Background A cumulative quantitative score is implemented for classification criteria in various rheumatic diseases, including rheumatoid arthritis (RA)1 and systemic lupus erythematosus (SLE).2 Previous reports indicate that self-report patient scores on a multidimensional health assessment questionnaire (MDHAQ) provide clues to identify patients with fibromyalgia (FM).3,4 A cumulative score of quantitative patient self-report scores as well as a physician estimate of a level of FM might be useful to identify patients with FM.
Objectives To analyze a simple 3-variable cumulative score to identify patients with FM, and distinguish them from patients with RA, SLE, and ankylosing spondylitis (AS) seen in a usual care setting in Korea.
Methods All patients seen at a Korean rheumatology setting complete an MDHAQ/RAPID3 (routine assessment of patient index data), which includes scores for physical function, visual analog scales (VAS) for pain, global status and fatigue, and a list of 60 symptoms as a checklist review of symptoms. Physicians also complete a RHEUMDOC form which includes a 0-10 VAS physician global estimate of patient status, and 0-10 VAS subscales for inflammation, damage, and level of symptoms not explained by inflammation and damage, such as FM. Preliminary cross tabulations were performed to compare scores for MDHAQ and RHEUMDOC variables in patients with FM versus RA, SLE, and AS. Based on these preliminary analyses, a 3-variable cumulative score was developed: 1) MDHAQ pain VAS score of ≥5 out of 10, 2) MDHAQ symptom checklist of ≥15 out of 60 symptoms, and 3) RHEUMDOC physician estimate of FM level ≥3 out of 10. The numbers of patients with FM, RA, SLE and AS who scored 0, 1, 2, or 3 were computed; chi square tests and receiver operator curves (ROC) were used to analyze statistical significance.
Results Data were available concerning 32 patients with FM, 18 with AS, 17 with SLE and 277 with RA. Overall, 53% of patients with FM had cumulative scores of 2 or 3 of the 3 criteria, compared to 5.5% with AS, 20.2% with RA and 5.8% with SLE (p≤0.001). ROC area was 0.734, with standard error of 0.047 and 95% confidence interval of 0.642-0.826. A score of 2, which does not require any physician data, was associated with a positive likelihood ratio for FM of 2.9, sensitivity of 53.1 and specificity of 81, with 79% correctly classified. A score of 3 was associated with a positive likelihood ratio for FM of 7.1, with 90% correctly classified.
Conclusions A cumulative score analogous to classification criteria for RA or SLE may be useful in helping to identify FM. A limitation of this study is that it did not identify patients with a diagnosis of RA, SLE,or AS who might also have FM, who may account for patients with inflammatory rheumatic diseases whose scores were 2 or 3 on the preliminary FM index.
Ann Rheum Dis 2010;69:1580-88.
Arthritis Rheum 2012;64:2677–86.
Arthritis Rheum 1990;33:1317-22.
Clin Exp Rheumatol 2004;22:453-61.
Disclosure of Interest None declared