Improvement in screening performance and diagnosis of congenital hypothyroidism in Scotland 1979-2003

Jones, J H and Mackenzie, J and Croft, G A and Beaton, S and Young, D and Donaldson, M D C (2006) Improvement in screening performance and diagnosis of congenital hypothyroidism in Scotland 1979-2003. Archives of Disease in Childhood, 91 (8). pp. 680-685. ISSN 0003-9888

Full text not available in this repository.Request a copy from the Strathclyde author

Abstract

Aim: To assess the Scottish newborn screening programme for congenital hypothyroidism from 1994 to 2003 (period 2) for performance and compare with an initial audit covering 1979 to 1993 (period 1). Design: Performance data-age at blood spot sampling, notification by screening laboratory, start of treatment, and the prevalence of late testing, notification or treatment-were compared, together with the incidence of congenital hypothyroidism. Results: Comparing data for period 2 with period 1, the mean annual incidence of true congenital hypothyroidism was 1:3655 live births v 1:4363. Median age for Guthrie sampling (all referrals) was 6 v 7 days (p<0.0001 ). Late sampling (>10 days) had fallen from 10.7% to 7%. For infants requiring repeat sampling before notification, the median (range) interval between initial and final repeat samples was 11 (1 to 52) compared with 14 (3 to 73) days. Median age at notification for true congenital hypothyroidism was 10 v 12 days (p <0.0001). Late notification (>15 days) was justifiable (mild TSH elevation) in 10 of 1 3 patients in period 2. Median age at start of treatment for true congenital hypothyroidism had improved to 11 days from 1 3.5 days. For true congenital hypothyroidism, late treatment (> 16 days) occurred in 7% of patients compared with 19% (p<0.0001). Conclusions: There has been an improvement in performance measures for the congenital hypothyroidism screening programme in Scotland. However, late sampling, occurring primarily in inpatients and which is never justified, remains a problem, while the interval between initial and recall sampling is a further source of delay.