Improving Diagnostic Accuracy in Medicine

1. What is diagnostic error?

  • Diagnosis that is wrong, missed or delayed (Graber 2005)

2. What do we know about diagnostic error?

  • It’s frequent and harmful
  • It’s under-recognized, under-studied and not integrated into quality assurance measures or activities (Newman-Toker and Pronovost 2010)

3. How often does it happen?

  • Approximately 5-15% of the time (Berner and Graber 2008 and other sources)

4. Where do we get information about diagnostic error?

  • Autopsy data
  • Physician self -reports of experiencing diagnostic error
  • Patient self-reports of experiencing diagnostic error
  • Databases of reported error
  • Peer reviewed journal studies
  • Medical malpractice claims data (Berner and Graber 2008)

5. What about malpractice claims data, isn’t it skewed?

  • If other data sources tend to under-represent diagnostic error and claims data tends to over-represent it (Graber and Berner 2008), then claims may serve as a balancing and confirming adjunct source
  • Claims analysis reviews the patient viewpoint in addition to the medical record, a viewpoint absent from most other data sources
  • There’s abundant claims data available on diagnostic error (NPDB, PIAA Data Sharing Report)
  • There are many peer reviewed claims studies in the literature (Griffen 2008, Studdert and Mello 2006, and others)
  • Claims are strongly associated with adverse events (Rand Corp. study 2010) such that if one is reduced or increased, so goes the other.

6. How often does diagnostic error lead to adverse events and death? How often is death due to diagnostic error?

  • Many errors are unknown (low rate of autopsy, patient going elsewhere for care) therefore exact relationship not known
  • Harvard Study showed diagnostic error accounting for 17% of adverse events (Leape, Brennan 1991)
  • In malpractice claims involving a death, diagnostic error is far and away the top allegation at 26% (Physician Insurer, PIAA 2010)

7. Isn’t diagnostic error more associated with hospital care than in the physician’s office?

  • No. Diagnostic error was the #1 cause of claims in ambulatory care and #2 in hospitals (after improper performance of a procedure), but the totals were fairly close (Bishop et al JAMA 2011, based on NPDB data)

8. What is the cause of diagnostic error?

  • It’s multi-factorial and can present as a perfect storm of multiple factors lining up: 6 factors on average were found per case of diagnostic error in an internal medicine study (Graber 2005).
  • Lack of physician knowledge is least often the problem. It is more often due to cognitive error, systems errors including communication errors, and most common of all, the combination of cognitive and systems errors (Graber 2005).

9. What is the difference between cognitive and systems errors?

  • Cognitive errors take place in the physician’s head— they concern the thinking process. An example is latching on prematurely to a diagnosis and abandoning the search for evidence to the contrary.
  • Systems errors occur between the inter-related pieces in healthcare systems. Examples include physicians or other practitioners dropping the ball in the referral-consultation process or in the hand-off process. Another is lost or unreported test results.
  • In combination, it is often the systems error that is the catalyst for adverse events and claims because 1) by definition a systems error tends to replicate and amplify, and 2) the patient perceives it as the error that didn’t have to happen, and 3) if the systems factor is peeled away, the perfect storm may not transpire

10. Is diagnostic error more closely associated with some specialties than others?

  • Yes, but it is a significant factor for claims in all specialties.
  • It is the #1 cause of medical misadventure claims for all the primary care specialties (internal medicine, family and general practice, pediatrics), radiology and emergency medicine, and most of the medical sub-specialties.
  • It is the #2 cause of medical misadventure claims for the surgical specialties (OB-gyn, general surgery, orthopedics and most of the surgical sub-specialties), but it is most often a close, not a distant second place.

