Objectives: To compare the results of rapid PCR screening for MRSA
using the GeneXpert system with those of cultures in an outbreak setting.
Methods: GeneXpert was used for screening MRSA in nose, throat,
groin, and other clinical samples during a 6-month period. Samples were
performed using a double-swab transystem. When >1 sample was found
positive in a screening set, all second swabs of the set were analysed by
culture.
Results: From June to October 2009, 7568 rapid tests were performed,
among which 432 (5.7%) were positive (nose: 149/2090, 7.1%; throat:
98/2078, 4.7%; groin: 152/2080, 7.3%; urine: 14/1090, 1.3%; wounds:
18/150, 12%; and others:1/27, 3.7%), and 84 (1.1%) were invalid. A
total of 1517 samples were analyzed by both rapid PCR and culture.
Rapid tests had a sensitivity of 0.896 compared to cultures, a specificity
of 0.769, a PPV of 0.763, and a NPV of 0.899. The rapid test was found
to be less sensitive in throat samples (0.81) than in nose or inguinal
samples (0.93 for both). In 32/192 (16%) patients a positive rapid PCR
result was not confirmed by culture, despite several subsequent screening
samples in some patients. Cycle threshold (Ct) for SCCmec of these PCR
positive reactions were all >30.
Conclusions: GeneXpert MRSA was found to be suitable for the rapid
detection in nose, inguinal, and throat samples, however with a lower
sensitivity in the later. Negative cultures in 16% of our PCR-positive
patients raised the question of false positivity or higher sensitivity of
GeneXpert. Further work is needed to investigate these cases.