anaesthesia will be recommended for medical reasons such as breech presentation, multiple births, pre-eclampsia or high BMI. Epidural anaesthesia could cause complications such as prolonged labour and a higher frequency of operative delivery. After delivery, the woman can experience symptoms such as temporary numbness of the legs and headaches.
During the period 2015–2017, about half of all primiparous women who gave birth vaginally (approx. 10,000 per year) and a quarter of all multiparous women (approx. 7,200 per year) received epidural pain relief.
There was moderate geographical variation in the use of epidural anaesthesia among primiparous women who gave birth vaginally. Among primiparous women resident in Diakonhjemmet and Bergen hospital referral areas, nearly twice as many per 1,000 births had an epidural as among primiparous women resident in the areas of UNN and Finnmark.
The geographical variation in the use of epidurals was greater for multiparous women who gave birth vaginally than for primiparous women. Nearly three times as many women per 1,000 births had an epidural among women resident in Diakonhjemmet hospital referral area compared with residents of the UNN area.
There was a clear correlation between the use of epidurals in connection with vaginal delivery in primiparous and multiparous women. The hospital referral areas with the highest rates for primiparous women are mostly among the areas with the highest rates for multiparous women (shown in the report).
There was moderate variation in the use of epidurals for primiparous women and high variation for multiparous women. Variation in the population composition, primarily the prevalence of overweight and proportion of high-risk births, cannot explain the observed variation.
The correlation between the use of epidural anaesthesia among primiparous and multiparous women indicates that the variation could be linked to variations in practice. Differences in the availability of anaesthetists, counselling and patient preferences are among the possible explanations. Overall, the variation is so high and systematic that it is deemed to be unwarranted.
The muscle and tissue around the vagina and anus are elastic and will stretch during childbirth. If the baby is born quickly, maximum stretching of the tissue is not achieved. For vaginal breech delivery or operative delivery, more space is needed. In such cases, an incision, known as an episiotomy, is made and will have to be stitched up afterwards. Other situations could also arise that will necessitate an episiotomy. Episiotomy is said to prevent damage to the anal sphincter, but this has not been scientifically confirmed.
Episiotomy was the rule rather than the exception in the 1960s, particularly for primiparous women. However, the frequency dropped as it has become clear that there is no scientific basis for the benefit of episiotomies. According to the Medical Birth Registry of Norway, approx. 200 episiotomies were performed per 1,000 births in the year 2000, compared with 170 in 2017.
The episiotomy rate increases with the maternity units’ birth volume. The lowest rates are found at maternity units with fewer than 500 births per year, while the highest rates are at units with more than 3,000 births per year. A spontaneous (first-degree or second-degree) tear will often heal faster, be less painful and cause less long-term problems than an episiotomy.
During the period 2015–2017, approximately 7,300 primiparous women and 2,300 multiparous women per year underwent episiotomies in connection with vaginal delivery. This corresponds to 36.5% of primiparous women and 8.3% of multiparous women who gave birth vaginally.
Nearly twice as many episiotomies were performed per 1,000 primiparous women resident in Telemark hospital area compared with the Førde area.
The variation was even higher than this for multiparous women who gave birth vaginally. More than twice as many episiotomies were performed per 1,000 births on women resident in Møre og Romsdal hospital referral area compared with the St. Olavs area.
Each year, approx. 10,000 episiotomies are performed and approx. 850 sphincter ruptures occur in Norway. It is not known how many sphincter ruptures are prevented by episiotomies.
Episiotomies are most common in connection with forceps deliveries and vaginal breech deliveries, and there are approx. 1,000 births in each of these categories per year. The number of episiotomies seems to be somewhat higher than the assumed need.
There was moderate or high variation in the use of episiotomy among primiparous and multiparous women. There was also a clear correlation between the episiotomy rates for primiparous and multiparous women in the different hospital referral areas (results shown in the report). Since the need for episiotomies is assumed to be evenly geographically distributed throughout Norway, these findings indicate significant variations in practice as regards episiotomy. The variation is deemed to be unwarranted.
Operative vaginal delivery
Operative vaginal delivery means that the baby is delivered using forceps or vacuum extraction (suction cup), and such procedures are performed by an obstetrician. Forceps delivery or vacuum extraction is used when it becomes necessary to deliver the baby quickly for some reason or if the birth is taking a long time, the labouring woman is exhausted and the contractions are becoming weaker.
Vacuum extraction is the most commonly used technique. It requires contractions and that the labouring woman is capable of pushing, while no such conditions apply to forceps deliveries. Operative vaginal delivery carries a risk of complications such as sphincter injuries and vaginal tears in the mother and facial injury in the newborn. During the period 2015–2017, the proportion of forceps deliveries that ended in severe (third- and fourth-degree) tears was 6.3% for primiparous women and 5.2% for multiparous women.
