Boston researchers reported today a novel use for electronic medical records -- using data in patient records, they say they were able to identify likely victims of domestic abuse an average of two years before a diagnosis was actually made.
Ben Reis, Dr. Isaac Kohane, and Dr. Kenneth Mandl of Children's Hospital Boston and Harvard Medical School studied six years of hospital admissions and emergency visits for patients over 18 years old. Based on the patient's history, including injuries and assaults, they determined whether patients met a definition of domestic abuse. Then they looked at actual diagnoses of domestic abuse.
"Our model predicted abuse two years before it appeared on medical records," Reis said in an interview. The article appears online in the British Medical Journal.
The risk factors linked to a future domestic abuse diagnosis differed between men and women. For women, the red flags were trips to the hospital to treat injuries, poisoning, and alcoholism. For men, depression and psychosis were associated with the greatest risk.
The researchers developed a visual display that could become part of a patient's electronic health record. The work is not ready to be implemented, they said, but the model could form the basis for an early warning system that would help busy doctors decide which patients need further screening and perhaps intervention.
"This is not a diagnosis but a screening support system," Reis said.
Their hope is to bring the wealth of information about a patient to the forefront during a doctor-patient encounter encumbered by competing demands. They plan to study other health problems, from diabetes to depression, that might lead to what they call "predictive medicine."
Dr. Thomas Bodenheimer, a professor of family and community medicine at University of California, San Francisco, has written about the pressures on primary care doctors to deal with all the problems patients have, not just the urgent ones that bring them to the doctor's office or hospital. Both paper and electronic health records already have "problem lists" designed to alert doctors, he said, but the typical 15-minute appointment may not allow all problems to be dealt with.
"I'm not arguing against the paper," he said. "Even if the researchers' study is absolutely correct and makes total sense, making it happen every day in every practice is going to be a huge challenge, especially with the time and number of agendas we have to deal with in the short time."
Dr. John Halamka, chief information officer for Harvard Medical School and an emergency doctor at Beth Israel Deaconess Medical Center, takes a more optimistic view.
"The paper is a great example of how electronic health record data can enhance care coordination and improve quality," he said in an e-mail interview. "As an emergency physician, I see a patient at a single point in time. To discover trends, make predictions about future health, and provide more patient centric medicine, we need to ensure clinicians record data electronically and share it among caregivers with patient consent. Early detection, as outlined in the paper, is just one of the breakthroughs we can achieve."
About white coat notes
|White Coat Notes covers the latest from the health care industry, hospitals, doctors offices, labs, insurers, and the corridors of government. Chelsea Conaboy previously covered health care for The Philadelphia Inquirer. Write her at firstname.lastname@example.org. Follow her on Twitter: @cconaboy.|
Gideon Gil, Health and Science Editor
Elizabeth Comeau, Senior Health Producer