National and international surveillance data are of great value for understanding global trends in antibacterial resistance and provide a repository of data for analyses to determine factors predictive of decreased susceptibility. It is, however, even more crucial to understand the pattern of antibacterial resistance trends locally and within the institution of interest in order to detect early signals of potentially serious problems. To this end, the appropriate compilation of susceptibility data is critical.
Until recently, hospitals followed their own set of guidelines for abstracting and presenting data in the form of an antibiogram. However, standardized guidelines to gather, analyze, and present cumulative antimicrobial susceptibility test data in the form of an antibiogram have been published in CLSI document M39-2A (50). Through this guideline, hospitals have a standardized methodology for data extraction for all drugs tested and for reporting results. For example, guidance is given as to which isolates should be included in the analysis (e.g., the first isolate from a patient within an analysis period), the analysis period (at least annual), population tested (e.g., inpatient, intensive care unit, or nursing home), specimen source, and a reasonable minimum of number of isolates for each organism (n = 30). A standardized approach to constructing antibiograms will facilitate internal and external benchmarking of antibacterial resistance patterns. The commitment to collecting antibacterial use and clinical outcome data will further enhance the value of data derived from benchmarking. Such activities will ultimately benefit and support antimicrobial stewardship activities and formulary decisions.
As described in this chapter, current susceptibility breakpoints are not optimal for all classes of agents or all patient populations and are in the process of being re-evaluated by groups such as CLSI and the European Committee on Antimicrobial Susceptibility Testing (EUCAST). As such, benchmarking of susceptibility data characterized as susceptible, intermediate, and resistant is less informative, and evaluations based on such comparisons may lead to suboptimal decisions with regard to formulary decisions and empiric prescribing. To optimize the value of data from microbiologic tests, institutions will need to implement an automated susceptibility testing methodology that can deliver MIC values, or at least a limited range of such values. Actual MIC data will allow for benchmarking of MIC distributions (including MIC50 and MIC90 values) in addition to the proportion of isolates that are resistant. The former will allow for detection of changes in susceptibility patterns prior to the occurrence of any large shift in MIC values into the resistant category. From a prescribing perspective, actual MIC data will also allow for interpretation of PK-PD target attainment analyses and, ultimately, better dose selection for an individual patient. For example, if testing for a particular microorganism demonstrates a MIC value in a higher range, dosing regimens may be optimized for the patient.
In conclusion, benchmarking can be a valuable and powerful tool in the fight against antibiotic resistance, but it is only as effective as the data used. Data from large multinational surveillance programs are of great value to understanding patterns of antibacterial resistance, and to date, these data have been under-utilized. The evaluation of MIC distributions using statistical tools that can accommodate quantitative MICs (including the pattern of MIC censoring) will allow for a better understanding of the impact of institution-, patient-, and microorganism-specific factors on changes in MIC, before high-level resistance develops in multiple regions. Future endeavors will require broad-based quality data from surveillance programs that are not dependent solely on funding from the pharmaceutical industry.
Was this article helpful?