Abstract
The RAZOR™ EX Anthrax Air Detection System was validated in a collaborative study for the detection of Bacillus anthracis in aerosol collection buffer. Phosphate-buffered saline was charged with 1 mg/mL standardized dust to simulate an authentic aerosol collection sample. The dust-charged buffer was spiked with either B. anthracis Ames at 2000 spores/mL or Bacillus cereus at 20 000 spores/mL. Twelve collaborators participated in the study, with four collaborators at each of three sites. Each collaborator tested 12 replicates of B. anthracis in dust-charged buffer and 12 replicates of B. cereus in dust-charged buffer. All samples sets were randomized and blind-coded. All collaborators produced valid data sets (no collaborators displayed systematic errors) and there was only one invalid data point. After unblinding, the analysis revealed a cross-collaborator probability of detection (CPOD) of 1.00 (144 positive results from 144 replicates, 95% confidence interval 0.975–1.00) for the B. anthracis samples and a CPOD of 0.00 (0 positive results from 143 replicates, 95% confidence interval 0.00–0.0262) for the B. cereus samples. These data meet the requirements of AOAC Standard Method Performance Requirement 2010.003, developed by the Stakeholder Panel on Agent Detection Assays.