RISK FACTORS PERINATAL MORTALITY IN PERU

Abstract

OBJECTIVE: To identify Peruvian perinatal mortality risk factors at both population level (ENDES) and hospital level (SIP) and to determine their predictive value. DESIGN:Epidemiologic, analytical, case-control study using information from both the National Survey of Demography and Health (ENDES) and the Perinatal Computer System (SIP) of 9 hospitals of the Ministry of Health corresponding to the year 2000. MATERIAL AND METHODS: The study included mothers with gestations more than 28 weeks or newborns weíght above 1000g. We used rate per thousand births, odds ratio (OR) with 95% confidence intervals, population attributable risk percentage (RAP%), logistical and curved regression ROC (receiving operating characteristic). RESULTS: The rate of perinatal mortality in Peru in the year 2000 was 23, 1 per thousand births according to ENDES, similar to the 22,9 rate of the hospital study. ENDES risk factors were rurality (OR= 1,5), altitude > 3000 masl (OR= 1,7), language different to spanish (OR= 1,8), agriculture occupation (OR= 1,6), low scholarship (OR= 1,5), short intervals between pregnancies (OR=4,7) and domiciliary childbirth (OR= 1,8), being the three last susceptible of control. The hospital risk factors were: mountain region (OR= 2,l), altitude>3000 mas1 (OR= 1,8), low scholarship (OR= 2,3), old maternal age (OR= 1,6), previous fetal death (OR= 1,9), short pregnancy interval period (OR= 4,5), multiparity (OR= 1,9), absence of prenatal control (OR= 2,l), anomalous presentation (OR= 4,3), maternal and neonatal pathology (OR= 2,4 and 56,7), low birthweight (OR= 9,8), prematurity (OR= 5,6), intrauterine malnutrition (OR= 5,5), low Apgar score at one and 5 minutes (OR= 4,3 and 46,6); the last seven were factors susceptible of control. Three ENDES risk factors-short interval between pregnancies, altitude> 3000 masl and absence or inadequate prenatal control- had low predictive value (62%); at hospital level predictive value was 96% with low birthweight, prematurity, newborn depression, neonatal morbidity and multiparity. CONCLUSIONS.- Risk factors identified by ENDES at population level and by SIP at hospital level are complemented, the latter with higher predictive value por perinatal mortality. We recommend the inclusion of hospital information in future ENDES.