Background. The term spondylarthropathy is used for several different
entities including undifferentiated forms. The latter are Particularly
frequent in children and often not recognized. Two Sets of criteria,
the Amor and the European Spondylarthropathy Study Group (ESSG) criter
ia, primarily devised for adults, were evaluated in a pediatric popula
tion. Methods. Three hundred and ten consecutive patients referred to
one pediatric rheumatology center over 6 months were prospectively ass
essed, whatever the rheumatic complaint. Twenty two items were analyse
d for each patient and compared with the criteria proposed by Amor (17
items) and the ESSG (12 items). Results. The Amor criteria performed
better than the ESSG criteria. The sensitivity was 84 and 69,7%, and t
he specificity 96,1 and 92.2% for the Amor and the ESSG criteria, resp
ectively. The sensitivity for pauciarthritis was much higher in childr
en, but the specificity was lower than in adults. The sensitivity of s
pinal involvement was very low in children. The undifferentiated type
accounted for 81% of all definite spondylarthropathies. Definite and p
ossible spondylarthropathies were 17% of the total referral over the 6
months, i.e. one patient in 6. Conclusions. These two sets of criteri
a, particularly the Amor criteria, can help pediatricians to recognize
the undifferentiated type of spondylarthropathy, in the absence of sp
inal or any other characteristic manifestation.