ABSTRACT
OBJECTIVES: To detect newborns with congenital hypothyroidism (CH) and to treat the affected infants as early as possible. STUDY DESIGN: Cord blood thyrotropin (TSH) screening for CH in Ramathibodi Hospital began in 1993. From October 1993 to December 1998, 35,390 neonates were screened. The infants with elevated TSH level of greater than 30 mU/L were recalled for verification of CH. Confirmation tests included total thyroxine, free thyroxine and TSH level. Thyroid scan and uptake were performed in some affected infants. RESULTS: Twelve infants with CH were detected resulting in an incidence of one in 2,949 live-births. All affected infants were asymptomatic at birth. Of 12 infants with CH, one premature neonate had a delayed TSH elevation and was diagnosed as having primary hypothyroidism at 2 months of age. The recall rate for validation of CH based on a cut-off value at serum TSH level of greater than 30 mU/L is 1.1 per cent. If the cut-off value of serum TSH level was raised to greater than 40 mU/L, the recall rate would decrease to 0.43 per cent. None of the affected infants had cord blood TSH level of less than 50 mU/L except one premature patient. Therefore, beginning in January 1997, the cut-off value of TSH was raised to 40 mU/L or greater. Pitfalls in this program include incomplete blood-specimen collection and incomplete follow-up. To strengthen the program, improvements were made in the follow-up system from 1996 onward. Therefore, the coverage for blood-specimen collection progressively increased from 84 per cent in 1994 to 96 per cent in 1998. Simultaneously, the patients' return after recalls also increased from 38 per cent to 100 per cent. CONCLUSIONS: The incidence of CH in Ramathibodi Hospital is approximately 1:3,000 live-births. The optimal cord blood TSH level for recall is 40 mU/L or greater. The intensification of follow-up strategy resulted in better response to recall and earlier treatment in the affected infants.