Objective: The main objectives of the present study were to evaluate the percentage of under-registration of infant mortality in 1993 and compare it with the ones found 1982; to analyze the agreement between the official death certificates and the ones made by the referees.
Methods: The infant mortality of all children born in Pelotas, in 1993, was monitored through daily visits to hospitals, as it was done in 1982; monthly, cemeteries and public registration offices were visited to detect any deaths outside the hospitals. Besides the official death certificates, two independent referees established the underlying cause of death based on information from pediatricians, case-notes, autopsies and through a home visit to the parents of the children.
Results: The percentage of under-registration fell substantially from 24%, in 1982, to 5.4%, in 1993. The agreement between the official death certificates and the ones made by the referees showed satisfactory Kappas, unless for ill defined diseases such as sudden infant death, where the agreement was null.
Conclusions: The authors conclude that there was a significant fell of the under-registration for infant death in Pelotas, and the ill defined causes such as sudden infant death have been hidden by the diagnosis of pneumonia.