PubMed2000
Knowing the vital statistics of a population is fundamental in controlling morbidity and mortality and improving living conditions. In Brazil, however, the available health information systems do not provide reliable vital statistics. This study was carried out in Quixadá, Icapuí, and Jucás, three municipalities in the state of Ceará that had good coverage by primary health care services. The study used an epidemiological instrument known as a "verbal autopsy" and investigated 215 (90%) of the 237 deaths of children younger than 1 year identified in 1993 and 1994 in the three communities. We investigated socioeconomic characteristics; sanitary conditions; nutritional status; the course of illness, health care, and death; the cause of death; and the operation of the national mortality information system and of the community health agents system. According to the verbal autopsies, diarrhea was the cause of death in 39% of the cases, followed by premature birth (17%), and acute respiratory infections (10%). Even though 79% of the families had sought formal health care services during the child's illness, 49% of the infants had died at home. This suggests limited effectiveness in the identification and treatment of sick infants. In 84% of the cases the family sought help from folk healers. Although community health agents reported 78% of the deaths, only 29% of the families had sought help from the agents during the children's illnesses. In terms of the statistical agreement between the information on the cause of death provided by community health agents and by the verbal autopsies, the agreement was good for diarrhea, intermediate for other causes, and low for acute respiratory infections. Making verbal autopsy a routine part of primary health care services in Ceará would provide invaluable information for local health care teams and would raise a critical consciousness fostering a reduction in infant mortality.