January 1987

Clinical Signs and Laboratory Tests in the Differential Diagnosis of Arthritis in Children

Author Affiliations

From the Children's Hospital (Drs Kunnamo and Pelkonen), and Department of Virology (Dr Hovi), University of Helsinki, and Aurora Hospital, Helsinki (Dr Kallio).

Am J Dis Child. 1987;141(1):34-40. doi:10.1001/archpedi.1987.04460010034018

• To develop a scheme for primary diagnosis, we analyzed the clinical findings and laboratory test results in 278 children with arthritis by using univariate analysis and multivariate logistic regression analysis. An elevated C-reactive protein (CRP) value, a temperature above 38.5°C, and a high white blood cell count were independent predictors for the diagnosis of septic joint infection in patients with acute monoarthritis. The presence of either of the first two signs had a sensitivity of 100% and a specificity of 87% for septic arthritis. Sixtyseven percent of all patients with arthritis were cured within two weeks from the onset of joint symptoms. In patients whose disease duration exceeded two weeks, a low CRP value, the absence of fever, and an elevated IgG value were independent predictors for the diagnosis of juvenile arthritis. Antinuclear antibodies had a specificity of 100% and a sensitivity of 25% for juvenile arthritis or other connective tissue diseases. We recommend that laboratory tests indicated for all children with joint symptoms include determinations of the erythrocyte sedimentation rate and the CRP value, both total and differential leukocyte counts, urinalysis, and a bacterial culture of a throat smear. When arthritis is prolonged or when enteroarthritis is suspected, tests for antinuclear antibodies and serum immunoglobulins, serologic tests for Yersinia and Salmonella, and stool bacterial cultures should be included.

(AJDC 1987;141:34-40)