A new set of criteria for the diagnosis of familial Mediterranean fever in childhood.

Objectives: Several sets of criteria mainly for adults have been proposed for the diagnosis of FMF. The aim of the present study is to validate the most widely used diagnostic 'Tel Hashomer' criteria in children and to establish a new set of criteria for use in childhood.

Methods: The study group consisted of 170 recently diagnosed FMF patients who had mutations at both alleles. They were interviewed about the presence of 35 features and manifestations of FMF at the time of diagnosis. Controls were consecutive patients without FMF (n = 141) who had episodes of fever and clinical features mimicking that of FMF. The diagnostic performance of the candidate features was assessed by multiple logistic regression analysis.

Results: The sensitivity and specificity of Tel Hashomer criteria in our study group were 98.8 and 54.6%, respectively. The multiple logistic regression analysis showed that 5 (fever, abdominal pain, chest pain, arthritis and family history of FMF) of the 35 candidate criteria discriminate FMF from controls with a sensitivity and specificity of 88.8 and 92.2%, respectively. The presence of two or more of these five criteria diagnosed FMF with a sensitivity of 86.5% and a specificity of 93.6%.

Conclusion: It was demonstrated that although the Tel Hashomer criteria were successful in diagnosing the FMF patients in childhood, its specificity was definitely low in children. The new set of criteria has a high sensitivity and specificity for the diagnosis of FMF and is practical to use on an everyday basis.