OBJECTIVE: to describe the common characteristics of elderly people admitted to public hospitals of Niterói (RJ) due to falls. METHODS: Cross-sectional study with 110 elderly patients (60 years and more) admitted to public hospitals in Niteroi - RJ, with fractures after falls. Data were collected by a multidimensional questionnaire. Multiple correspondence and cluster analysis were used to describe the characteristics of this population. RESULTS: The results indicated four groups. The first group was formed by individuals with better physical conditions who did not have a hip fracture, were more independent, used to go out more than twice a week, had good vision, had a short hospital stay, and were discharged to their homes. There was a group with worse conditions including those who were less independent, had a long hospital stay (more than 30 days), and died at the hospital. A third group was formed by those who had had a hip fracture, had poor vision, and used to go out less than once a week. The fourth group was comprised of those who stayed in the hospital from 11 to 30 days and used to go out once or twice a week. CONCLUSIONS: The multiple correspondence analysis proved to be a good technique to identify subgroups with common characteristics, which provides tools for the creation of strategies for fall prevention programs.
Correspondence analysis; Cluster analysis; Fractures; Falls; Elderly; Hospitalization