Peggy Zuckerman was on her way to a concert at the Hollywood Bowl back in 2004 when her doctor called. Blood work she’d had in preparation for minor surgery showed she was dangerously anemic, he said, and she needed to get to the emergency room. Instead of enjoying “Madama Butterfly,” Zuckerman spent that night and the next at the hospital, getting three blood transfusions, an endoscopy and a colonoscopy to figure out where the problem might lie. She was told the bleeding was likely due to a tiny, scabbed-over stomach ulcer and sent home. But seven months later, she still wasn’t feeling well. “By this time, I was just dragging tail,” recalls Zuckerman, 67. Finally, an ultrasound revealed a softball-sized mass on her right kidney. The cancer had also spread to her lungs.
Zuckerman was fortunate. She responded well to surgery and immunotherapy, and has been free of kidney cancer since. (She’s now receiving care for a separate early-stage breast cancer.) But as she recovered, she started to look into why she hadn’t been diagnosed earlier. “When I was well enough to lift my head up and read,” she says, she learned that hemoglobin levels as low as hers would have required significant bleeding, not “a small, theoretically scabbed-over ulcer.” She reviewed her records and spotted missed opportunities, including an ER physician’s recommendation — never passed on to her — that she see a rheumatologist or a hematologist, and a pathology report showing no ulcer. Misdiagnosis is common with kidney cancer, which often presents with seemingly unrelated symptoms. As Zuckerman, who is now a patient advocate, did more research, though, she realized poor calls are a far broader problem. “It’s not just odd cancers or rare diseases,” she says.
That worrisome reality was underscored last year by the National Academies of Sciences, Engineering, and Medicine in a report on the urgent need to improve diagnosis. The authors defined diagnostic error as the failure to establish an accurate and timely explanation of the problem or to communicate that information to the patient. And they concluded that most people will experience at least one diagnostic error, “sometimes with devastating consequences.” Autopsy reports suggest that these mistakes contribute to about 10 percent of patient deaths, while medical record reviews suggest they cause up to 17 percent of adverse events in hospitals.
[See: 5 Common Preventable Medical Errors.]
One way doctors err is by leaping to conclusions before all the evidence is in. Tamara Patterson, a 35-year-old therapeutic yoga instructor in Mount Shasta, California, recalls telling her physicians over and over again that her lower abdominal pain wasn’t alleviated by a hernia repair and that her pain and bleeding varied with her menstrual cycle. “I said, ‘I don’t think it’s the hernia.’ And I literally had four clinicians look at me like I was a lunatic.” It turns out she had endometriosis. In other cases, key information isn’t passed on from one clinician to the next. Or test results fall through the cracks. “This is a safety issue,” says Dr. Mark L. Graber, a member of the committee that authored the report and president and founder of the Society to Improve Diagnosis in Medicine. “It might be the biggest one.”
A report in the journal Diagnosis in October 2014 described one example. A feverish patient had walked into a Dallas emergency department complaining of dizziness, nausea, abdominal pain and headache. He told the nurse that he’d recently traveled to Liberia, which should have raised a red flag. But according to the case report, that information “was not appreciated or acted upon.” The patient was prescribed antibiotics, told to take Tylenol and discharged. He was back two days later, severely ill with Ebola, and he infected two nurses who were caring for him before he died.
That misdiagnosis made headlines. Most aren’t even identified. It can be extremely difficult to pin down exactly how often the diagnostic process goes wrong or at what step, or even to define exactly when that occurs. A rare disease, for example, may take a while to diagnose correctly, not because physicians missed clear signals but because knowledge of or experience with the condition is lacking, says Dr. Hardeep Singh, a patient-safety researcher at the Houston Veterans Affairs Health Services Research Center of Innovation and an associate professor of medicine at Baylor College of Medicine, who co-authored the Ebola case report.
Singh is one of the researchers trying to get a handle on the problem. A 2013 study in JAMA Internal Medicine that he co-authored found that in primary care, commonly missed conditions included pneumonia, heart failure, cancer and infections. The breakdowns happened most frequently during the patient-clinician encounter — a medical history or physical exam was incomplete, for example. But many cases also went wrong because of factors such as a specialist referral not made or incorrect test interpretation. Experts say much more work needs to be done to characterize and measure misdiagnosis so they know where to focus their efforts.
In the meantime, some hospitals and physicians have begun to tackle the problem in multiple ways, from changing the processes that lead to a diagnosis and the health system culture to teaching physicians to think differently about, well, how they think.
[See: 8 Questions to Ask Your Pharmacist.]
One important innovation: encouraging reporting. “Diagnostic errors were having a significant effect on patient safety, but we weren’t identifying or responding to them in the way we did to other safety problems,” says Dr. Robert Trowbridge, a hospitalist and medical educator at the Maine Medical Center in Portland. In 2011, he led a six-month pilot program that educated doctors about misdiagnosis and let them anonymously report diagnostic errors by clicking a computer desktop icon and briefly describing what went wrong.
