SEARCH

SEARCH BY CITATION

Thumbnail image of

Cancer specialists participating in a national survey on misdiagnosis that was released this past January agreed that steps should be taken to improve the accuracy of diagnosis, ranging from incentives for the improved confidential gathering and reporting of data on misdiagnosis to increased funding for the study of misdiagnosis.[1]

“We have a database of cancer specialists with reputations as being among the best in the country, and we were interested in knowing what they think about the quality of diagnosis in cancer care,” says Evan Falchuk, vice chairman of Best Doctors Inc., a Boston, Massachusetts-based medical research firm that co-sponsored the survey with the National Coalition on Health Care (NCHC). The survey was conducted among 400 cancer specialists nationwide and audited and certified by Gallup. The goal was to determine the most significant barriers to accurately diagnosing and characterizing cancers, and to identify which tools and technology physicians believe will help them better diagnose the disease.

Misdiagnosis More Frequent Than Believed

  1. Top of page
  2. Misdiagnosis More Frequent Than Believed
  3. Need for Well-Trained Pathologists
  4. Multidisciplinary Communication
  5. References

Survey respondents said the number 1 reason misdiagnoses occur is due to missing or fragmented information, says Falchuk. “One interesting point is the extent to which physicians underestimate the extent to which issues of diagnostic accuracy occur.” A fairly high majority of respondents (60.5%) estimated that misdiagnoses or incomplete characterizations occur 0% to 10% of the time, whereas another 33% estimated that they occur 10% to 20% of the time.

Published studies, however, indicate that incidence of misdiagnoses ranges from 28% in general to 44% for certain types of cancer. Falchuk cites 2 studies–1 published by the Journal of Clinical Oncology in 2007, which reviewed the literature to understand the prevalence, origins, and prevention of errors in cancer; and another study by the University of Michigan that was published in Cancer in 2006 and reviewed breast cancer imaging studies, resultant changes in interpretations, and, ultimately, changes in the surgical management of those cases.[2, 3]

In the latter study, a retrospective review of 149 patients referred to the University of Michigan's multidisciplinary breast center demonstrated that a review of imaging studies led to a change in interpretation in 67 of the patients (45%), whereas a review of the pathology led to changes in interpretation for 43 patients (29%). Both contributed to changes in recommendations for the surgical management of the patients.

Falchuk points to poor “feedback loops” that do not enable a good mechanism for reporting back to the original physician about a change in the original diagnosis. “That's a huge missed opportunity to improve quality of care,” he says.

He also notes that some countries are ahead of the United States in this area. In Australia, for example, a government agency known as the Clinical Excellence Commission is putting such feedback loops into practice. In Massachusetts, meanwhile, where Best Doctors is headquartered, legislation was introduced recently to create a task force on diagnostic accuracy. “A lot of the cases that we get are cancer, and it's a wonderfully simple analogy of what's happening throughout the health care system,” Falchuk says. “All the issues of not enough time, fragmented information, and the skill of the pathologist are magnified in the context of cancer.”

Need for Well-Trained Pathologists

  1. Top of page
  2. Misdiagnosis More Frequent Than Believed
  3. Need for Well-Trained Pathologists
  4. Multidisciplinary Communication
  5. References

Asked to weigh in on the topic of misdiagnosis, Robert Mayer, MD, vice chair for academic affairs in the department of medical oncology at the Dana-Farber Cancer Institute in Boston, Massachusetts, and former president of the American Society of Clinical Oncology, says the 2 cancers that are most commonly linked to changes in diagnoses are lymphoma and breast cancer. Because of the various subtypes of these diseases and the need for highly specialized pathologists, his institution mandates that pathology and scans for these and other cancers are thoroughly reviewed for patients coming from other centers or entering clinical trials.

Specialists at smaller community cancer centers may not deal with more complicated diagnoses on a routine basis nor possess some of the more sophisticated technology. For these reasons, Dr. Mayer believes physicians at these facilities should take time to send the slides and tissue blocks to a major facility to have them confirm or refute the diagnosis. “When there is a gray area, that is a quality measure, and it provides much better security and safety for the patient,” he says.

