Stricter microbiological criteria are useful surrogate in estimating the prevalence of nontuberculous mycobacterial pulmonary disease in Croatia

Abstract

Background: nontuberculous mycobacterial pulmonary disease (NTM-PD) is not reportable disease and its prevalence is difficult to assess. Aims and objectives: to assess the use of microbiological and stricter microbiological criteria in the estimation of the NTM-PD in Croatia. Methods: Retrospective analysis of the data from the Croatian national NTM registry. We included all patients with pulmonary NTM (without exclusion of any NTM species) isolated in the period from 2006 through 2013 that also had complete medical records data. Microbiological criteria of the American Thoracic Society (ATS) and stricter microbiological criteria (>3 positive sputum cultures or 1 bronchoscopic or lung biopsy culture + 1 positive sputum culture) were used to establish diagnosis of NTM-PD or NTM colonization, and then correlated to the whole ATS criteria. Results: We identified 368 patients with enough data for firm establishment of diagnosis with both criteria. By comparison of the rate of established NTM-PD diagnosis, we found good correlation between the ATS microbiological criteria and the whole ATS criteria (phi coefficient 0.66 ; 95% CI 0.61 to 0.72, p<0.0001). The strength of correlation increased when we compared stricter microbiological with full ATS criteria (phi coefficient 0.73 ; 95% CI 0.68 to 0.77, p<0.0001). Furthermore, use of ATS microbiological criteria had a positive predictive value (PPV) of NTM-PD of only 59.8%, while stricter microbiological criteria yielded a PPV of 93.3%. Conclusion: In our setting, stricter microbiological criteria may be used as surrogate criteria for laboratory based estimation of national prevalence of NTM-PD.