[Differences in perinatal mortality between districts of Amsterdam].

Objective: To study the effect of district on perinatal mortality in Amsterdam, the Netherlands, taking into account various risk factors including ethnicity and social economic status (SES).

Design: Cohort study.

Method: The investigation related to 73,661 singleton births in Amsterdam, Diemen and Ouder-Amstel recorded in the Netherlands Perinatal Registry over the years 2000-2006. Logistic regression analysis was used to determine if perinatal mortality differed by district, taking into account various risk factors.

Results: Each year in Amsterdam an average of 10,525 singleton children were born of whom 114 infants died (10.8 per 1,000 births (‰)). National perinatal mortality was 9.9 ‰. In three districts, perinatal mortality was 1.5-2 times higher than the national average: Zuidoost (21‰), Slotervaart (14‰) and Zeeburg (14‰). However, mortality in the districts of ZuiderAmstel (5‰), Oud-Zuid (7‰), Centrum and Osdorp (8‰) was 20-50% lower. The high risk of perinatal mortality in the Zuidoost district (odds ratio: 2.1; 95% CI: 1.9-2.6) was explained by the high prevalence of women with higher risk factors; African or South Asian Surinamese ethnicity, low SES and preterm birth. The effects of parity and ethnicity on perinatal mortality differed by district. In Zeeburg increased effect for higher parity and for Turkish/Moroccan ethnicity was seen. In Slotervaart the perinatal mortality risk was increased (odds ratio: 1.8; 95% CI: 1.3-2.5), but this was not explained by the risk factors studied.

Conclusion: Amsterdam had districts with both highly elevated and reduced perinatal mortality rates. The prevalence of risk factors differed by district and the effects of ethnicity and parity were not homogenous. Therefore, tailored policy and research by district is necessary.