Neonatal screening for congenital hypothyroidism in Sweden 1980-2013: effects of lowering the thyroid-stimulating hormone threshold.

Abstract

This study aims to evaluate the neonatal screening for congenital hypothyroidism (CH) and the diagnosis CH in the national health registers and to study the effects of lowering screening thyroid-stimulating hormone (TSH) threshold on the incidence of CH and birth characteristics of screening positive and negative CH children. This is a nationwide register-study of all children (n = 3 427 240) in the Swedish Medical Birth Register (MBR) and national cohort for screening positive infants (n = 1577) in 1980-2013. The study population was further linked to several other Swedish health registers. Evaluation of the CH screening and CH diagnosis was performed with levothyroxine use in the first year of life as reference. The incidence of CH was estimated by the Clopper-Pearson method. Regression models were used to study associations between CH and birth characteristics. The neonatal CH screening had high efficacy, but 50% of all children with a CH diagnosis were screening negative. The incidence of screening positive CH increased (1/3375 to 1/2222), and the incidence of screening negative CH decreased (1/2563 to 1/7841) after lowering the TSH screening threshold in 2009. Screening negative CH was associated with female sex, twinning, prematurity, low birth weight, birth defects, and need of neonatal intensive care, and 42% had transient disease. Despite high efficacy of the CH screening, 50% of children diagnosed as CH was screening negative. Although other factors influencing the incidence of the CH diagnosis cannot be ruled out, the incidence of screening negative CH decreased with lowering of the TSH threshold. Birth characteristics differed between screening positive and negative CH.