Introduction: Testosterone (T) is a hormone that is crucial for primary and secondary sexual development in both males and females. Free testosterone (FT) represents the biologically active form of T, and its measurement is of great importance in clinical practice. While application of either equilibrium dialysis or ultrafiltration is considered to be the gold standard for FT assessment, these methods are expensive and not widely accessible. As an alternative, the Vermeulen formula is a commonly utilized calculated method.
Methods: This clinical study, including 190 consecutive patients, was carried out to compare FT levels obtained through direct immunoluminometric assay and the Vermeulen formula. The comparison was performed using Passing-Bablok and Deming regression as well as the Bland-Altman plot. Sensitivity, specificity, accuracy, positive predictive value, and negative predictive value were assessed.
Results: The calculated method employing the Vermeulen formula was considered the gold standard. Passing-Bablok regression indicated a good agreement between the two methods, with slopes close to 1 for the whole series. Although the Bland-Altman plot demonstrated overall agreement, a potential proportional bias was observed in females. Deming regression confirmed excellent agreement and reliable estimates. Sensitivity and specificity analysis revealed that the direct method had a sensitivity of 75.0% and a specificity of 93.4% in all patients. However, sensitivity improved to 81.0% in males and dropped to 18.2% in females likely due to the low number of true positive cases.
Conclusion: The direct method exhibited comparable performance to the calculated method, but caution should be exercised when interpreting results, particularly in females. Further studies are necessary to validate its sensitivity and specificity in larger series.
Keywords: Clinical series; Direct method; Free testosterone; Vermeulen formula.
© 2024. The Author(s), under exclusive licence to Hellenic Endocrine Society.