[Variables associated with underreporting of AIDS patients in Rio de Janeiro, Brazil, 1996]

The underreporting of AIDS cases in the municipality of Rio de Janeiro, Brazil, is significant. The study intends to analyze the factors associated to this event.

Using data provided by the Hospital Information System for the year of 1996, in Rio de Janeiro city, and by the National Surveillance System, patients were randomly selected and their medical records reviewed to verify an AIDS diagnosis. A multinomial model was used to perform an analysis of the variations on the chances of underreporting of AIDS cases versus reporting and on the chances of underreporting without evidences to fulfill the case definition of AIDS versus reporting.

No significant associations were found between the variables such as "age", "marital status", "level of education", "occupation", and "severity of illness" and the underreporting of AIDS cases. The variable "female gender" showed a strong association with hospitalization without evidence of an AIDS diagnosis. A strong association was found between two or more admissions in an inpatient unit care and reporting. The existence of a epidemiological surveillance department in the hospital is inversely associated with the underreporting of AIDS cases.

The significant association between organizational variables and underreporting of AIDS cases found in the study point out to the need of standardization of the surveillance procedures, the especial need for the creation and maintenance of surveillance departments in hospitals to improve the quality of the health information system and, therefore, AIDS prevention and care.