11. Is it the rare diagnosis that is the subject of diagnostic error?

  • No, it is the common diagnosis and the common killers: heart attack, cancer and stroke.
  • Overall, the top diagnosis in claims related to diagnostic error is breast cancer (PIAA Data Sharing Report 1985-2009).
  • Acute myocardial infarction is the top subject of diagnostic error in claims for the specialties of adult primary care, emergency medicine and cardiology (PIAA Data Sharing Report 1985-2009).
  • Stroke is associated with diagnostic error 9% of the time (Newman-Toker et al 2008).
  • For family and general practice, the top diagnoses involved in diagnostic error in descending order were myocardial infarction, breast cancer, appendicitis, colorectal cancer and lung cancer.
  • In a study of physician self-reported diagnostic errors, the diagnoses most often involved were pulmonary embolism, drug reaction or overdose, lung cancer, colorectal cancer, acute coronary syndrome, breast cancer and stroke (Schiff et al 2009).
  • Certain diagnoses like pulmonary embolism and aortic dissection may not be found until autopsy, but the rate of autopsies performed in the US has declined steeply, so these and others are under-detected at an unknown rate.

12. What can be done to reduce diagnostic error and harm?

  • Raise awareness of diagnostic error and its importance among physicians, the public, healthcare organizations and funding sources
  • Increase funding and attention to research into causes and remedies for diagnostic error
  • Develop the databases and guidelines to improve clinical decision-making tools
  • Integrate teaching about cognitive error and diagnostic error into medical school curricula, resident training, and continuing education
  • Include goals related to improving diagnostic accuracy in quality improvement/quality assurance activities and measures
  • Invest in and make available the following things to the physician at the point of care: clinical decision making support systems, electronic medical records, integration of physician office-hospital-lab-imaging facility records, and other aids to improve diagnostic accuracy
  • Develop the means to measure diagnostic error and provide report cards to practitioners
  • Develop a corps of physician-champions dedicated to improving accuracy in medical diagnosis, make international connections to spread this initiative internationally


Berner ES, Graber ML. Overconfidence as a cause of diagnostic error in medicine. AmJ Med. 2008;121:S2–S23.

Bishop TF, Ryan AK, Casalino LP. Paid malpractice claims for adverse events in inpatient and outpatient settings. JAMA. 2011:305;2427-2431.

Data Sharing Project 1985-2009 and Special Reports. Physician Insurance Association of America (PIAA), Rockville, MD. For more information

Graber ML. Diagnostic errors in medicine: a case of neglect. Jt Comm J Qual Patient Saf. 2005;31:106-113.

Graber ML, Franklin N, Gordon R. Diagnostic error in internal medicine. Arch Intern Med. 2005;165:1493-1499. Free full text

Greenberg, MD, Haviland AM, Ashwood SJ, Main R. Is Better Patient Safety Associated with Less Malpractice Activity? Evidence from California. Santa Monica, CA: Rand Corporation, 2010. Free full text 

FD, Stephens LS, Alexander JB, et al. Violations of behavioral practices revealed in closed claims reviews. Ann Surg. 2008;248:468-474.

Leape L, Brennan TA, Laird N, et al. The nature of adverse events in hospitalized patients: results of the Harvard Medical Practice Study II. N Engl J Med.1991;324:377-384. Free full text

Newman-Toker DE, Pronovost PJ. Diagnostic errors—the next frontier for patient safetyJAMA. 2009; 301:1060-1062.

Newman-Toker DE, Robinson KA, Edlow JA. Frontline diagnosis of cerebrovascular events in the era of modern neuroimaging: a systematic review. Ann Neurol. 2008:64(suppl 12):S17-18.

PIAA Data Sharing Project, Top alleged medical errors in claims where the patient expired (DSP data 1985-2009). Physician Insurer Fourth Quarter 2010: 55.

Schiff GD, Hasan O, Kim S, Abrams R et al. Diagnostic error in medicine, analysis of 583 physician-reported errors. Arch Intern Med. 2009;169:1881-1887. Free full text

Studdert DM, Mello MM, Gawande AA, et al. Claims, errors, and compensation payments in medical malpractice claims. N Engl J Med. 2006;354;2024-2033. Free full text