The findings: 36 mistakes, none of which had been flagged by other monitoring systems. The ability to report diagnostic mistakes is now built into the error reporting system, and while there’s no hard evidence, Trowbridge thinks the system is reducing their frequency.
The switch to electronic medical records is often blamed for missed signals because a focus on the computer screen can divert clinicians’ attention from the patient. But plenty of experts think health IT has the potential to improve diagnosis. For example, involuntary weight loss is associated with a higher risk of death since it can indicate a serious health problem. But changes in a patient’s weight over time may not be noticed by his or her physician, says Dr. Gordon Schiff, an internist and associate director of the Center for Patient Safety Research and Practice at Brigham and Women’s Hospital in Boston. When he and colleagues crunched electronic medical record data, they found that 45 percent of patients who unintentionally lost 10 percent or more of their body weight in a year had medical conditions emerge in the next two years that might have been related. What if software automatically searched for that kind of pattern? “We have to make the computer work harder,” Schiff says.
Kaiser Permanente is already using such triggers. The SureNet system is a series of alerts that tips health professionals off to potential lapses in care “before harm comes to the patient,” says Dr. Michael Kanter, regional medical director of quality and clinical analysis at Kaiser Permanente, Southern California Permanente Medical Group. Take chronic kidney disease, which is diagnosed on the basis of two elevated blood creatinine levels taken three months apart. With SureNet, if there’s no second test in the record within 90 days of the first, the system alerts both the physician and patient that another one is needed. A study published last year found that over a four-year period, SureNet identified 12,396 people in need of a repeat test, and 30 percent of them had chronic kidney disease confirmed.
While systems are clearly important, physicians are ultimately the ones making a diagnosis. That process — gathering information, framing a problem, evaluating possible solutions and choosing one — “is no different from the way we solve any other problem in our life,” says Dr. Gurpreet Dhaliwal, professor of clinical medicine at the University of California–San Francisco. It’s similarly subject to the mental shortcuts and biases everyone relies upon to get through the day. Even an experienced doctor who sees a particular infection several times in a week, for example, is going to be more likely to suspect his next patient’s identical symptoms are due to the same problem. Or a doctor might see an initial test result and quickly home in on a diagnosis, then filter out subsequent information that points to something else entirely.
Doing away with those cognitive shortcuts is neither feasible nor desirable, says Dhaliwal, since they often aid correct diagnosis, too. But it’s important to teach clinicians to be aware that their thinking is sometimes flawed. He urges physicians to follow up on their diagnoses to see if they made the right call and, if not, to figure out how they can do better next time. At the University of Pittsburgh Medical Center, Dr. William Follansbee, the master clinician professor of cardiovascular medicine, is using a checklist to help trainee doctors break the decision-making process into five specific steps, including describing the case briefly and asking themselves whether any available information is discordant with the diagnosis. “We can’t turn every physician into a cognitive psychologist, but we can make them more aware of when they might be getting into trouble,” Follansbee says.
To be sure, it’s impossible for every doctor to be deeply familiar with the latest science on every possible illness. So in some cases, fixing misdiagnosis may just be a matter of more widely and quickly disseminating new knowledge, says Dr. David Newman-Toker, an associate professor of neurology at Johns Hopkins who will head up a new diagnostic innovation center there. For example, he says, an estimated 4.4 million people show up in the ER with dizziness every year, and 3 to 5 percent of them are having a stroke. Specialists can help identify those people using tests that measure certain eye movements. Hopkins is starting to test a “tele-dizzy” program that would capture those eye movements with video goggles and send the recordings to specialists for evaluation, all within an hour.
Patients, too, play a key role in the process. You can improve your odds of getting good answers by providing an accurate description of symptoms, a list of your medications and your personal and family medical histories.
[See: 10 Lessons From Empowered Patients.]
To help reduce mental shortcut mistakes, patients also have to ask questions. “The universal antidote for patients is, when given a diagnosis, to ask ‘What else could this be?'” Graber says. You should also ask how the illness is expected to evolve, and what to do if it doesn’t go away or worsens. Ask for a second opinion if you receive a worrisome diagnosis.
You can help protect yourself by getting copies of all your medical records, checking for obvious errors and making sure they reflect what you’ve been told. And, says Dhaliwal, “you can’t work under the assumption that no news is good news.” Follow-up, he says, “should land on your to-do list as well as ours.”
More from U.S. News
5 Common Preventable Medical Errors
HIPAA: Protecting Your Health Information
8 Questions to Ask Your Pharmacist
Surviving a Bad Diagnosis originally appeared on usnews.com
