Characterization of clinically significant strains of coagulase-negative staphylococci

On occasion, a patient may have two or more clinical cultures yielding a coagulase-negative staphylococcus If these multiple isolates have the same phenotype, one might conclude that the same strain was reisolated from the patient, indicating its persistent and pathological presence. We examined the validity of this conclusion when we applied a number of characterizing systems to a collection of 143 isolates of coagulase-negative staphylococci collected during an outbreak of intravascular catheter-associated sepsis. The probability of classifying two random isolates as the same phenotype or species was as follows: P = 0.356 for phage typing, P = 0.348 for Baird-Parker biotyping, P = 0.346 for the API STAPH-IDENT (Analytab Products) system, P = 0.327 for Bentley et al. biotyping, and P = 0.077 for antimicrobial susceptibility patterns. Although antimicrobial susceptibility patterns had the lowest probability, a variability in test results of 7.7% and a tendency for strains to have similar antibiograms effectively raised the probability to P = 0.897. The combination of the API STAPH-IDENT with antibiograms resulted in a probability of P = 0.037 to P = 0.147. When all of the above methods were used together a probability of P = 0.014 was achieved. Five patients had isolates from two or more blood cultures spaced more than 1 day apart that were identical by all of the above criteria, thus confirming prolonged bacteremia. The collection was also examined for the incidence of slime production. Slime production was not associated with any of the above groups, but was associated with symptomatic infections (P less than 0.05) and gentamicin resistance (P less than 0.01). Slime production was strain stable and was of assistance in typing strains of coagulase-negative staphylococci.