Propuesta para mejorar la sensibilidad del algoritmo diagnóstico de evaluación de la función tiroidea

Abstract

The measurement of thyrotropin and thyroxine concentrations in serum is the basis of the biochemical evaluation of thyroid function. The reference interval of thyrotropin is frequently used as an initial screening to assess the need for thyroxine measurement. This study was carried out to obtain a different and more adjusted interval of thyrotropin, in order to improve the diagnostic sensitivity. All of the results of thyrotropin and thyroxine requested at the same time on outpatients were selected: for the first part of the study, those from year 2008 (n = 10,900), and for the second part, those from 2009 with no follow-up in the previous year (n = 5,367). Two ROC curves were used to define the algorithm decision interval with a sensitivity of 90% and the number of false negative results was calculated. The thyrotropin intervals obtained in the first and second studies were (2.11-3.50) mIU/L and (2.04-3.41) mIU/L, respectively. In both studies, the sensitivity increased approximately from an average of 70% to 90% of the confidence interval using the algorithm interval. The number of false negatives was reduced from 75 to 30 in the first case, and from 37 to 13 in the second case. The application of a calculated thyrotropin interval to assess thyroid function in outpatients with or without prior monitoring, leads to an increase of the diagnostic sensitivity with regard to the use of the thyrotropin reference interval.