Clinical agreement in pulmonary nontuberculous mycobacterial disease

Purpose: To consider the clinical agreement among respiratory and infectious disease physicians, working in a tertiary chest diseases center serving a population with a low incidence of pulmonary tuberculosis (< 3/100,000/year), in the assessment of cases of pulmonary nontuberculous mycobacterial (NTM) lung disease. Method: A series of previously notified cases of NTM disease was abstracted and anonymously presented to a cohort of seven respiratory and infectious disease physicians. Their individual decisions to notify, treat, and follow the cases was evaluated and compared using the intraclass correlation coefficient. Results: A wide range was demonstrated in the diagnostic and management decision triage of each case by the physicians participating in the study. Clinical agreement on the likelihood of disease was limited, with an intraclass correlation coefficient of 0.394. Indication to notify the case to the state registry was linked to the clinical intent to initiate a treatment program. Conclusion: There appears to be limited agreement on the clinical significance of NTM isolates from pulmonary specimens among this cohort of experienced clinicians. If this trend is generalizable to a wider population of respiratory and infectious disease physicians, the number of notified and treated cases of disease is likely to be an underestimate of the true burden of disease in the general population. Keywords: diagnostic certainty, Kappa index, intraclass correlation coefficient, lung disease