Objective To review the scientific evidence for routine postnatal testing of thyroid function in different groups of newborns within the framework of an effective universal screening for congenital hypothyroidism, and analyse the clinical effectiveness of the local laboratory services for thyroid function testing (TFT).

Method Review of literature from Medline, Cochrane library and CINHAL. Analysis of all TFT on children between birth and 3 months of age at two local hospitals in South-West of England (Royal Devon and Exeter, a large teaching hospital, and Torbay District Hospital) over a three-year period (2005 to 2008).

Results Literature review and local clinical audits in many UK centres suggest that too many laboratory tests of thyroid functions are unnecessarily performed in the newborn and young infants. For example, routine repeated screening in premature babies could be avoided by delayed sampling and by setting a lower threshold of TSH level for re-screening. Routine testing of thyroid function in infants of mothers with Hashimoto’s disease and babies with prolonged jaundice has very insignificant yield and has been abandoned in many centres. Routine screening of babies born to mothers with hyperthyroidism can be replaced with a more selective testing based on clinical surveillance and maternal levels of stimulating TSH receptor antibodies.

2–5% of all TFT (total 639) performed locally in the two hospitals in babies up to 3 months of age showed physiological derangements (high TSH with normal or high T4) which often necessitated multiple repeat testing. There was no confirmed diagnosis of hyperthyroidism.