Criteria for the diagnosis of fibromyalgia: validation of the modified 2010 preliminary American College of Rheumatology criteria and the development of alternative criteria.

OBJECTIVE

To validate the 2011 modification of the 2010 American College of Rheumatology (ACR) preliminary criteria for the diagnosis of fibromyalgia (2011ModCr) and develop alternative criteria in a sample of patients with diverse pain disorders that are commonly seen in everyday practice by pain specialists, rheumatologists, and psychologists.

METHODS

Eight clinicians from geographically varied locations in the US evaluated patients with chronic pain and psychiatric disorders using a standard set of questions that included the 2011ModCr questions, the Symptom Impact Questionnaire (SIQR), a 28-area pain location inventory (PLI), and the Short Form 36. Alternative diagnostic criteria were developed from the same data set using logistic regression and receiver operating curve analysis.

RESULTS

Complete data on 321 patients were evaluated; there were 135 patients with fibromyalgia (according to the 1990 ACR criteria) and 186 patients with 16 other common chronic pain problems. Comparing the 2011ModCr with the 1990 ACR criteria provided a sensitivity of 83%, a specificity of 67%, and a correct classification of 74%. Alternative criteria were derived from the 10-item symptom score from the SIQR symptoms and the 28-area PLI. Maximal diagnostic accuracy was obtained with ≥17 pain sites (range 0-28) and an SIQR symptom score of ≥21 (range 0-50). These alternative criteria had a diagnostic sensitivity of 81%, a specificity of 80%, and a correct classification of 80%.

CONCLUSION

The 2011ModCr had robust operating characteristics. Alternative criteria based on symptom items from the SIQR and pain locations from the PLI had comparable operating characteristics, with somewhat better specificity and ease of use.