[The determinants of the high Dutch perinatal mortality in a complete regional cohort, 1990-1994].

Objective: Analysis of the effects of population-based determinants (maternal age, parity, multiple pregnancy and ethnicity) and of professional and organisational factors (conservative management in case of early preterm birth, the policy on prenatal screening and the Dutch obstetric-care system in general) on perinatal mortality.
Design: Population-based prospective cohort study.
Method: In a regional cohort (Zaanstreek) of 8031 pregnancies in the period 1990-1994 data were prospectively collected in an electronic database by deliverers of primary and secondary care. Analysis focussed on the effect on perinatal mortality of maternal age, parity, multiple pregnancy, ethnicity and professional and organisational factors (the policy in case of early preterm birth, the policy on prenatal screening and the Dutch obstetric-care system in general). Perinatal mortality was defined as mortality from a gestational age of 22 weeks until 28 days post partum.
Results: The perinatal mortality in the Zaanstreek was 12.6/1000. Increasing the gestational age to 28 weeks decreased the perinatal mortality by 29%. The perinatal mortality in this cohort was significantly affected by parity, multiple pregnancy and maternal age (relative risk: 2.8), but not by ethnicity if corrected for the previous factors. Conservative management in case of early preterm birth and a restrictive screening policy for lethal birth defects were associated with an increase in perinatal mortality. In 31 of 92 singleton pregnancies followed by perinatal mortality, a relationship to substandard care was established. In 7 cases this relationship was probable and in 1 case the midwife was responsible.
Conclusion: Given the magnitude of their effects, both independently and via interaction, stratification for maternal age, parity, multiple pregnancy and ethnicity must precede any interpretation and comparison of perinatal mortality rates. Although clinical policy played a modest role, a negative role of the organisation of obstetric care was unlikely in this cohort. A definitive judgement as to the quality of perinatal care would require extension of the evaluation to at least the entire first year of life in connection with the morbidity. The most favourable effects can be expected from stimulatory measures directed at lowering the age at first pregnancy.