Context: Congenital hypothyroidism (CH) is reportedly increasing in the United States, possibly reflecting changes in screening methods. In Québec, the same initial TSH cutoff (15 mU/liter) has been used for the last 20 yr, but in 2001, the cutoff was decreased from 15 to 5 mU/liter for the second test, which is requested when TSH is intermediate (15-30 mU/liter) on the first.
Objectives: Our objective was to assess the incidence of CH over the last 20 yr in Québec.
Design, setting, patients, and main outcome measure: This is a population-based retrospective study. Incidences by etiology based on thyroid scintigraphy with technetium were compared between 1990-2000 and 2001-2009.
Results: Of 1,660,857 newborns over 20 yr, 620 had CH (incidence 1:2679). Etiology was dysgenesis (n = 389, 1:4270), either ectopy (n = 290) or athyreosis (n = 99), goiter (n = 52, 1:31,940), normal-size gland in situ (n = 115, 1:14,442), and unknown (n = 64, 1:25,950). The new screening algorithm identified 49 additional cases (i.e. 25 normal-size gland in situ, 12 unknown etiology, 10 ectopies, and two goiters). Consequently, the incidence of normal-size gland in situ or of unknown etiology more than doubled (1:22,222 to 1:9,836, P = 0.0015; and 1:43,824 to 1:17,143, P = 0.0018, respectively) but that of dysgenesis and goiter remained stable. Had the 1990-2000 algorithm been applied in 2001-2009, no change in incidence would have been observed in any category.
Conclusion: Estimating the incidence of CH is influenced by minimal changes in TSH screening cutoffs. Lower cutoffs identify additional cases that have predominantly functional disorders whose impact on intellectual disability, if left untreated, remains to be determined.