Effect of successive tuberculin skin test readings on healthcare professionals' proficiency: an operational observational study in five high-tuberculosis-burden Brazilian municipalities, 2023-2024

Objective To verify the effect of repeated readings of the tuberculin skin test on the proficiency of healthcare professionals. Methods This was an operational observational study that compared the readings of professionals and lead trainers during training sessions conducted by the Ministry of Health in five Brazilian municipalities. The mean difference between readings, agreement, sensitivity, and specificity were calculated, considering the lead trainer as the gold standard for each of the first 50 readings. The Kappa index assessed interobserver agreement. The database originated from a training program on tuberculin skin test reading. The Statistical Package for the Social Sciences Statistics (SPSS) software was used for descriptive analyses. Results 168 professionals were trained, and a total of 5,929 readings were performed. The average number of readings per professional was 35.3±29.2 (minimum=5, maximum=122). The mean difference between the readings of the professionals and those of the lead trainers was 0.01 mm (±0.70 mm). The agreement was 93.0% for readings between 0 mm and 4 mm, 74.8% for readings between 5 mm and 9 mm, and 88.2% for readings of 10 mm or more (weighted Kappa coefficient 0.864). Sensitivity and specificity were 96.5% and 93.0% for the 5 mm cutoff point (the most commonly used in Brazil) and 88.2% and 95.0% for the 10 mm cutoff. None of the evaluated parameters changed during the first 50 readings. Conclusion There was no effect of successive readings on the professionals' proficiency, which was high from the first reading according to all parameters. These findings support the newly published training recommendations of the Ministry of Health.