Outcomes of lowered newborn screening thresholds for congenital hypothyroidism

Abstract

BackgroundNewborn screening (NBS) has largely eliminated the physical and neurodevelopmental effects of untreated congenital hypothyroidism (CH). Many countries, including Australia, have progressively lowered NBS bloodspot thyroid‐stimulating hormone (b‐TSH) thresholds. The impact of these changes is still unclear.ObjectivesTo evaluate the performance of CH NBS following the reduction of b‐TSH thresholds in New South Wales (NSW) and the Australian Capital Territory (ACT), Australia, from 15 to 8 mIU/L, and to determine the clinical outcomes of cases detected by these thresholds.MethodsNBS data of 346 849 infants born in NSW/ACT, Australia from 1 November, 2016–1 March, 2020 inclusive were analysed. A clinical audit was conducted on infants with a preliminary diagnosis of CH born between 1 January, 2016–1 December, 2020 inclusive.ResultsThe lowered b‐TSH threshold (≥8 mIU/L, ~99.5th centile) detected 1668 infants (0.48%), representing an eight‐fold increase in recall rate, of whom 212 of 1668 (12.7%) commenced thyroxine treatment. Of these 212 infants, 62 (29.2%) (including eight cases with a preliminary diagnosis of thyroid dysgenesis) had an initial b‐TSH 8–14.9 mIU/L. The positive predictive value for a preliminary diagnosis of CH decreased from 74.3% to 12.8% with the lowered threshold. Proportionally, more pre‐term infants received a preliminary CH diagnosis on screening with the lower threshold (16.1% of 62) than with the higher threshold (8.0% of 150).ConclusionClinically relevant CH was detected using the lowered threshold, albeit at the cost of an eight‐fold increase in recall rate. Further clinical and economic studies are required to determine whether benefits of lowered screening thresholds outweigh potential harms from false‐positive results on infants, their families and NBS programs.