A procedure for evaluation and documentation of susceptibility test methods using the susceptibility of Klebsiella pneumoniae to ciprofloxacin as a model.

A new procedure for the evaluation and documentation of susceptibility test methods is described. To illustrate the procedure, four basically different susceptibility test methods were examined in a routine laboratory. The test parameter detection of decreased susceptibility to ciprofloxacin (breakpoint MIC 0.25 mg/L) among 94 selected isolates of Klebsiella pneumoniae was used. In addition to comparison of frequency histograms and regression analysis, the accuracies of the susceptibility test methods were determined using the receiver operating characteristic procedure. For each of the methods, the sensitivity (SN), specificity (SP), positive predictive value (PV+) and negative predictive value (PV-) for detection of decreased susceptibility to ciprofloxacin were calculated and plotted against a range of ciprofloxacin inhibition zones determined by the various susceptibility test methods or MICs determined by the Etest (Etest MICs). The results illustrate the accuracy and the robustness of the methods, which can be used to expose the need for training and instruction of laboratory staff. It becomes possible to optimize and justify the choice of inhibition zone breakpoints or Etest MIC breakpoints according to the SN and SP of the method employed. Furthermore, the consequences of adjustments of these breakpoints on the PV+ and PV- can be analysed and related to different clinical and epidemiological situations. We believe that our approach can be used as a model for the evaluation and documentation of susceptibility test methods in general.