INFANTILE MORTALITY-RATES IN A SEMIURBAN AREA OF THE TUNISIAN PORTIONOF THE TUNISIAN SAHEL

High infantile mortality rates remain a public health problem in devel oping countries. Only the prospective approach can provide reliable es timates of infantile mortality rates (IMR). The cohort of 1145 infants born alive in 1991 in a semi-urban part of the Sousse region, Tunisia , were monitored for two years. There were 38 deaths, yielding an IMR of 33.19 per 1000. Early neonatal mortality contributed a large portio n of these deaths: 28 deaths occurred before the age of four weeks, in cluding 22 during the first week of life. Ten infants died between the ages of one and six months. The main risk factors were related to the course of the pregnancy and delivery and to the status of the infant during the first few hours of life. Thus, infection, neonatal distress , prematurity, and complications of pregnancy took the heaviest toll. Many of the deaths were related to multiple factors. Over the last few decades, postneonatal mortality has decreased in Tunisia, and most in fant deaths now occur during the early neonatal period. This new situa tion requires the development of programs aimed at improving prenatal, obstetrical, and neonatal care. Local surveillance of IMRs as part of research-action programs is the best approach for raising the quality of health care available to mothers and children.