OBJECTIVE: This study sought to determine if violence against women is accurately documented in the trauma registry, and if poor documentation in the medical record is associated with incorrect coding in the registry. DESIGN: Retrospective cohort study. MATERIALS AND METHODS: We identified women aged 15 to 49 in the trauma registry of a regional medical center who had unintentional and intentional injuries over three years, and retrospectively reviewed their medical records to verify registry coding. MEASUREMENTS AND MAIN RESULTS: Of the 41 assault victims in the registry, 32 were verified by the medical record. Of the 87 unintentional injuries, only 28 were verified; 21 were assault victims according to the medical record, and for the remaining 38, the medical record was too vague to determine intentionality. Thus, the sensitivity of the trauma registry in documenting violence against women was only 57%. Injuries correctly coded in the registry had the details well documented in the medical record, whereas injuries incorrectly coded had poor documentation in the medical record. CONCLUSIONS: Violence against women often goes undocumented in hospital data systems.
OBJECTIVE: This study sought to determine if violence against women is accurately documented in the trauma registry, and if poor documentation in the medical record is associated with incorrect coding in the registry. DESIGN: Retrospective cohort study. MATERIALS AND METHODS: We identified women aged 15 to 49 in the trauma registry of a regional medical center who had unintentional and intentional injuries over three years, and retrospectively reviewed their medical records to verify registry coding. MEASUREMENTS AND MAIN RESULTS: Of the 41 assault victims in the registry, 32 were verified by the medical record. Of the 87 unintentional injuries, only 28 were verified; 21 were assault victims according to the medical record, and for the remaining 38, the medical record was too vague to determine intentionality. Thus, the sensitivity of the trauma registry in documenting violence against women was only 57%. Injuries correctly coded in the registry had the details well documented in the medical record, whereas injuries incorrectly coded had poor documentation in the medical record. CONCLUSIONS: Violence against women often goes undocumented in hospital data systems.