We thank Dr. Hoffman for his critical comments (1). According to his calculations, based on our cross-sectional study data, he found a specificity of 80% and only a low sensitivity of 35% for the thyroid antibody–screening test. Thus, he does not support our recommendation for yearly examinations of thyroid antibodies (2).
We agree with him that screening tests should be highly sensitive, but the data of our multicenter study do not allow a true estimation of the sensitivity, since patients have not been followed longitudinally and since we missed those patients who will develop hypothyroidism later on during their course of diabetes. Indeed, we examined patients with type 1 diabetes at the Pediatric Diabetes Outpatient Clinic of the Otto-Heubner-Center, Charité, Berlin, and found that 8 of 16 patients with positive antibodies developed thyroid stimulating hormone (TSH) elevation after an observation time of 2–6 years (median 3.5) (3). Therefore, patients with positive antibodies should be monitored for TSH elevation at yearly intervals.
In addition, the German Association of Pediatric and Adolescent Medicine recommends that patients with positive thyroid antibodies and concomitant thyroid gland enlargement with a typical hypoechogenic pattern in ultrasound studies should receive treatment with l-thyroxine.
For these reasons, we have been performing thyroid antibody–screening tests at our institution in all patients with diabetes since 1998.
References
Address correspondence to Olga Kordonouri, MD, Klinik für Allgemeine Pädiatrie, Otto-Heubner-Centrum, Charité, CVK, Augustenburger Platz 1, 13353 Berlin, Germany. E-mail: [email protected].
