Births at night examined

Friday July 16 2010

Babies born at night are “three times more at risk of death,” said the Daily Mail. The newspaper suggested that a lack of senior staff available outside of normal working hours is putting newborns at risk.

The well-conducted study behind the story reviewed records of over one million full-term, single baby births in Scotland from 1985 to 2004. It found an increased risk of newborn death out of hours (i.e. during the night time and weekends), with deaths mostly related to lack of oxygen during delivery. However, these results must be interpreted in the correct context, as the risk of newborn death was very low in both groups: 4.2 out of 10,000 births in normal working hours, and 5.6 out of 10,000 births out-of-hours.

As the researchers say, the observations could be due to many different causes, as the reasons behind this relationship were not examined and should not be assumed to be due to ‘hospital staffing shortages’. Many women deliver out-of-hours (three-quarters in this sample) and research will continue to examine the link between time of birth and adverse outcomes for mothers and babies.

Where did the story come from?

The study was carried out by researchers from University of Cambridge and University of Glasgow, and was funded by Medical Research Council and the Royal College of Obstetricians and Gynaecologists. The study was published in the peer-reviewed British Medical Journal.

The papers have reflected the findings of this research but, in general, fail to clarify that the risk of death was small for both day and night births. It is misleading to report that the associations may be ‘due to hospital staffing shortages’, as the causes of different death rates have not been examined in this research and any such claims are based on speculation.

What kind of research was this?

This was a population-based cohort study that examined data from Scottish birth certificates and relevant databases between 1985 and 2004. It aimed to assess whether the time and day of birth had an effect upon the risk of newborn death. This particular study has the advantages of having access to a large quantity of data covering over one million births. However, it does rely on records being filled in accurately and completely.

The study did not assess the reasons behind any observed associations between time of birth and mortality risk, which could be due to a number of factors. As such, it should not be assumed that this is due to lack of skilled staff being available out of hours.

What did the research involve?

The researchers used various data sources. The Scottish Morbidity Record collects information on the outcomes for mothers and babies discharged from Scottish maternity hospitals. This record was used to identify all single-baby births between 1985 and 2004. The babies were linked to the Scottish Stillbirth and Infant Death Survey, which uses codes to record the cause of death for all babies who die around the time of birth.

The researchers were only interested in single babies born at term (between 37 and 42 weeks), with no congenital abnormalities, who were ‘cephalic’ (presenting head first) at full-term, and for whom the delivery method was recorded. They also performed a separate analysis only looking at babies who were delivered at hospital units that delivered more than 10 babies a year.

The main outcome of interest was death in the newborn. This was defined as death of a liveborn baby within the first four weeks of its life. The grouped births based on their day and time of delivery:

  • Weekday births: between 09:00 and 17.00 Monday to Friday
  • Overnight weekday births: between 17:01 and 08:59 on weekday nights (includes Saturday morning until 08:59)
  • Weekend births: from 09:00 on Saturday mornings until 08:59 on Mondays
  • All out-of-hours births: collectively, all births at any time other than 09.00-17.00, Monday to Friday

The association between risk of death and time of birth was adjusted for various possible confounding factors, including characteristics of the birth, maternal characteristics and obstetric history, social and demographic characteristics, and ‘hospital throughput’ (total number of births for a given hospital in a given year).

What were the basic results?

A total of 1,039,560 live births met the specified inclusion criteria, which corresponded to over 95% of all single full-term births in Scotland for 1985-2004. Within the study cohort, 72% of births occurred out-of-hours. In total there were 539 (0.05%) newborn deaths, which was equivalent to a rate of 5.2 out of 10,000 live births. Analysis revealed that just over half of these births (273) were related to anoxia (lack of oxygen) during delivery.

During normal weekday working hours (Monday to Friday, 0900-1700) the risk of newborn death was 4.2 out of 10,000, and for all other times (out-of-hours) it was 5.6 out of 10,000: equivalent to a 30% greater incidence of death (odds ratio 1.3, 95% confidence interval 1.1 to 1.6).

They found that the increased chance of death out-of-hours was mostly related to a higher number of deaths due to anoxia (70% increased risk of death due to anoxia [a total decrease in the level of oxygen] out of hours; odds ratio 1.7, 95% CI 1.2 to 2.3). The attributable fraction of newborn deaths ascribed to anoxia during delivery out-of-hours was 26% (that is 26% of deaths related to anoxia during birth might not occur if women could deliver during normal hours rather than out-of-hours).

The associations seen were not due to confounding by maternal, infant and obstetric characteristics.

How did the researchers interpret the results?

The researchers conclude that delivery of an infant outside of the normal working week was associated with an increased risk of newborn death due to anoxia during delivery.


This is a well-conducted study that analysed Scottish medical records on over one million single, full-term births to determine whether there were any association between time of birth and risk of newborn death. A particular strength of this study is the accuracy of the records used: the Scottish morbidity record reportedly has an almost 99% completion since the late 1970’s and receives regular quality assurance checks. Also, the Stillbirth and Infant Death Survey was completed using the General Register Office, and is reportedly 100% complete.

Though there was an increased risk of newborn death out-of-hours, mostly related to deaths due to lack of oxygen during delivery, these results must be interpreted in the appropriate context:

  • The risk of newborn death, regardless of birth time, is very low. The rate in this large population study was 4.2 out of 10,000 during normal working hours, increasing to 5.6 per 10,000 out-of-hours. Therefore, although this relates to a 30% increased risk, the actual number of deaths for out-of-hours births is still very small.
  • The reasons for the observations, particularly the excess deaths due to anoxia, cannot be easily explained as the situations surrounding the adverse birth outcomes were not examined in detail.
  • As the researchers say, there are many possible reasons for the deaths, which may or may not be due to the variation in staffing availability at different times of the day or fewer clinical facilities available out-of-hours. However, they caution that this cannot be assumed.

Many women deliver out-of-hours (almost three-quarters of this cohort) and this is not something that can easily be controlled. A number of studies have examined the relationship between time of birth and adverse outcomes for the mother or baby, finding similar associations in some cases but no association in others. Research in this area is likely to continue, with the hope of possibly identifying any interventions that could reduce any discrepancy in outcomes between births during the normal working week, and those occurring at night or over the weekend.

Analysis by Bazian
Edited by NHS Choices