[Evaluation of a diagnostic protocol of thyroid dysfunction].
OBJECTIVE: To assess a three-stage protocol for the study of thyroid dysfunction. DESIGN: Descriptive and retrospective. SETTING: Primary Care Centres on the island of Mallorca. PATIENTS: 8,818 non-hospital patients without any previous treatment. MEASUREMENTS AND MAIN RESULTS: In the first stage only TSH was determined. In function of its result FT4, considered a confirmation parameter, was proceeded with. Determination of T3 was reserved for the few cases of discrepancy between FT4 and TSH, when the latter was low or suppressed. We studied costs and diagnostic yield and then the possibility of improving the relationship between them by varying the cut-off used in the TSH (< 0.5 or > 4.5 microUI/ml) against the reference values (0.35-5.5 microUI/ml). The results of these two hypotheses were studied, along with a third where the significant difference of the FT4 reference limits was considered. Analytic sensitivity and specificity were acceptable in all cases. Determinations not made because they were inappropriate (FT4 and T3) comprised 45.7% of the requested or "hypothetical" determinations which would have been carried out if the protocol had not been applied, and a year's saving of 3,637,075 pesetas; which would increase by 181,753 pesetas (2.30%) if we changed the TSU cut-off to the reference values. CONCLUSIONS: This study protocol of thyroid dysfunction is adequate, although it could be improved by changing the TSH cut-offs to the reference values.