Objectives: To assess the test performance of antinuclear antibody (ANA), rheumatoid factor (RF), and cyclic-citrullinated peptide (CCP) tests in children and adolescents with undiagnosed musculoskeletal (MSK) pain or joint swelling, compared with clinical diagnoses of pediatric systemic lupus erythematosus (pSLE) and juvenile idiopathic arthritis (JIA). To explore differences in test performance for accuracy modifiers including age, sex, race or ethnicity, comorbidities, and recent infections. To evaluate the impact of test results on clinical decisionmaking and clinically important outcomes such as referrals, ordering of additional tests, clinical management, and anxiety experienced by children and parents.
Data Sources: We conducted comprehensive searches in nine electronic databases. We also hand searched reference lists and conference proceedings. There were no restrictions on language, year of publication, and study design.
Review Methods: Study selection, quality assessment, data extraction, and grading the evidence were conducted independently by two reviewers. A combination of qualitative and quantitative approaches was used to synthesize the data. We calculated sensitivity (Sn) and specificity (Sp).
Results: The search identified 11,695 citations; 28 were included in the review. Only one cohort study examined the test performance of RF to diagnose JIA among children with undiagnosed MSK pain. It demonstrated an Sn of 5 percent and an Sp of 98 percent. Fifteen case-control studies did not specifically address the test performance of RF among children with MSK pain. The strength of evidence is low for both Sn and Sp. The 12 case-control studies that examined other test-disease combinations did not specifically address the prevalence of positive tests for ANA or CCP among children presenting with undiagnosed MSK pain. The strength of evidence is insufficient to determine the test performance of ANA or CCP to diagnose JIA or pSLE in children with undiagnosed MSK pain. No studies addressed children with joint swelling. There was no evidence addressing the prespecified accuracy modifiers or clinically important outcomes.
Conclusions: Most of the evidence from the 28 studies included in the review was not applicable to the population of interest as most studies examined children with known disease rather than with undiagnosed MSK pain. No studies provided a complete investigation on accuracy modifiers. No studies examined clinically important outcomes such as the impact of the test results on referrals, ordering of additional tests, patient management, and patient and parent anxiety levels. Because the Sn and Sp of these tests have yet to be verified, current evidence does not support their use as diagnostic tests for children with undiagnosed MSK pain. They have a potential application as an adjunct to a clinical assessment that suggests the presence of an inflammatory arthritis or connective tissue disease.