Study objective: The tuberculin skin test is the best diagnostic method to detect tuberculous infection. How accurate is interpretation of the test?
Design: Observational study.
Setting: Both general hospital and university hospital.
Participants: One hundred seven health-care professionals, including 52 practicing pediatricians, 33 pediatric house officers, 10 pediatric academicians, 11 registered nurses, and 1 pediatric nurse practitioner.
Study: A tuberculin skin test (Mantoux) was applied to the arm of a known tuberculin converter. As participants entered/left the room, they were guided to the tuberculin converter. At no time did a participant observe readings other than his/her own.
Results: Mantoux tuberculin reaction measuring 15 mm induration was read individually by a group of 52 practicing pediatricians, 33 pediatric house officers, 10 pediatric academicians, 11 registered nurses, and one pediatric nurse practitioner. The median induration recorded by this group of 107 health-care professionals was 10 mm, and 17 (33%) practicing pediatricians read the reaction as <10 mm induration. Using the ≥15-mm induration indicator to identify a positive reaction, 93% of those in the study (99/107 participants) would have identified our known converter as tuberculin negative.
Conclusion: This study confirms a general inaccuracy in interpretation of the tuberculin skin test reaction. It raises two questions. (1) Is there a general tendency toward underreading? (2) Does this general tendency to underread tuberculin skin test reactions raise some question as to the American Academy of Pediatrics, American Thoracic Society, and Centers for Disease Control and Prevention move in raising the amount of induration considered tuberculin positive to 15 mm in low-risk individuals?