"Intelligent" Medical Histories Could Flag Domestic Abuse Sooner
Large study shows a long-range predictive view can provide an early warning
September 30, 2009
Boston, Mass. -- Tapping commonly available electronic health records, predictive computer models could help doctors diagnose domestic abuse an average of 10 to 30 months earlier, by highlighting subtle patterns that are easy to miss, report researchers from Children's Hospital Boston.
Their study, based on a review of anonymized electronic data from more than half a million adults, was published online September 29 in the British Medical Journal.
Domestic abuse is the most common cause of nonfatal injury to U.S. women, accounting for more than half of all murders of women every year. Medical records, if carefully examined, often contain clues that hint at possible abuse. But with patient encounters typically lasting under 10 minutes, doctors often lack the time and resources to carefully review and interpret information from multiple visits over many years. While universal screening for abuse is encouraged, actual screening rates remain low in many settings.
The researchers, led by Ben Reis, PhD, of the Children's Hospital Informatics Program (CHIP) and the Division of Emergency Medicine at Children's, analyzed six years of anonymized insurance claims for hospitalizations and emergency-department visits by more than 560,000 patients over 18. All patients had visits recorded over at least a four-year period; overall, 1 to 3 percent had an abuse diagnosis on record, depending on the case definition used.
"Risk gel" visualizations:
These sample visualizations present a quick, broad overview of two anonymized patients' diagnostic histories during the four years preceding their abuse diagnosis. Each small colored bar represents a diagnosis recorded at a particular visit (most recent visits at the bottom), grouped into 12 categories (injury, psychiatric, etc.). Each bar's color denotes the degree of abuse risk it reflects (green = low; yellow = medium; red = high). The blue arrows at right indicate when the model would have detected a high abuse risk, assuming a false-positive rate of 5%. For patient A, with few recorded visits, abuse risk would have been detected 27 months before the recorded abuse diagnosis. For patient B, with many visits, abuse risk would have been detected even earlier--34 months before it was actually diagnosed.
Using data from two-thirds of the patients, a computer model was trained to differentiate those who ultimately received a diagnosis of abuse from those who didn't, based solely on their history of visits. The variables associated with abuse (such as a higher number of annual visits, mental health diagnoses, and visits for injury) were used to create a predictive model, which was tested on the remaining third of the patients.
Reis and his CHIP colleagues Kenneth Mandl, MD, MPH and Isaac Kohane, MD PhD, found that the model was able to identify these patients an average of 10 to 30 months before the diagnosis was made, with high sensitivity and specificity.
"This is an important step towards the goal of Predictive Medicine," says Reis. "This decision support tool can help doctors identify abuse earlier by reminding them to perform standardized screening when certain patterns appear. By providing doctors with this additional safety net, we're hoping to minimize the chances that a high-risk patient falls through the cracks."
The analysis also found some gender differences in the statistical signs associated with abuse. Alcoholism, poisoning and injuries from external causes were more predictive of abuse in women than in men, while psychoses, affective disorders, and other mental conditions were more predictive in men than in women.
Reis notes that the database used may underestimate the true incidence of abuse, which is sometimes undiagnosed or simply not recorded in medical records. "The data set we used to train and evaluate the predictive model has all the challenges of real-world data," says Reis. "The important finding here is that in spite of these real-world limitations, the model was able to produce useful and reliable results."
Reis also developed a color-coded visual display to help physicians quickly process large amounts of information, integrating the patient's entire diagnostic history into one easy-to-digest view (see accompanying images) and displaying an alert if the history is suggestive of abuse. "Our goal is to communicate an overview of a patient's longitudinal history to the doctor in 10 seconds," says Reis. "We're hoping doctors with access to this kind of view will be able to provide better, more-informed care, including detecting certain medical conditions earlier."
Planned future studies will validate, refine and adapt these models to different health care environments, including HMO patients and pediatric populations - thus extending the model to child abuse. Reis and colleagues also plan to develop similar models to help predict other conditions that sometimes go undiagnosed, such as depression, or diseases like diabetes that can be missed in their early stages.
The study will also put various prototype patient-history displays to the test with focus groups of doctors operating in mock health-care settings, to see which display is most helpful to physicians in spotting possible cases of abuse or other diagnoses.
Children's Hospital Boston is home to the world's largest research enterprise based at a pediatric medical center, where its discoveries have benefited both children and adults since 1869. More than 500 scientists, including eight members of the National Academy of Sciences, 11 members of the Institute of Medicine and 12 members of the Howard Hughes Medical Institute comprise Children's research community. Founded as a 20-bed hospital for children, Children's Hospital Boston today is a 396-bed comprehensive center for pediatric and adolescent health care grounded in the values of excellence in patient care and sensitivity to the complex needs and diversity of children and families. Children's also is the primary pediatric teaching affiliate of Harvard Medical School. For more information about the hospital and its research visit: www.childrenshospital.org/newsroom.