Abstract | December 13, 2021
Deficiencies of Hospital Antibiograms Generated from Rule-Based Technology and Application in Patients with Prolonged Lengths of Stay
Learning Objectives
- Discuss potential issues with rule-based technology generated antibiograms.
Introduction:
Antibiograms are unit specific annual reports of cumulative pathogen incidence and antibiotic susceptibilities, used to guide selection of empiric antibiotic therapies. Rulebased technology (RBT) expedites data compilation, but follows “first pathogen, per patient” and limits applicability, especially for prescribing antibiotics in units with higher lengths of stay. Using a single patient as an example, a Staphylococcus wound infection on day 5 and a Pseudomonas pneumonia on day 30 may be included, if both were the first time the pathogen grew via culture. Therefore, the objective was to compare the pathogens and susceptibilities of the current automated RBT antibiogram with one manually collected through chart review with additional rules accounting for days since admission, risk factors for hospital-acquired infections, and initial courses of antibiotic therapy.
Methods:
This is a single-center, retrospective cohort study utilizing chart review to assess patients admitted to the Burn Center between January 2018 and December 2019 from whom significant bacterial cultures were obtained. Demographics and burn injury characteristics were collected. Treatment data related to the infection and antibiotic usage was also collected. Specific to the culture, timing, site, pathogen, and sensitivity were collected. All cultures within the first 30 days of admission were included. The current RBT antibiogram served as the control. And new antibiogram versions were created using additional rules and compared to the RBT antibiogram.
Results:
During the 2-year study period, 657 patients were admitted. Two-hundred four patients remained after applicable exclusions. The majority of exclusions (61%) was due to lack of cultures. Mean age was 50.6 ± 16.5 years and 66% were male. Forty-nine percent were Caucasian. Seventy-two percent were admitted for an acute burn injury with flame as the primary mechanism and a median percent total body surface area of 10 (3, 21). Fifty-nine percent had at least one hospital acquired risk factor with over one-third having recent illicit drug use and one-third having a recent hospitalization. Of the 410 cultures included, 57% were Gram-negative and half were from wound infections. Sensitivities were significantly different when comparing the RBT to those created from significant cultures within 7 days of admission, cultures within 7 days of admission and with hospital-acquired infection risk factors, and only those being treated with initial course of antibiotics. Recommended empiric antibiotic changed from double coverage to a single β-lactam with > 90% susceptibility. The susceptibilities between first and subsequent courses were dramatically different.
Conclusions:
The antibiogram was significantly different from the RBT version after including factors, such as days since admission, presence of hospital acquired risk factors, or previous antibiotic courses. Before developing an antibiogram or interpreting the output, it is important to consider which automated criteria are utilized, especially for units with extended lengths of stay. Utilizing the RBT generated antibiogram over the manually-der