E Bulliard1, B Grandbastien1, L Senn1, G Greub2, D S Blanc3. 1. Service of Hospital Preventive Medicine, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland. 2. Institute of Microbiology, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland. 3. Service of Hospital Preventive Medicine, Lausanne University Hospital and University of Lausanne, Lausanne, Switzerland; Swiss National Reference Centre for Emerging Antibiotic Resistance (NARA), Fribourg, Switzerland. Electronic address: Dominique.Blanc@chuv.ch.
Abstract
OBJECTIVES: Screening for methicillin-resistant Staphylococcus aureus (MRSA) is part of many recommendations to control MRSA. Several rapid PCR tests are available commercially and updated versions are constantly released. We aimed to evaluate the performance of three consecutive versions (G3, Gen3 and NxG) of the XpertMRSA test. METHODS: Routine samples for MRSA screening were simultaneously tested by culture and rapid PCR. The three versions of XpertMRSA were used successively and compared with culture. RESULTS: A total of 3512, 2794 and 3288 samples were analysed by culture and by the G3, Gen3 and NxG XpertMRSA versions, respectively. The rates of positive-by-culture in the three groups were 5.0%, 4.7% and 4.3%, respectively. The sensitivity improved over time (71.4, 95% CI 64.0-77.9; 82.3, 95% CI 74.4-88.2; and 84.3%, 95% CI 77.0-89.7, respectively), but not significantly. The specificity (98.4, 95% CI 97.9-98.8; 96.8, 95% CI 96.0-97.4; and 99.1, 95% CI 98.7-99.4, respectively) and the positive likelihood ratios (45.7, 95% CI 34.4-60.8; 25.6, 95% CI 20.5-32.0; and 97.1, 95% CI 66.3-142.4) were significantly lower in the Gen3 version (p < 0.00001). CONCLUSIONS: These significant differences in performance show the importance of evaluating each new version of a commercial test.
OBJECTIVES: Screening for methicillin-resistant Staphylococcus aureus (MRSA) is part of many recommendations to control MRSA. Several rapid PCR tests are available commercially and updated versions are constantly released. We aimed to evaluate the performance of three consecutive versions (G3, Gen3 and NxG) of the XpertMRSA test. METHODS: Routine samples for MRSA screening were simultaneously tested by culture and rapid PCR. The three versions of XpertMRSA were used successively and compared with culture. RESULTS: A total of 3512, 2794 and 3288 samples were analysed by culture and by the G3, Gen3 and NxG XpertMRSA versions, respectively. The rates of positive-by-culture in the three groups were 5.0%, 4.7% and 4.3%, respectively. The sensitivity improved over time (71.4, 95% CI 64.0-77.9; 82.3, 95% CI 74.4-88.2; and 84.3%, 95% CI 77.0-89.7, respectively), but not significantly. The specificity (98.4, 95% CI 97.9-98.8; 96.8, 95% CI 96.0-97.4; and 99.1, 95% CI 98.7-99.4, respectively) and the positive likelihood ratios (45.7, 95% CI 34.4-60.8; 25.6, 95% CI 20.5-32.0; and 97.1, 95% CI 66.3-142.4) were significantly lower in the Gen3 version (p < 0.00001). CONCLUSIONS: These significant differences in performance show the importance of evaluating each new version of a commercial test.