Background: Perinatal mortality is considered a good effectiveness health care indicator during pregnancy, deliver and early perinatal period. The aims of this investigation were to describe the magnitude and tendency of underreporting perinatal mortality in Barcelona, Spain from 1985 to 1994 and to determine underreporting predictive variables.
Subjects and methods: Perinatal deaths between 1985 and 1994 were studied according to WHO international statistics criteria and deaths from 1988 to 1994 were also studied according to the WHO national statistics criteria. Sources of information were: Statistical Deaths Certificates and the Active Surveillance Registry that collect information about perinatal deaths directly from hospitals. A perinatal death was considered underreported if it was registered in the Active Surveillance Register but not in the Statistical Death Certificate.
Results: 24.5% perinatal deaths were underreported from 1985 to 1995 (according to international statistics criteria) and 24.9% from 1988 to 1994 (according national statistics criteria). In both periods underreporting decreased. Deaths in the first 24 hours of live (OR = 1.8; CI 95% = 1-3), newborns weight between 1,000 and 1,499 g (OR = 1.5; CI 95% = 0.6-3.8), hospitals with mortality registry (OR = 4; CI 95% = 2.2-7.1) and the first years of the study (OR = 0.7; CI 95% = 0.7-0.8) were predictive of underreporting.
Conclusions: Underreporting perinatal deaths decreased during the study period. Deaths of newborns with low birth weight and who lived less than 24 hours were more underreported. The work developed by Active Surveillance Registry contributed to the decense of underreporting perinatal deaths.