964. Physicians' acceptable treatment failure rates and interpretation of antibiogram for Gram-negative infections: A pilot survey study of infectious diseases specialists

Abstract

Abstract Background The recommendations of many clinical guidelines include an antibiogram as a key component when making empiric therapy decisions. However, little is known about how clinicians interpret antibiograms. We aim to characterize the threshold of the acceptable risk of selecting inappropriate antibiotics among clinicians through clinical case-based questionnaires. Methods We conducted an online survey of infectious diseases (ID) specialists at the University of Iowa, as an internal pilot study. We created realistic clinical case scenarios and asked clinicians to provide the desired minimum percentage of susceptible isolates on an antibiogram for them to feel comfortable with selecting that agent for empiric therapy. We included four scenarios in the questionnaire: i) a patient with urinary tract infection (UTI) in an outpatient setting, ii) a patient with UTI in an inpatient setting, iii) a patient with Gram-negative rod bloodstream infection (GNR-BSI) in an intensive care unit (ICU) setting, iv) a patient with GNR-BSI in a non-ICU setting. Results Of the 16 of 26 providers (62%) who responded, 15 were considered effective responses. ID Physicians reported wide ranges in thresholds of minimum percentage to feel comfortable with prescribing a hypothetical antibiotic and higher thresholds for patients with more severe illnesses (outpatient with UTI: 85% [range: 75-100%]; inpatient with UTI: 95% [range: 50-99%]; inpatient with GNR-BSI in non-ICU settings: 90% [range: 80-95%]; inpatient with GNR-BSI in ICU: 95% [range: 88-100%]) (Figure 1). Figure 1.The distribution of physicians’ acceptable likelihood of susceptibility in four realistic clinical case-scenarios. Abbreviations: GNR-BSI, Gram-negative rod bloodstream infection; ICU, intensive care unit; UTI, urinary tract infection. Conclusion There were wide ranges in thresholds for interpretation of antibiogram among frontline clinicians, especially for less severe infections. Understanding physicians’ antibiogram usage can be helpful to determine how and if antibiogram guides clinicians toward more appropriate empiric therapy decisions. As this was a preliminary study, we plan to conduct a study with a larger sample size to better characterize the use of antibiogram among clinicians from diverse practice settings. Disclosures Michihiko Goto, MD MSCI, Merck & Co.: Grant/Research Support.