Should births be centralised in higher level hospitals? Experiences from regionalised health care in Finland


Prof E Hemminki, THL (National Institute for Health and Welfare), PO Box 30, 00271 Helsinki, Finland. Email


Please cite this paper as: Hemminki E, Heino A, Gissler M. Should births be centralised in higher level hospitals? Experiences from regionalised health care in Finland. BJOG 2011; DOI: 10.1111/j.1471-0528.2011.02977.x.

Objective  To describe: (i) trends in centralisation and unplanned out-of-hospital births; (ii) perinatal mortality by place of birth; and (iii) health and birth outcomes in areas served by hospitals of different levels.

Design  Cross-sectional analysis of medical birth register data.

Setting  Finland, from 1991 to 2008, and Uusimaa district from 2004 to 2008.

Population  All births.

Methods  In the hospital-based analysis, birthweight was adjusted by logistic regression. In the area-based analysis results were calculated according to where women lived, grouping them into areas served by different hospitals. The mother’s background characteristics were adjusted for by logistic regression.

Main outcome measures  Place of birth, unplanned out-of-hospital birth, perinatal mortality, newborn outcomes, and birth procedures.

Results  The number of birthing hospitals declined, the mean number of births per hospital increased, and more births, particularly high-risk births, occurred in university hospitals. Unplanned out-of-hospital births were rare, and their numbers increased in the 2000s, but regional differences declined. Perinatal mortality was higher in the university hospitals than in other hospitals, but after adjusting for birthweight, it was lower. Among children weighing more than 2500 g, mortality was similar for all hospital levels. In out-of-hospital births, perinatal mortality was much higher than in other children. The area-based analysis did not systematically show better or worse results for the areas served by lower level hospitals: after adjusting for the background characteristics of the mothers, all differences were found to be small.

Conclusions  The health and service data do not support the need to close down small hospitals in a regionalised system where there is a referral system that functions well.