Objective: To analyse the mode and cause of perinatal mortality. Setting: a
rural Dutch region. Study design: Over a two-year period ;(1994-1995), dat
a were collected in the 's Hertogenbosch region. A perinatal audit group in
vestigated and classified the cause of death in an "intention to treat" and
concensus model. We then analyzed who was responsible for the patient at t
he moment perinatal death occurred, or became inevitable. Results: Out of 8
509 newborns, 73 died between the 24th week of pregnancy till the 7th day p
ost-partum (8.58 promille). Twenty-three cases (31.50%) were classified as
probably or possibly avoidable. In the primary health care group (midwives,
general practitioners) 6 out of 32 (18.75%), in the secondary care group (
obstetricians) 15 out of 35 (44.86%) and in the tertiary care group 1 out o
f 4 (25.00%) were judged as probably or possibly avoidable. The degree of c
oncensus in the perinatal audit committee was high (Kappa=0.9). Impact: The
analysis of perinatal mortality identifies the cause of death and may help
to improve perinatal health care. Conclusion: In this study, 31.55% of per
inatal mortality was avoidable in the three levels of care. Intra-uterine g
rowth retardation, congenital malformations and antepartum haemorrhage were
the most determinant factors for perinatal mortality. The Dutch obstetrica
l care system as such, for example home deliveries, did not effect the peri
natal mortality rate. Perinatal mortality rates presented by the Dutch Cent
ral Bureau of Statistics still shows a slight underregistration. (C) 2000 E
lsevier Science Ireland Ltd. All rights reserved.