Reliability, Factor Structure and Predictive Validity of the Widespread Pain Index and Symptom Severity Scales of the 2010 American College of Rheumatology Criteria of Fibromyalgia

Abstract

Fibromyalgia syndrome (FMS) is a chronic condition of widespread pain. In 2010, the American College of Rheumatology (ACR) proposed new diagnostic criteria for FMS based on two scales: the Widespread Pain Index (WPI) and Symptoms Severity (SS) scale. This study evaluated the reliability, factor structure and predictive validity of WPI and SS. In total, 102 women with FMS and 68 women with rheumatoid arthritis (RA) completed the WPI, SS, McGill Pain Questionnaire, Trait Anxiety Inventory, Fatigue Severity Scale, Oviedo Quality of Sleep Questionnaire, and Beck Depression Inventory. Pain threshold and tolerance and a measure of central sensitization to pain were obtained by pressure algometry. Values on WPI and SS showed negative-skewed frequency distributions in FMS patients, with most of the observations concentrated at the upper end of the scale. Factor analysis did not reveal single-factor models for either scale; instead, the WPI was composed of nine pain-localization factors and the SS of four factors. The Cronbach’s α (i.e., Internal consistency) was 0.34 for the WPI,0.83 for the SS and 0.82 for the combination of WPI and SS. Scores on both scales correlated positively with measures of clinical pain, fatigue, insomnia, depression, and anxiety but were unrelated to pain threshold and tolerance or central pain sensitization. The 2010 ACR criteria showed 100% sensitivity and 81% specificity in the discrimination between FMS and RA patients, where discrimination was better for WPI than SS. In conclusion, despite their limited reliability, both scales allow for highly accurate identification and differentiation of FMS patients. The inclusion of more painful areas in the WPI and of additional symptoms in the SS may reduce ceiling effects and improve the discrimination between patients differing in disease severity. In addition, the use of higher cut-off values on both scales may increase the diagnostic specificity in Spanish samples.