Correspondence (letter to the editor): Comment on the Positive Predictive Value

I enjoyed reading this informative and well structured article. Everyone involved in neonatal screening—in hospital and clinical practice and in the laboratory—would be well advised to have this review article and its practical tables ready to hand.

I do not think, however, that the positive predictive value reported by the authors on page 16 for the immunochemical 17-OH progesterone screening (3%), can be deduced from the authors’ data on the specificity (adrenogenital syndrome [AGS] 99.39%, page 16) and the confirmed cases (AGS: 216) or the incidence (1:12 771) in Table 1 on page 14; and neither can it be concluded from the cited bibliographical source (2).

The positive predictive value for AGS screening can therefore be assumed to be in the region of 1.3%, which means that about 80 (rather than 30) neonates would have to be examined in order to identify one baby with the syndrome.