Early neonatal mortality, asphyxia related deaths, and timing of low risk births in Hesse, Germany, 1990-8: observational study

BMJ 2000; 321 doi: (Published 29 July 2000) Cite this as: BMJ 2000;321:274
  1. Günther Heller (hellerg{at}, assistant professora,
  2. Björn Misselwitz, public health researcherc,
  3. Stephan Schmidt, professor of gynaecology and obstetricsb
  1. a Institute of Medical Sociology and Social Medicine, Medical Centre of Methodology and Health Research, Philipps-University of Marburg, Medical School, D 35033 Marburg, Germany
  2. b Department of Obstetrics, Centre of Gynaecology and Obstetrics, Philipps-University of Marburg
  3. c Institute of Quality Assurance Hesse, D 65760 Eschborn, Germany
  1. Correspondence to: G Heller
  • Accepted 17 April 2000

A higher neonatal mortality related to intrapartum events during the night has been reported in Great Britain.1 2 We investigated whether the time of birth affects early neonatal mortality or deaths related to asphyxia in low risk births.

Participants, methods, and results

Data from the perinatal birth register of the federal state of Hesse, Germany, 1990-8, were used ( The register comprises detailed information about all infants born in birth clinics (more than 95% of all births in Hesse); about the mother, including the pregnancy; and about the delivery, as documented by the obstetrician in charge of the birth, using an evaluated standardised questionnaire comprising 67 items.3 Detailed information is available about the child's morbidity and reasons for death coded in 40 predefined categories adapted from ICD-9 (international classification of diseases, 9th revision).

Outcome events were deaths during labour or within the first seven days of life (early neonatal deaths) and asphyxia related deaths during the same period. Completeness of the record of all early neonatal deaths was validated by comparison with corresponding death rates as reported by the statistical office of Hesse.4

To control for the effects of planned births with respect to the time of birth we excluded antepartum deaths, caesarean sections, infants born before the 37th week of gestation, and infants with congenital malformations or hereditary metabolic diseases. Those births occurring between 9 pm and 6 59 am were defined as night time births. Otherwise births were assumed to have taken place during the day.

Relative risks and 95% confidence intervals were calculated to assess the effect of night and day on death rates. A total of 380 930 births met the criteria for inclusion. Fifty seven early neonatal deaths were observed, reflecting the low mortality risk of these selected births. Babies born at night were almost twice as likely to die as babies born during the day (relative risk=1.86; 95% confidence interval 1.10 to 3.13). For deaths related to asphyxia an even more pronounced relation was observed (3.89; 1.51 to 10.03; table).

Early neonatal mortality and timing of birth in low risk births in Hesse, Germany, 1990-8

View this table:


Our results confirm a higher early neonatal mortality in low risk babies born at night. This may be a result of staff's increased physical and mental fatigue during the night, when doctors in charge, at least in Germany, have usually worked through a complete day shift. Overreliance on less experienced staff may be another important reason for the higher risk of early neonatal death during the night. These phenomena are not specific to the NHS or the British population. Better designed shifts, resulting in shorter working hours or decreased workload with greater supervision by experienced staff at night, should be considered to reduce early neonatal mortality during the night.

Although our analyses are consistent with previously reported British results, some differences should be considered.1 2 We used slightly different definitions of night and day because the hours of day shifts are different in Germany. Applying the time categorisation of the British studies yielded almost identical results. Because of the nature of our database only deaths occurring during labour or in the first seven days of life could be traced. There was some concern that babies born during the day are more likely to be preterm or high risk babies who have had induced births. We therefore restricted our study population rigorously. The assumption that a death was related to asphyxia relied solely on the obstetrician's documentation of morbidity and reasons for death, which could be prone to error. Nevertheless, a higher early neonatal mortality in general and a higher mortality related to asphyxia were seen. Additionally, for each deceased child each author reviewed all the information available from the register's database for other potential confounding factors. No alternative explanations for the reported relationship were found.


We thank Christiane Gasse for revising the manuscript and helpful comments.

Contributors: GH had the idea for the paper, performed statistical analysis, wrote the paper, and is the study guarantor. BM gave access to data, participated in performing statistical analysis, and commented on the draft. SS helped to write the paper. All authors reviewed the database for each deceased child.


  • Funding None

  • Competing interests None declared.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
View Abstract