English

Abstract

Objective: To measure Tuberculin Skin Test (TST) at 48 and 72 hours and, if there are differences, to determine the concordance and whether it is of clinical significance (change in the TST score). Material and method: Prospective study of those admitted to prison between March and June 2023, screened by means of TST read at 48 and 72 hours by trained personnel, assessed according to criteria in force in Spain. When the measurement varied, concordance was checked using the kappa index and its clinical significance. To determine variables associated with clinical significance, bivariate and mutivariate analysis was performed using logistic regression. Results: 488 cases were studied. TST was positive in 20.1, 23.4 and 23.8% at 48 hours, 72 hours or any of them, respectively.There was more TST positive in foreigners or BCG vaccinated. The reading varied in 35.2% (mean deviation 5.5 +/–4.4 mm) and was higher at 72 hours (86% of different cases). Inter-reading concordance was excellent overall (κ = 0.892) and in BCG vaccinated (κ = 0.805), but moderate in immunocompromised (κ = 0.421). Discordance was clinically significant in 3.7% and was not associated with any variable. Discussion: TST concordance at 48-72 hours is excellent, but in some patients (3.7% in this work) the score is incorrect, more so in readings at 48 hours. The proportion is low, but can be reduced if the reading is taken at 72 hours and Interferon Gamma Release Assay tests are used in some cases (immunocompromised, etc.).