MON-101 Costa Rican National Primary Congenital Hypothyroidism Screening Program Evaluation. Retrospective Cohort Trial Between 2015 and 2017

Abstract

Congenital hypothyroidism (CH) is one of the leading causes of intellectual impairment worldwide in infancy. The newborn screening has been able to prevent this mental disability, by a prompt initiation of therapy. Over the last years the incidence of HC has been increasing, mainly by lowering the screening cut-off level that leads to detection of milder cases. There is conflicting evidence if children with mild CH without treatment may develop neurological impairment in the future.Costa Rican newborn screening program in divided in three stages, measuring serum TSH concentrations from a heel prick aliquot of capillary blood dried onto a filter paper. Each test has different TSH cut-off values to determine if the newborn needs a clinical evaluation by an endocrinologist, needs another screening test or rules out hypothyroidism.We developed an observational, descriptive, retrospective study, based on medical records, to evaluate our newborn screening program performance. The study included the total national population of screened newborns from 2015 to 2017. Descriptive analysis and analytical analysis of variables were done, and test’s sensitivity and specificity were determined.The study analyzed 204.241 screened newborns, and 145 children referred to the Endocrinology Department of the National Children Hospital. This population represents 97% of births in these 3 years. The recall rate for a first positive test was 0.3%. Congenital hypothyroidism was confirmed in 73% of children referred to the Endocrinology Department because of a positive screening. Incidence was 1/1926 births. Detection rate was independent of birth weight nor gestational age. 45.3% of children diagnosed with CH were detected after the first screening test, 52.8% after the second screening test and 1.9% after the third screening test. Screening test analysis showed high sensitivity and specificity, with diagnostic accuracy above 90%, except for the third screening test. Free thyroxine measurements alone weren’t useful to predict CH diagnosis.The coverage of the national neonatal screening program extends almost to the entire population. Our test specificity is within international standards. Incidence of CH in Costa Rica is similar to that reported in medical literature. With current detection cut-off level, there is no need of performing a second mandatory evaluation to preterm and low weight newborns. The third screening test has a low performance rate, does not improve detection of children with CH significantly, and delays clinical evaluation by the endocrinologist. Prognosis of treated children with CH is excellent, with no evidence of severe cognitive deficit.