Perspective: Diagnostic Errors in Medicine 2016

By Charles A. Pilcher MD FACEP
December, 2016.

Over 300 physicians, educators, insurance companies executives, risk managers, quality & safety specialists, patient advocates and medical society representatives assembled at the 2016 Diagnostic Error in Medicine (DEM) Conference in Los Angeles November 6-8 hosted by the Society to Improve Diagnosis in Medicine. The common goal of this diverse group of attendees is to reduce the incidence of diagnostic error in medical practice, first by understanding why errors happen and second by educating health providers on how to avoid them. I have thoroughly enjoyed this gathering the past two years.

A 3 hour seminar hosted by PIAA (formerly Physicians Insurance Association of America), produced the material of greatest interest to readers of the Medical Malpractice Bulletin. The session was titled The PIAA Medical Professional Liability Perspective: Improving Diagnosis and Patient Outcomes. Presenters and panelists included:

  • Pat Kischak RN DPHRM, VP Risk Management and CNO Hospitals Insurance Company
  • Gordon Wallace MD, Director of Education and Managing Director of Safe Medical Care, Canadian Medical Protective Association
  • Alan Lembitz MD, Chief Medical Officer, CRICO
  • Dana Welle DO JD, Chief Medical Executive, The Risk Authority, Stanford
  • Anne Whitehead JD RN, VP of Risk Management & Patient Safety, Cooperative of American Physicians
  • David Troxel MD, Medical Directore, The Doctors Company
  • Divya Parikh MPH, VP of Research and Risk Management, PIAA

Here are a few of the pearls readers might find interesting:

  1. Cancer is the #1 diagnosis leading to a malpractice claim.
  2. Diagnostic error (misdiagnosis, delayed diagnosis or failure to diagnose) is the #1 mistake.
  3. Radiology is the #1 specialty involved.
  4. 25% of claims in the IOM database are related to diagnosis.
  5. 34% of non-surgical claims are diagnosis related, the #1 cause of claims. Internal medicine is the most common specialty with 40% of their clams due to diagnostic error, and all of the top 5 diagnoses are common. Hospital Medicine, Family Practice, Cardiology, Emergency Medicine and Pediatrics are frequently involved in cases of diagnostic error.
  6. For surgical specialties, diagnosis-related error ranks #3 at 14%, usually failure to recognize a complication.
  7. Most diagnostic errors occur during times of rapid decision making, not over longer periods of time when reconsideration is possible. For example, the window of opportunity to make a diagnosis and assure a successful outcome has a different trajectory for an MI than it does for cancer.
  8. The top reasons for diagnostic error are:
    • inadequate differential Dx (not thinking broadly enough) impaired data synthesis (having the data but misinterpreting it)
    • failure to order tests (when the presentation indicates the need). Note that “taking a history” remains the best “test,” the example used being missed aortic dissection. Sir William Osler’s advice from over 100 years ago remains sound: “Listen to your patient, he is telling you the diagnosis.”
    • anchoring bias (e.g., assuming a pre-existing condition is the cause of the current presentation)
    • context errors (e.g., missing a lung nodule on a chest x-ray obtained for trauma)
    • failure to follow diagnostic protocols (e.g., failing to get a CT scan on a patient with minor head trauma who is on Coumadin)
    • human factors (fatigue, etc.)
    • system factors (e.g., MRI machine too small for obese patient)
    • in contrast, faulty knowledge is rarely the cause of diagnostic error
  9. Because no standard has been set for EHR’s, they are not inter-operable and do not allow the free flow of information among providers. When individual vendors drive operability, diagnostic error results.
  10. In addition to education and raising awareness, improving clinical decision support in our IT systems has the most promise to reduce diagnostic error.
  11. EHR’s have significantly increased the “signal to noise ratio.” For now, medical records are primarily billing documents rather than ways to communicate information among care providers.
  12. “Click-tation” has replaced communication, resulting in a “pixelated medical record”; the record must show the “intellectual footprint” of providers, i.e., demonstrate “medical decision making.”
  13. The Institute of Medicine’s (IOM) Goal #6 to reduce diagnostic error is “Learn from diagnostic errors & near misses.” This goal is aimed at malpractice insurance companies. The Doctors Company already uses closed claims (anonymized, no $ amounts) to educate specialty societies and requires the data to be published in society journals, presented at national meetings and/or posted on the specialty’s website.
  14. Communications and Resolution Programs (CRP) may reduce incidents of diagnostic error by bringing transparency to such mistakes. These should be done in a way that does not “blame and shame” and include a commitment to “Just Culture” principles.
  15. Faulty communications systems lead to diagnostic errors, which could be reduced through improved coordination during handoffs and transitions of care, improving protocols for telephone triage, follow-up of diagnostic tests and clearly defining staff roles and responsibilities.
  16. Canadian physicians are sued 1/5 as often as American physicians on a per capita basis. After screening 54% of cases are dismissed, 38% settle and 9% go to trial, where doctors prevail 80% of the time. Diagnostic error is Canada’s #1 cause of claims and is present in 84% of cases.

I will most likely forego the SIDM DEM Conference next year in Boston October 8-10 and instead attend the national conference of the American Society of Healthcare Risk Management in Seattle October 15-17, 2017. Perhaps I will see you there.

{ 1 comment… read it below or add one }

Patrick Croskerry January 2, 2017 at 6:06 AM

Charles: thanks for this, and for the cases that you send.
I am planning on publishing a book this year that will have ‘cognitive autopsies’ on a number of cases from the ED.
Like you, I believe we learn a lot from our mistakes. The main problem is that there is usually no tangible proof of cognitive error – it has to be inferred from events. Nevertheless, students usually find them interesting and informative.
Besides the selected cases I have (about 30) I also have 4 that I have autopsied in depth from the legal documentation.
I would like to build this to about 10 – do you have any suggested examples that could be included? The only requirement is that I have sufficient detail to make the analysis.

Leave a Comment