The proportion of vacuum-assisted deliveries that ended in severe perineal tears during the same period was 5.0% for primiparous women and 3.8% for multiparous women. There are no indications that operative vaginal delivery have a negative effect on the long-term outcome for the baby.
During the period 2015–2017, approximately 4,500 primiparous women and 1,400 multiparous women per year underwent a forceps delivery or vacuum extraction. About 85% of them underwent vacuum extraction. Forceps deliveries were most frequent in Bergen and Fonna hospital referral areas.
For Norway as a whole, approx. 23% of all vaginal births to primiparous women ended in forceps or vacuum-assisted delivery, and there was considerable variation between hospital referral areas. There were more than twice as many operative vaginal deliveries per 1,000 births among women resident in Stavanger hospital referral area compared with women resident in the Vestfold area. The proportion of vacuum extractions varied from 43.6% in Bergen hospital referral area to 99.7% in the UNN area.
For multiparous women, approx. 5% of all vaginal births ended in forceps or vacuum-assisted delivery, and there was considerable variation between hospital referral areas here too. There were more than twice as many operative vaginal deliveries per 1,000 births among women resident in Stavanger hospital referral area compared with women resident in the Vestfold area. The proportion of vacuum extractions varied from 50.7% in Bergen hospital referral area to 99.0% in the UNN area.
There was a clear correlation between the use of operative vaginal delivery in primiparous and multiparous women. Moreover, hospital referral areas with low operative vaginal delivery rates generally had higher caesarean section rates, and vice versa. This indicates that the variation is largely due to variations in practice.
The observed variation in operative vaginal delivery for primiparous women is deemed to be unwarranted. It is more challenging to assess the variation for multiparous women, since the low number of operative deliveries means that there is a greater element of random variation. Nevertheless, we believe that there is reason to ask whether the observed variation for multiparous women could also be unwarranted.
Emergency caesarean sections are performed when the life of the mother or baby is in danger. Planned caesarean sections are usually performed because risk factors have been identified that could have unfortunate consequences if the baby is delivered vaginally. The probability of caesarean delivery is highest if the baby is in the breech position, in connection with multiple births and if the pregnant woman has previously undergone caesarean section.
The number of births that end in a caesarean delivery has remained very stable in Norway over the past years, and in 2016 it was approx. 16%. However, the proportion increased strongly from 2% in the 1960s to 16–17% in 2006. Caesarean section increases the risk of serious, but rare complications. It also increases the risk of complications in subsequent pregnancies as well as the probability (45%) of a caesarean section in the next pregnancy. Several studies indicate that a caesarean section has negative consequences for newborns in the head presentation, while breech babies may benefit from caesarean delivery.
About one third of all caesarean sections were planned, meaning that the decision was made more than eight hours before the procedure. For primiparous women, one in five caesarean sections were planned, while half of all caesarean sections performed on multiparous women were planned.
Moderate geographical variation in the number of caesarean sections was observed for primiparous women. Approximately 80% more caesarean sections were performed per 1,000 births in Nord-Trøndelag hospital referral area than in the Bergen area. The proportion of emergency caesarean sections varied from 72.8% for women resident in Diakonhjemmet hospital referral area to 85.3% for women in the Møre og Romsdal area.
The geographical variation in the number of caesarean sections was higher for multiparous women. Women resident in Nord-Trøndelag hospital referral area had more than twice as many caesarean sections per 1,000 births as women resident in the Bergen area. If we exclude Nord-Trøndelag hospital referral area, the variation for multiparous women was also moderate. The proportion of emergency caesarean sections varied from 38.1% for women resident in Diakonhjemmet hospital referral area to 57.4 % for women in the Nordland area.
There was a clear correlation between the use of emergency caesarean sections in primiparous and multiparous women. The hospital referral areas with the highest rates for primiparous women were mostly among the areas with the highest rates for multiparous women. There was also a clear correlation for planned caesarean sections.
The rates for emergency caesarean sections were particularly low in Diakonhjemmet, Bergen and Fonna hospital referral areas. If we see this finding in conjunction with the results for operative vaginal delivery (forceps or vacuum-assisted delivery), the overall variation for operative intervention was low.
This indicates that the indications for operative intervention to deliver a baby in an emergency are more or less the same, but that different and sometimes strong preferences as regards the choice of method prevail in different geographical areas.
About 60–80% of all births to primiparous or multiparous women take place without major interventions or complications. However, interventions such as caesarean section, vacuum extraction or forceps delivery are sometimes necessary in order to save lives or avoid short-term or long-term complications. Fortunately, complications such as babies with low Apgar scores and major haemorrhage or severe perineal tears in the mother are rare.
