Overconfident Docs Need Dose of Reality

Misdiagnoses Occur up to 15% of the Time, and Physician Overconfidence May Be Partly to Blame, Study Shows

Medically Reviewed by Louise Chang, MD on April 30, 2008

April 29, 2008 -- Most of the time a medical diagnosis is on point. Butmisdiagnoses do occur, and an overly confident doctor may be partly to blame, anew review suggests.

The rate of diagnostic error is as high as 15%, Eta S. Berner, EdD, and MarkL.. Graber, MD, write in a special edition of The American Journal ofMedicine dedicated to understanding and addressing diagnostic errors.

Physician overconfidence and a lack of feedback following a diagnosis aretwo important contributors to the problem, they note.

"When directly questioned, many clinicians find it inconceivable thattheir own error rate could be as high as the literature demonstrates,"Berner and Graber write. "They acknowledge that diagnostic error exists,but believe the rate is very low, and that any errors are made by others whoare less skillful or less careful."

Berner says it is often the cases physicians perceive as routine andunchallenging that end up being misdiagnosed.

"With the hard cases, doctors generally seek out different opinions orturn to (computer-based) decision support tools," she tells WebMD.

(Have you everbeen misdiagnosed? Tell us about it on WebMD's Health Cafe message board.)

A Missed Diagnosis

Retired engineer Paul Mongerson is all too aware of the problem of medicalmisdiagnosis, and he has spent the last 28 years addressing the issue.

In 1980, Mongerson was incorrectly told by four different doctors that hehad pancreatic cancer, a highlydeadly cancer that kills most peoplewho have it within five years.

Mongerson made up a matrix charting his symptoms and test results to helphim assess the probability that his doctors were right.

"I determined from that matrix that I didn't have cancer," he tellsWebMD.

Just two days before he was scheduled for cancer surgery, a fifth doctoragreed and Mongerson canceled the operation.

"I said at the time that if I survived I was going to see what I coulddo to help other people," he says.

What Mongerson did was form a foundation to fund the development ofcomputer-based programs designed to assist physicians in diagnosingdisease.

While such programs are being used in many hospital and educationalsettings, they are not yet widely used by private practice physicians.

Mongerson says performing more autopsies and having systems in place tocrosscheck medical diagnoses would help address the issue of lack offeedback.

Barriers to Patient Follow-up

In one of the newly published essays, Gordon D. Schiff, MD, associatedirector of the Center for Patient Safety Research and Practice at Brighamand Women's Hospital, addressed the barriers to the follow-up of patients inthe real-world, clinical practice setting.

Not surprisingly, lack of time was at the top of his list, followed byfragmentation of care, the large number of symptoms for which there is no cleardiagnosis, cost and managed care barriers, and physician defensiveness aboutcritical feedback from peers.

"Learning and feedback are inseparable," Schiff writes. "The oldtools (used by physicians) -- individual idiosyncratic systems to trackpatients, reliance on human memory, and patient adherence to or initiating offollow-up appointments -- are too unreliable to be depended upon to ensure highquality in modern diagnosis."

He calls for a systematic approach to link diagnoses with patientoutcomes.

In a different essay, Mark Graber, MD, of the department of medicine atState University of New York at Stony Brook and VA Medical Center in Northport,N.Y., proposes new roles for patients that can help. One is to have the patientbecome a "watchdog for cognitive errors" by having doctors communicateto patients more about what diagnoses they are considering rather than justtelling patients what tests to get or what medications to take. Sharingmore information with patients can help patients be more active in checking forerrors.

A second role is as a "watchdog for system-related errors" to helpkeep track of their own medical information such as test results and medicationlists. By doing so, "the patient can play a valuable role in combatingerrors related to latent flaws in our healthcare systems and practices,"Graber writes.

Berner adds that patients can help by questioning their doctors carefullyduring the diagnostic process, and, especially, letting them know when theymight have made the wrong call.

"If your doctor says you should be better in a week, and you aren't,call the office and let them know," she says, adding that a surprisingnumber of patients do not do this.

Patients who aren't sure about their diagnosis should also ask their doctorswhat else their condition might be, she says.

The simple suggestion was a major focus of the best-selling 2007 book HowDoctors Think by Harvard Medical School physician Jerome Groopman.

In it Groopman writes that instead of being intimidated by their doctors,patients should ask questions like, "Is there anything that doesn't fityour diagnosis?" and "Is it possible that I may have more than oneproblem?"

Mongerson tells WebMD that the point is not to put physicians on thedefensive, but to explore all medical possibilities.

"After everything I went through I am still very high on doctors,"he tells WebMD. "They are very dedicated people who work very hard and gothrough hell when they find out they have made a mistake. The problem is, theydon't normally find out."

Show Sources


Berner, E.S. The American Journal of Medicine, supplemental issue, May 2008; vol 121.

Eta S. Berner, EdD, professor, department of health services administration, University of Alabama at Birmingham.

Gordon D. Schiff, MD, associate director, Center for Patient Safety Research and Practice, Brigham and Women's Hospital, Boston.

Paul Mongerson, retired engineer; founder, Paul Mongerson Foundation, Naples, Fla.

Groopman, J. How Doctor's Think, Houghton Mifflin, 2007.

Graber, M. The American Journal of Medicine, supplemental issue, May 2008, vol 121.

© 2008 WebMD, LLC. All rights reserved. View privacy policy and trust info