BackgroundThe aim of this study was to estimate the prevalence of haploinsufficiency of short stature homeobox containing gene (SHOX) deficiency (SHOXD) in a population of short-statured children, and to analyze their phenotype and the performance of clinical scores.MethodsScreening for SHOXD was performed in 281 children with short stature by direct sequencing and multiplex ligation probe-dependent amplification. Subjects with SHOXD were compared with 117 matched short patients without SHOXD. We calculated the cutoff of growth velocity associated with the highest sensitivity and specificity as a screening test for SHOXD by receiver operating characteristic curves.ResultsThe prevalence of SHOXD was 6.8%. Subjects with SHOXD showed a lower growth velocity (P<0.05) and a higher prevalence of dysmorphic signs. The best cutoff for growth velocity was -1.5 standard deviation score (SDS) both in the whole population and in subjects with a Rappold score <7 and <4 points. Growth velocity was ≤-1.5 SDS or Rappold score was >7/>4 points in 17/17 of 19 children with SHOXD and in 49/65 of 117 subjects without SHOX mutations.ConclusionsGrowth rate ≤-1.5 SDS, even with negative Rappold score, may be useful to detect precociously children with SHOXD. Selecting children deserving the genetic test by using growth velocity or the Rappold score significantly increases the sensitivity in detecting mutations and decreases the specificity.