[Investigation of risk factors for infant mortality by linking health databases].

In order to identify risk factors for infant mortality (< 1 year of age) in Salvador, Bahia State, Brazil, by means of data bank linkage, a case-control study was performed, selecting individuals from the Mortality Information System (SIM; 2000 and 2001) and the Information System on Live Births (SINASC; 2000). The database linkage or data-sharing technique was used, with the Access 2000 software, version 9.0. Independent variables were collected from the SINASC database. The association between potential risk factors and infant death was evaluated by logistic regression. Prematurity, maternal occupation as a domestic servant, housewife, or student, delivery in public health services, insufficient number of prenatal visits, and low birth weight were predictors of infant death. Linkage revealed missing and incomplete data. Only 40.9% of data were recorded electronically. Despite these limitations, data linkage allowed better use of the two systems and the identification of critical points to further improve their quality.