Screening for congenital hypothyroidism: comparison of borderline screening cut-off points and the effect on the number of children treated with levothyroxine.

BACKGROUND The newborn screening programme for congenital hypothyroidism (CH) has led to the prevention of severe developmental delay associated with this condition. In the UK, thyroid-stimulating hormone (TSH) screening cut-off points have changed over time, in some instances prompted by changing methodological platforms. The use of borderline cut-off points varies throughout the country. OBJECTIVE To use discordance in cut-off points to assess the performance of the UK Newborn Screening Programme Centre (UKNSPC) definitions. METHODS Between January 2006 and December 2007, 223,658 newborn infants were screened by the Great Ormond Street Hospital (GOSH) for CH. All children with positive results and those with blood-spot TSH concentrations >6 mU/l on repeat screening were referred to GOSH. We compared the numbers of children detected and treated for CH using the GOSH cut-off points (>6 mU/l) and those of the national screening programme (>10 mU/l). Children were defined as transient CH if levothyroxine treatment had been discontinued by 3 years. RESULTS Of the children screened between January 2006 and December 2007, 167 out of 223,658 fulfilled the GOSH screening criteria; 136 of these required levothyroxine treatment, but 29 (21%) of the children treated would not have been detected by the current UKNSPC guidelines. Transient CH was found in 17/47 (36%) of the treated children detected with a cut-off point >6 mU/l. Raising the cut-off point to >10 mU/l reduced the number of children treated for transient CH to 4/18 (22%). CONCLUSION A significant number of children with true and transient CH are missed with a screening cut-off point of >10 mU/l. Our data suggests that a cut-off point of 6 mU/l is appropriate.