From the Journals

Criteria-based fibromyalgia diagnosis and rheumatologists’ clinical diagnosis often disagree

View on the News

Classification criteria should never replace clinical acumen

Despite voicing concern regarding the “self-report nature of fibromyalgia,” Wolfe et al. selected the 2011 self-report fibromyalgia criteria as their “gold standard” for fibromyalgia diagnosis. Their findings therefore should not be surprising.

The authors’ conclusion that expert physicians often misdiagnose fibromyalgia implies that published criteria are superior to expert clinical judgment for individual patient diagnosis.

However, this view fails to take into account the myriad variables seen by the clinician in the clinic. Until biomarkers or genetic testing allows for more objective disease markers, common conditions like fibromyalgia will continue to be symptom-based diagnoses. Therefore, the gold standard diagnostic criteria for fibromyalgia and other common illnesses is expert opinion.

Rheumatologists are the go-to experts for the diagnosis of fibromyalgia, whether or not we readily accept that role.

These comments are adapted from an accompanying editorial by Don Goldenberg, MD, of Oregon Health & Science University, Portland, and Tufts University, Boston (Arthritis Care Res. 2019 Feb 6. doi: 10.1002/acr.23727).



New research has revealed considerable disparity between rheumatologists’ clinical diagnosis of fibromyalgia and fibromyalgia criteria-based diagnosis, particularly the most recent revision of criteria from 2011 that rely on patients’ self-report of symptoms.

Dr. Frederick Wolfe of the National Data Bank for Rheumatic Diseases and the University of Kansas in Wichita

Dr. Frederick Wolfe

Frederick Wolfe, MD, of the National Data Bank for Rheumatic Diseases and the University of Kansas, Wichita, and his colleagues reported in Arthritis Care & Research that clinicians failed to identify almost half of cases that met the 2011 modification of the American College of Rheumatology’s self-report criteria for fibromyalgia during a 3-month period at Rush Medical College, Chicago. This led them to conclude that the “overall agreement between clinicians’ diagnosis of fibromyalgia and diagnosis by fibromyalgia criteria is only fair.”

The findings call into question studies of fibromyalgia based on ICD-10 diagnosis, according to the authors.

Several widely accepted criteria sets have provided definitions and methods of diagnosis, but a number of studies have suggested that fibromyalgia is overdiagnosed, the authors wrote.

“In the current study, we consider underdiagnosis as well as the form of overdiagnosis that lead to diagnosis when fibromyalgia criteria are not satisfied and symptoms are insufficient for diagnosis,” they added.

The research included 497 consecutive patients attending the Rush Medical College rheumatology clinic who completed a questionnaire assessing fibromyalgia diagnostic variables used in the ACR 2010 preliminary diagnostic criteria for fibromyalgia and its 2011 modification for self-report immediately prior to being seen at the clinic. Clinicians were not given instructions on how any diseases were to be diagnosed, including no instructions on using 2011 fibromyalgia criteria.

The results showed that 121 (24.3%) of the patients satisfied the 2011 fibromyalgia criteria. The agreement between clinicians and criteria was 79.2%, but agreement beyond chance was considered “fair” based on a kappa score of 0.41 and a probabilistic benchmark interval of 0.21-0.40. The researchers wrote that this benchmark represents “the probability for each coefficient of falling into the selected benchmark interval along with the cumulative probability of exceeding the predetermined threshold associated with the interval.”

A total of 104 (20.9%) received a clinician ICD-10 diagnosis of fibromyalgia, but only 61 (58.7%) actually satisfied criteria. Physicians failed to identify 60 (49.6%) criteria-positive patients and incorrectly identified 43 (11.4%) criteria-negative patients.

In a subset of 88 patients with RA, agreement was 84.1%, and the kappa score was 0.32, indicating “slight to fair” agreement (probabilistic benchmark interval, 0.00-0.20). Among 13 RA patients with criteria-positive fibromyalgia, 5 were identified by clinicians; among those who were criteria negative, 6 were deemed positive by clinicians.

The authors noted than “even worse” results were obtained in patients with systemic lupus erythematosus, where the agreement between criteria and clinician ICD diagnosis yielded a kappa value of 0.08. In patients with osteoarthritis, the kappa score was 0.51 (probabilistic benchmark interval, 0.21-0.40).

Overall, the results showed that women and patients with more symptoms but fewer pain areas were more likely to receive a clinician’s diagnosis than to satisfy fibromyalgia criteria. “Clinicians gave greater weight in making a diagnosis to being a woman and having increased symptoms and were willing to diagnose patients with lower WPI [Widespread Pain Index] and PSD [polysymptomatic distress] scores,” the researchers noted.

Dr. Wolfe and his associates wrote that it is unclear why the physicians in the study who misclassified fibromyalgia and missed patients with the diagnosis did not do better. “They may have simply misdiagnosed the condition or not accepted the use of the formal diagnostic system as necessary or clinically useful for identifying fibromyalgia in routine care.”

The researchers noted that it is likely that diagnosis in the community by general physicians, for whom the criteria are not as well known, is even more inaccurate. “It is likely that misdiagnosis is a public health problem and one that can lead to overdiagnosis and overtreatment as well as to inappropriate treatment of individuals not recognized to have fibromyalgia symptoms.”

No disclosures or funding were reported.

SOURCE: Wolfe F et al. Arthritis Care Res. 2019 Feb 6. doi: 10.1002/acr.23731.

Next Article: