Computerized antimicrobial decision support: an offline evaluation of a database-driven empiric antimicrobial guidance program in hospitalized patients with a bloodstream infection

Abstract

Introduction: We developed a computerized antimicrobial guidance program based on the last 5 years of our laboratory culture data augmented by expert infectious disease logic. The program is designed to assist physicians with the targeting of empiric antimicrobials for hospitalized patients by tracking pathogenic bacteria and their evolving antimicrobial resistance profiles. Costs, toxicities, and environmental impact of antimicrobial use also influence the final recommendations. We undertook the following analysis to verify its potential safety and efficacy in hospitalized patients with a bloodstream infection. Methods: We retrospectively enrolled all inpatients with a positive blood culture for a previously undetermined pathogen during the first 6 months of 2002 and determined the empiric therapy initiated within the 12 h before and after the time of culture. Antimicrobial recommendations from the microbiologic decision support tool were then determined by matching specimen (blood), hospital unit, community- versus hospital-acquired category, age category, and gender. Generated antimicrobial recommendations were tailored to patient allergies, age category, and presence of pregnancy, lactation, or hepatic impairment. Results: The microbiology laboratory recorded 226 unique patient/pathogen blood cultures during the study period. Physicians initiated effective empiric therapy in 150 of the 226 cases, for an effectiveness rate of 66%. The computer-guided therapy was effective in 195 of the 226 cases for a rate of 86%. A contingency table analysis showed 55 cases where the computer recommendation was effective but the physicians’ selection was not, and eight cases where the physicians’ antimicrobials were effective but the computer’s were not (P<0.0001). Discussion: For patients with a bloodstream infection, we found that our computer-guided statistically-derived antimicrobial therapy would potentially improve the rate of effectiveness of empirically chosen antimicrobials.