The International Working Group on Mycobacterial Taxonomy conducted a cooperative study to standardize and evaluate the reproducibility of laboratory tests used for identifying and classifying mycobacteria. Sixteen laboratories examined 18 bacterial strains using 25 different techniques. The researchers established rigorous criteria for reproducibility and differential power, requiring at least 90% mean agreement between laboratories, clear positive and negative results, and adequate replicate data. Five tests met all criteria: urease activity, pigment production (photochromogenic, scotochromogenic, or nonchromogenic), tolerance to 5% sodium chloride, hydrolysis of Tween 80, and β-galactosidase activity. These tests demonstrated high reproducibility across laboratories, with mean agreement ranging from 90% to 96.7%. The study identified potential sources of error including labeling mistakes, procedural deviations, strain variability, and technical factors. The authors recommend incorporating these five standardized tests into minimal standards for mycobacterial identification, noting that future studies with broader organism representation and greater replication may identify additional reproducible tests.

Key findings

• Five laboratory tests achieved ≥90% reproducibility across 16 participating laboratories: urease, pigment production, NaCl tolerance, Tween 80 hydrolysis, and β-galactosidase activity
• These standardized tests demonstrated high mean agreement with consensus results (90.0-96.7%) and high internal laboratory consistency for differentiating mycobacterial species
• Some laboratories showed disproportionately high error rates despite overall test reliability, suggesting that procedural adherence and experience are critical factors
• The extended 10-day incubation criterion for Tween 80 hydrolysis improved reproducibility compared to the original 5-day criterion
• The five highly reproducible tests are recommended for inclusion as minimal standards in mycobacterial taxonomy and nomenclature