Domestic abuse often goes undiagnosed until too late — yet medical records often contain subtle clues that doctors often lack the time to fathom out. Now, researchers from the Children’s Hospital Informatics Program and Division of Emergency Medicine demonstrate that tapping commonly available electronic health records could help doctors spot abuse early. This display, designed for physicians, pulls a patient’s diagnostic history into one view, sounding an alert when the pattern of visits suggests possible domestic abuse.
Each colored bar above represents a diagnoses recorded in the patient’s chart, grouped by category, during the four years before her abuse diagnosis; the most recent diagnoses are shown at the bottom. The color coding denotes the degree of abuse risk, calculated using data from the study (green, low risk; yellow, medium risk; red, high risk). As indicated by the blue “detect” arrow, the system would have sounded an alert as early as 34 months before domestic abuse was actually diagnosed.
In the future, the researchers hope that their models can detect when a person is at risk for abuse before abuse even occurs. Although the study, published in the British Medical Journal, was done in adults, the plan is to validate the model in children, too, and to develop similar models for conditions that are often missed, like depression and early-stage diabetes. According to Ben Reis, PhD, who led the project, such “intelligent histories” are an important step toward the larger goal of predictive medicine — helping busy physicians not by making diagnoses for them, but by offering a decision support tool that can flag patients who merit specific screening.