Aims: Two commercial methods for the identification of coagulase-negative staphylococci (CNS) were compared with the restriction fragment length polymorphism (RFLP) of the amplified tuf gene, which served as the reference method.
Methods and results: One hundred and forty-five CNS were evaluated using the API 32 Staph ID and the Crystal GP/ID BBL systems. The PCR-RFLP of the tuf gene served as the reference method. The APIStaph and the GP/ID BBL had an overall rate of agreement with the molecular method of 58.6% and 46.2% respectively, with the inability of the GP/ID BBL to characterize 11.7% of the isolates. The APIStaph showed higher sensitivity and better agreement than the GP/ID BBL with the PCR-RFLP, except for Staphylococcus hominis and Staphylococcus capitis.
Conclusions: Neither of the commercial systems was as reliable as the PCR-RFLP method for identifying isolates of CNS. Overall the APIStaph had better agreement with the PCR-RFLP than the GP/ID system.
Significance and impact of the study: The results indicate that the PCR-RFLP method is more reliable than the two commercial systems tested, suggesting that it is more reliable for routinely identifying CNS.