Analysis of parameters and validation of method for normalized interpretation of antimicrobial resistance

The method of normalized resistance interpretation (NRI), uses the high-zone side of the susceptible peak in a zone diameter histogram as an internal calibrator to construct the real standard distribution of susceptible isolates even in the presence of resistant isolates. NRI parameters were optimized using control strain histograms from microbiology laboratories in Stockholm, Argentina, and the Philippines. A moving average based on four-zone values was slightly better than based on two-zone average values. The optimal peak adjustment from the switch position of the moving average was 1.0 for two-zone averages and 2.5 for four-zone averages. A comparison between true means and NRI-calculated means showed a highly significant correlation (R-2 = 0.963). Coefficients of variation (CV), comparing the CV of the true distribution of control strain test results with the NRI calculated distribution, identified two types of aberrant histograms. NRI calculations on clinical isolates of Escherichia coli and Staphylococcus aureus from selected laboratories showed a good agreement between the local resistance interpretations with the NRI calculated levels. One type of deviation was most marked with cephalothin histograms for E coli isolates where the regular zone breakpoints used cut through the population of susceptible strains. With proper markers for required quality of disc test results, the NRI method might be a valuable tool for both resistance surveillance and for quality control of the disc diffusion method. (C) 2003 Elsevier Science B.V. and the International Society of Chemotherapy. All rights reserved.