BACKGROUND: Accurate diagnosis of TB infection (TBI) is challenging due to the lack of a gold standard. Tuberculin skin test (TST) and interferon-gamma release assay (IGRA) are currently useful in TBI diagnosis, but both have several limitations. This study aims to evaluate inter-operator variability in TST measurements and determine its impact on TBI diagnosis and treatment.METHODS: This was a retrospective analysis of patients screened for TBI using at least TST at a public outpatient clinic specialised in TB from January 2019 to August 2021. TST readings performed by five experienced nurses were compared.RESULTS: A total of 671 screenings were analysed. TST positivity rate (P < 0.001) and mean TST measurements obtained by our nurses were significantly different (P < 0.001). Concordance of TST and IGRA results was of 83.4% in the overall population (κ = 0.479). However, TST/IGRA agreement was significantly different among nurses (P = 0.003).CONCLUSION: Our analysis of TST measurements by experienced nurses shows significant differences in TST positivity rate, mean measured values and overall concordance with IGRA. This led to significant different outcomes in TBI diagnosis and subsequent treatment. TST measurement differences could potentially be more pronounced if we considered untrained operators or those with occasional reading experience.