The Apgar score method is an assessment of the condition of the newborn baby just after birth based on heart rate, respiration, muscle tone, colour and response to stimuli. The assessment is carried out at 1, 5 and 10 minutes after birth. Scores between 7 and 10 are considered normal. Scores between 4 and 6 require immediate action, usually with supplemental oxygen and non-invasive ventilation support. A newborn with an Apgar score of 0–3 will need resuscitation.
A major postpartum haemorrhage with a blood loss of more than 1,500 ml represents a risk to the mother’s life and health. Reducing the incidence of severe perineal tears has been a focus area since 2006, and the proportion of women who suffer such tears while giving birth has been reduced from 3.2% in 2006 to 1.7% in 2017.
During the period 2015–2017, there were approximately 780 births per year where the newborn had an Apgar score below 7 at 5 minutes after birth. Of these babies, approx. 450 were born to primiparous women (1.9%) and approx. 330 were born to multiparous women (1.0%). A high proportion of the observed variation was due to chance.
Approximately 3.7% of primiparous women and 2.2% of multiparous women who gave birth vaginally suffered major postpartum haemorrhage where they lost more than 1,500 ml of blood. The low incidence and considerable variation between years in the rates for individual hospital referral areas indicate that much of the observed variation was due to chance.
During the period 2015–2017, just under 600 primiparous women (2.9 %) and approx. 250 multiparous women (0.9 %) per year suffered a third- or fourth-degree perineal tear (shown in the report). There was high variation from year to year and a considerable element of random variation in the occurrence of severe perineal tears.
Complications of childbirth, such as babies with low Apgar scores or major haemorrhage or severe perineal tears in the mother, are relatively rare in Norway.
Due to the low number of events, rates vary from year to year in many hospital referral areas. Despite random fluctuations, some areas had relatively stable rates throughout the period. Maternity units belonging to hospital referral areas with consistently high rates over time should consider introducing measures aimed at reducing the number of birth complications.
The low number of events and considerable element of random variation mean that it is not possible to identify any systematic geographical variation.
The postnatal period
The postnatal period lasts for approximately six weeks from the birth. This is the period it will normally take for the body to adjust, physically and mentally, to the fact that the pregnancy is over and life as the mother of a small child has begun. Postpartum women, their newborns and family may need close follow-up during the initial period after the birth.
Postnatal follow-up is partly related to the postpartum woman’s medical needs and the baby’s vulnerability while it adjusts to life outside the womb and stable feeding and growth is established. It is particularly important to identify and offer assistance to women and families in cases where there is a risk that mental health issues, the family or care situation or other factors could pose a threat to stability and good development. Between one and three days after birth, the newborn undergoes an examination that focuses on the hips, heart, hearing and rare diseases.
In most cases, these examinations take place while the mother and baby are staying at the hospital's maternity unit or maternity hotel. Some hospitals practise what is known as ambulant delivery, which means discharge from hospital a few hours after the birth with subsequent follow-up being provided by midwives, health visitors and postnatal outpatient clinics. The length of postnatal hospitals stays decreased from 4.1 to 2.8 days during the period from 1999 to 2017.
Each year, postpartum women had 11,000 contacts with their RGP/the emergency primary healthcare service and 25,000 outpatient contacts with the specialist health service. Postpartum women resident in Bergen hospital referral area had nearly four times as many contacts per 1,000 women with their RGP/the emergency primary healthcare service and the specialist health service births as postpartum women from the Stavanger area. The variation is still high even if we exclude the two hospital referral areas that had the highest rates.
Beginning in 2017, Norwegian municipalities reported the number of home visits made by midwives, in 2017 within one or two days and from 2018 within three days after discharge from a delivery and maternity unit. There was considerable geographical variation in the number of home visits. Postpartum women resident in Førde hospital referral area received five times as many visits from a municipal midwife per 1,000 women as postpartum women resident in the Stavanger area.
The results for mothers and babies during the postnatal period are characterised by considerable geographical variation in the use of both the specialist health service and the municipal midwife service. This could indicate that the division of responsibility between the health trusts and municipalities is not agreed and clear as required by the Health and Care Services Act.
User surveys show that women report more positive experiences of labour and delivery than of their postnatal stay. At the same time, the peer breastfeeding support group Ammehjelpen reports an increasing demand for their volunteer services.
Taken together, the results may seem to indicate that follow-up of mother and child during the postnatal period is not as highly prioritised nor as well organised as their follow-up during pregnancy and childbirth. The observed geographical variation is deemed to be unwarranted, and we believe that there is reason to question whether postpartum women have good and equitable access to health services regardless of where they live.