Consistency between the 2016 ACR criteria and a previous diagnosis or hypothesis of fibromyalgia in a specialised referral clinic

Abstract
Objectives: Fibromyalgia (FM) is a complex syndrome whose hallmark features are chronic widespread pain, sleep disturbances, fatigue and cognitive dysfunctions. However, it is still difficult to apply validated diagnostic criteria. The aim of this study is to examine the accuracy of a previous diagnosis/diagnostic hypothesis of FM according to the 2016 ACR diagnostic criteria.

Methods: All of the patients newly referred to a private rheumatological clinic with the specific request for a consultation because if FM over an 18-month period were evaluated by means of a standardised protocol in order to determine whether they fulfilled the 2016 ACR diagnostic criteria for FM. They were initially divided into three groups: those with a previous diagnosis of FM (group 1), those with a physician's diagnostic hypothesis of FM (group 2) and those who personally hypothesised FM (group 3). They were subsequently classified as having FM, IFM (borderline scores) or not having FM (non-FM) on the basis of the 2016 ACR diagnostic criteria.

Results: The study involved 216 patients (25 males and 191 females): 112 in group 1, 49 in group 2, and 55 in group 3. Only 89 patients (41.2%) fulfilled the ACR criteria; 42 (19.44%) met the study protocol-defined scores for IFM; and 85 (39.35%) were diagnosed as not having FM. Only 50% of the patients with a previous diagnosis of FM fulfilled the ACR criteria, and just under 25% did not have FM. Almost 50% of the patients with a physician's diagnostic hypotheses of FM did not have FM, whereas 20% of the patients who personally hypothesised FM fulfilled the ACR criteria. GP scores and TPCs were significantly different (FM > IFM, FM > non-FM, and IFM > non-FM) as were WPI, SSS and PSD scores for FM > IFM group. Rheumatologists made the previous diagnosis in 92.85% of patients, 53.84% of whom met the ACR criteria and about 20% did not have FM; and as many as 37.5% of the patients with a previous diagnosis made by a non-rheumatologist did not have FM. The non-FM patients were given 84 alternative diagnoses, 78.5% of which referred to rheumatic diseases. One hundred and thirty-one patients had 86 closely pain-related co-morbidities, 94.1% of which were rheumatic diseases.

Conclusions: Our findings confirm the inaccuracy of FM diagnoses and highlights the possibility that in everyday clinical practice, they are not always made with reference to very specific criteria and that there is a high risk of classifying non-FM patients as having FM. They also underline the importance of an accurate differential diagnosis. Separately classifying as IFM those patients who do not meet the ACR criteria, but have clinical findings indicating FM, may help to prevent their exclusion from specific treatment(s).