INTRODUCTION: Women very rarely die during pregnancy and childbirth in Denmark. Although maternal deaths are registered worldwide, various studies indicate that underreporting does occur. This paper presents validated Danish register data for two periods between 1985 and 2017.

METHODS: Maternal deaths were identified from 1985 to 1994 and from 2002 to 2017 by linking four national health registers, death certificates and notifications from maternity wards. A group of obstetricians categorised and assessed all medical records, classifying each case by cause of death.

RESULTS: Linkage of four registers yielded valid data, leading to the identification of 143 maternal deaths in the abovementioned periods. From 1985-1994 there were 73 deaths and 618,021 live births, resulting in a maternal mortality rate of 11.8 per 100,000 live births with a non-significant 2% annual increase (95% confidence interval (CI): -6.0-11.0%). From 2002 to 2017 there were 70 maternal deaths and 999,206 live births, resulting in a maternal mortality rate of 7.0 per 100,000 live births (95% CI: 5.5-8.9) with a significant 9% annual decrease (95% CI: 4.0-14.0%).

CONCLUSIONS: Overall maternal mortality decreased in the course of the two periods (n = 33 years), with a significant decrease during the last period. This is suggested to be a result of multiple clinical and organisational improvements as discussed in the paper.

FUNDING: none.

TRIAL REGISTRATION: not relevant.