Study: Centralising hospitals has reduced birth mortality

Every year, around 55,000 children are born in Norway. 

But a long, thin, mountainous country with a coast incised by fjords means it can be tricky for expecting mothers to get to the hospital quickly.

Some women only have to travel a short distance to the nearest maternity ward, while others have to cross fjords and mountains. This can create great anxiety and uncertainty, especially as labour approaches.

Researchers at NTNU have now investigated the effect of both the size of the maternity ward and travel time on a baby’s risk of dying in connection with birth. 

The study used data from more than one million births in Norway over a period of 17 years.

There were several clear findings:

• Longer travel times significantly increase the risk of giving birth on the way to hospital, but there is little to indicate that longer travel times increase the risk of death.
• There is a clear correlation between the size of the maternity ward and mortality. The more births a maternity ward has, the rarer the deaths.
• Smaller hospitals have more babies delivered by caesarean section
• There is a 55 per cent higher risk of the child dying shortly before or shortly after birth if the mother is expected to give birth in a hospital with 500 births a year compared to a hospital with 2,000 births a year. The effect levelled off at 2,000 births per year.
• There is approximately a doubling of the risk of giving birth on the way to the hospital for every 30-minute increase in travel time.

Fewer than one birth a week

“There may be other good arguments for maintaining a highly decentralised maternity service, but lower mortality associated with shorter travel times does not seem to be one of them. Our findings support Norway’s decision in recent decades to centralise hospitals to reduce childbirth-related mortality,” says Fredrik Carlsen. 

He is a professor of economics at NTNU, and one of the researchers behind the new study.

Fewer children are born in Norway than before, and fewer people live in rural areas than before. The population base for maintaining birth centres in this elongated country is therefore gradually disappearing on its own.

From 1967 to 2016, the number of birthing places in Norway was reduced from 182 to 48.

Norway now has 43 maternity services.

While birthing centres are required to maintain their professional competence, there is no lower limit to the number of births needed to retain a maternity service.

The smallest maternity institutions in Norway have fewer than one birth per week, according to Norway’s Medical Birth Registry.

“The smallest birthing places are basically for births with little risk. At the same time, our study shows that there's a greater propensity to use caesarean sections in small hospitals. This may suggest that obstetricians in larger hospitals have more training in using less invasive techniques such as forceps and vacuums,” says Andreas Asheim.

He is a researcher at NTNU’s Department of Clinical and Molecular Medicine.

Mostly a political decision

When labour starts, women generally know where they should go to give birth. 

According to national guidelines, pregnant women are offered follow-up by healthcare workers so that potentially risky births can be identified in advance.

“Births that are believed to pose an extra risk to the mother or child are referred to one of the large hospitals. Women will give birth in a hospital where the professional services are best adapted to the needs of both the woman and the child. A prerequisite for ensuring safe follow-up of women in labour is a well-functioning system for identifying high-risk births. However, it’s not possible to identify all high-risk births in advance,” says Johan Håkon Bjørngaard.

He is a professor at NTNU’s Department of Public Health and Nursing.

Women who have apparently uncomplicated pregnancies and who live close to small hospitals may therefore face a weaker maternity service if they nevertheless have complications during childbirth, according to the research team.

“It's important that we have the knowledge to have the discussion about how best to organise maternity care. There's a trade-off between the risk of long travel times and the need for an adequately sized maternity ward. This is a complex, and ultimately a political decision. As researchers, we can only make sure that the consequences of the choices made are as clear as possible,” Bjørngaard says.

Designed the study in a unique way

Quality in maternity care is difficult to study because women who have high-risk births are referred to large hospitals. 

The statistics will therefore show a higher incidence of birth complications at larger hospitals, even thought the size of the hospital isn’t really a factor in determining where it is safest to give birth.

“For the same reason, we might expect to find a higher risk with long travel distances, as the most high-risk births will be sent to large hospitals and thus have a longer travel distance,” says Bjørngaard.

To avoid this, the researchers accounted for both the division of tasks and other differences between those who give birth at large and small hospitals.

The researchers have therefore compared women based on which hospital they were supposed to give birth at in the first place – and not where they actually ended up.

Safe to give birth in Norway

The study included all births in Norway (approximately 1.1 million) from 1999 to 2016.

The researchers used data from two subgroups: 203,464 births from women who had moved between pregnancies, and 460,776 births from women who lived in neighbouring municipalities.

“It's important to emphasise that it's safe to give birth in Norway. There are few deaths, but we are facing major changes over where people choose to live. If we maintain many small maternity wards, it will be very important to ensure prenatal care so that everyone who needs the expertise that the large hospitals possess can get it when they need it,” Bjørngaard says.

Reference:

Asheim et al. The Effects of Hospital Delivery Volume and Travel Time on Perinatal Mortality and Delivery in Transit: Causal Inference with Triangulation, Epidemiology, 2025. DOI: 10.1097/EDE.0000000000001840

———

Read the Norwegian version of this article on forskning.no

This content is paid for and presented by NTNU

This content is created by NTNU's communication staff, who use this platform to communicate science and share results from research with the public. NTNU is one of more than 80 owners of ScienceNorway.no. Read more here.

More content from NTNU:

• 

  Early testing of schoolchildren: “We found absolutely no effect”

• 

  This determines whether your income level rises or falls

• 

  Why is nothing being done about the destruction of nature?“We hand over the data, but then it stops there"

• 

  Researchers now know more about why quick clay is so unstable

• 

  Many mothers do not show up for postnatal check-ups

• 

  This woman's grave from the Viking Age excites archaeologists