ALEXANDRIA, Va. -- What are some of the biggest keys to reducing diagnostic error? Communication and teamwork, according to , senior fellow at RTI International.
"If you look at studies of diagnostic error, what's the root cause of the problem?" Graber said last week at the annual meeting of the . "Communication is always at the top of the list."
Communication is a great thing to focus on because "it plays out so many ways in the diagnostic process," said Graber, who is president of the society and also served on the Institute of Medicine (IOM) committee that wrote the recently released report on reducing diagnostic errors. "Handoffs ... and so many [other] things aren't communicated accurately. It's not just test results -- it's the whole patient story."
Just look at the example of the misdiagnosed Ebola case in Dallas, Graber said. In that instance, the patient, Eric Duncan, had told the nurse in the emergency department (ED) that he had traveled recently to Africa, and the nurse had put a note to that effect in Duncan's medical record, but that information was not considered by the doctor.
Duncan was initially given antibiotics and discharged, only to be readmitted several days later when his symptoms worsened; he eventually died from the disease. Duncan "might have been diagnosed had the nurse been more effective member of that team," said Graber. "It's the ward nurses, the nurses who work in the clinic, who need to feel and be respected as a member of the team."
Graber gave another example, this one from his own clinic at the VA Medical Center-Northport in New York. An elderly patient with chronic medical problems came into the office to talk about her issues. She mentioned to the nurse -- but not to Graber -- that she had fallen down the week before and her wrist was really bothering her. "She was in the [ED] a week later and had a fractured wrist ... because the nurse was not an effective part of the diagnostic team," he said.
One piece of low-hanging fruit when it comes to reducing diagnostic error is health information technology, Graber continued. "I love my electronic medical record (EMR)," he said. "For me it is a life-saver, especially in regard to diagnosis. I can read it; the tests are there; it helps me organize my thoughts; I can use it to communicate with specialists and consultants; and I can see graphic displays so I notice trends [more easily]."
He acknowledged that some features of EMRs are "just horrible," such as the copy/paste function. "We have to get rid of it; the minute you see a copy/paste note, you can't trust anything that's in that record," he said. Similarly, menu-driven documentation "is great for research and billing but not great for patient care."
Some people, such as , of the University of California San Francisco, say EMRs are also "communication killers."
"We just look at a report and don't think we have to talk to the radiologist," Graber noted. "But we do; it's absolutely critical."
But despite those problems, EMRs "are low-hanging fruit because we know what the problems are and know what the solutions are," he said. "The agenda is there; we just need to act upon it." That includes having health IT vendors work with the federal Office of the National Coordinator (ONC) for Health Information Technology to make sure EMRs demonstrate usability and incorporate human factors knowledge.
Graber lauded the IOM report for including a definition of misdiagnosis: the failure to establish an accurate and timely explanation of the patient's health problems or to communicate that explanation to the patient. "The definition gives us things to study in the research sense -- we can study accuracy, we can look at timeliness, and we can study communication," he said. "And one thing I like a lot is that it's got 'patient' in it twice."
The report did have a weak spot, he said -- it contained no policy recommendations, including recommendations on changing the reimbursement system. Graber noted that often when a physician is trying to make a diagnosis, there is no pathologist available to discuss a case, because there's no reimbursement for them to do so.
"If we change that we could perhaps improve the diagnostic process," he said. "The way the payment system is set up rewards procedures over diagnostic thinking; it's time to change that. Diagnosis takes time; that needs to be accommodated in our payment systems."
More research funding needs to be dedicated to reducing diagnostic errors, Graber suggested, noting that the current funding in this area is "a couple million dollars compared to billions spent on other problems. We have to achieve some better balance."
Graber has two favorite research questions: first, what are the cost implications of diagnostic error? "If we had that number, it might drive the policy changes we need to see," he said. "Every penny you spend on treating the wrong diagnosis is wasted -- that's a big issue."
The second question is: What interventions work to reduce diagnosis errors? "We need to start trying [interventions out] ... you never know until you try," Graber said. He mentioned a poster at the meeting showing that using a checklist for diagnosing patients could work. "I'm partially enthusiastic those things might work -- let's start looking at them."