Because specific subtypes will determine how certain cancers are managed, the work of diagnosis is much more complicated and requires physicians to set a much higher standard for the type of information they need, says Dr. Mayer. In colon cancer, for example, research has demonstrated that 11 to 12 lymph nodes must be removed to determine prognosis and treatment, a requirement now part of the facility accreditation process.

Dr. Mayer emphasized the importance of well-trained pathologists, which was echoed by physicians in the Best Doctors survey, who believed the leading cause in most errors was due to interpretations of pathology specimens. Some 47% of respondents attributed errors to “the pathologist's lack of subspecialty expertise.” At the same time, 36% called for new or improved pathology tools or resources in improving accuracy rates in cancer. “As pathologists are trained in these new technologies, they should take courses and become recertified so that we can all feel much more secure that changing diagnoses in the future will be a little less common,” Dr. Mayer says.

Breast cancer specialist Leon Dragon, MD, medical director of the Kellogg Cancer Center in Highland Park, Illinois, and assistant professor of medicine at the University of Chicago, also notes that determining cancer subtypes is not straightforward. “One specialist might see HER2/neu-negative while another at a different institution might interpret it as positive,” he says. “Does that reflect state-of-the-art variations in how these tests can be done? There are alterations inherent in the technology.”

Other variations in diagnoses can range from different interpretations of specific grades of cancer, different determinations of neuroendocrine versus adenocarcinoma, and even different conclusions as to whether a lesion is cancerous or not. “Very highly qualified, experienced pathologists may have differences of opinion,” Dr. Dragon says, noting that guidelines for performing special studies may vary among institutions. Different institutions may also use different reagents or types of instruments to perform the studies.

At the same time, he adds, some of the variations are simply inherent in the practice of medicine. “Medicine is part science and part art, and the art part varies from person to person,” he says. “These variations in medicine take time to resolve.”

Multidisciplinary Communication

  1. Top of page
  2. Misdiagnosis More Frequent Than Believed
  3. Need for Well-Trained Pathologists
  4. Multidisciplinary Communication
  5. References

Differing interpretations can be very serious, particularly in examples such as the HER2/neu-positive subtype, for which a highly effective treatment exists. Generally, institutions that handle higher volumes of specific cancers (such as large comprehensive cancer centers) are more likely to have uniform outcomes, Dr. Dragon says. “But, even at the best of institutions, you sometimes need to get help on more complicated cases,” he adds. “We have had troublesome cases where we will ask for an opinion from a specialist at another institution.”

In a further effort to improve standardization of interpreting results, the American College of Pathologists has established guidelines and procedures for many of the special studies that are now used, Dr. Dragon says.

In addition to improving standardization of technology, communication between the clinician and other specialists such as pathologists and radiologists is vitally important, and that means ensuring that the medical record has sufficient information and is easily accessible. However, guaranteeing those efforts at a time when reimbursement and resources are diminishing is a challenge, Dr. Dragon adds.

Thumbnail image of

Medicine is part science and part art, and the art part varies from person to person. These variations in medicine take time to resolve.

One way to foster strong communication is through multidisciplinary, site-specific reviews of each patient's case help to ensure correct diagnosis, particularly for complex problems. Although this approach is routine at National Cancer Institutedesignated cancer centers, Dr. Dragon notes that an increasing number of centers are adopting this method.

On the public policy front, John Rother, JD, president and CEO of NCHC, believes misdiagnosis should play a larger role in future national health policy and budget discussions, noting that it contributes to the nearly $700 billion wasted in the US medical system each year, according to Thomson Reuters.

One indirect, potential solution is through efforts in Congress to change the reimbursement system for physicians, moving it in the direction of pay-for-performance. Rother anticipates that legislation will likely be introduced this summer that proposes to hold down increases in fee-for-services payments and place greater emphasis on paying for value, including higher quality and better outcomes. He hopes that focus will improve the climate for the more accurate diagnoses of cancer and other diseases.

References

  1. Top of page
  2. Misdiagnosis More Frequent Than Believed
  3. Need for Well-Trained Pathologists
  4. Multidisciplinary Communication
  5. References