Epidemiologic surveillance to detect false-positive Mycobacterium tuberculosis cultures

Abstract

This study was aimed to investigate the ability of potential indices from epidemiologic surveillance to detect false-positive cultures of Mycobacterium tuberculosis (MTB). All clinical specimens for mycobacterial culture from April 1 to August 31, 2010, were reviewed. Single-positive cultures without relevant clinical and pathologic information were categorized as suspected false-positive cultures. Genotyping methods were used to confirm false-positive cultures. The performance of epidemiologic surveillance indices to detect potential false-positive cultures was evaluated. A total of 14,462 specimens were sent to the laboratory and 214 batches were processed in 107 work days (average 67.6 specimens per batch, ranging from 21 to 130 specimens per batch). Seventy-one single-positive cultures were identified, among which 5 cultures of multidrug-resistant MTB in 1 batch were false-positive, confirmed by genotyping methods. Epidemiologic surveillance with statistical process control charts for single-positive cultures per day showed good performance in epidemiologic surveillance. The false-positive rate was 38.5% in the 13 potential false-positive cultures according to the statistical process control chart for single-positive cultures per day. Although the incidence of tuberculous disease is high in Taiwan, clustering of multidrug-resistant MTB in 1 batch or clustering of single-positive cultures still suggested the occurrence of false-positive MTB cultures. Therefore, epidemiologic surveillance for the clustering of single-positive cultures with the statistical process control chart could be used to monitor the occurrence of false-positive results.