Nonparametric Calibration of Two Common Antimicrobial Susceptibility Tests Based on Interval-Censored Data with Measurement Error

• Activity Number: 370
• Type: Contributed
• Date/Time: Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): Bruce A. Craig*+ and Xiaoli Qi
• Companies: Purdue University and Oracle Corporation
• Keywords: nonparmetric regression ; diagnostic test ; error-in-variables ; susceptibility testsing ; nonparametrics

Abstract

Drug dilution (MIC) and disk diffusion (DIA) are the two common tests to determine pathogen susceptibility to antibiotics. For each, two drug-specific breakpoints classify a pathogen as either susceptible, intermediate, or resistant. Unfortunately, DIA breakpoints are not straightforward to calculate. Current Clinical and Laboratory Standards Institute guidelines require a scattergram of test results and the DIA breakpoints are based on limiting the MIC/DIA classification discrepancies. This approach is limited by the fact that certain experimental errors are ignored. We will discuss a nonparametric approach to this error-in-variables problem which factors in the uncertainty of both tests, the relationship between the two tests, and the underlying distribution of pathogens. A loss function is then used to determine the DIA breakpoints. We conclude by comparing these two approaches.