Evaluation of the usefulness of a sensitive immunoradiometric assay for thyroid stimulating hormone as a first-line thyroid function test in an unselected patient population.

A new sensitive immunoradiometric assay for serum thyroid stimulating hormone (TSH) was applied to 1627 consecutive patients from four large clinics at Malmö General Hospital, in whom a thyroid function disorder was suspected. The final diagnoses were taken from the patients' charts at a follow-up 1-2 years after the primary evaluation. If only those patients without thyroid disease were used as a reference population, the 5th percentile was 0.8 mIU/l. When this cutoff limit was applied to the unselected patient population, the TSH assay had a high sensitivity for the diagnosis of hyperthyroidism (96%), a lower specificity (64%) and a very low predictive value (10%). Therefore, in the majority of the patients (72%) the sensitive TSH assay cannot be used alone to define thyroid function.