Abstract
Analytical and visual tools can be used to monitor progress for a variety of ASP key performance indicators, but few data describe the process of building disease-state specific tools to retrospectively monitor antimicrobial choice and duration. We describe process and methods for development of a pneumonia dashboard.
In late 2019, the Carolina ASP began construction of a dashboard to monitor antimicrobial selection and duration in patients admitted with a diagnosis code (ICD-10) consistent with pneumonia. Data extracted from the medical record after discharge included: admission date and time, admission and discharge ICD-10s, inpatient orders and administrations for agents included in the NHSN Antimicrobial Use (AU) option, and antimicrobials ordered at discharge with associated ICD-10. Extracted data fields were validated using a one-month sample. Displays were constructed to trend selection during the first 48 hours of admission, inpatient days of therapy, and total length of therapy (sum of inpatient + outpatient days) for patients who received a discharge prescription for an antimicrobial included in the AU option that was associated with an ICD-10 consistent with pneumonia. Trends observed between Jan 2020 and Mar 2021 are reported.
341 admissions were trended. Within the first two days of admission, monthly proportions of patients receiving an antimicrobial by category were: anti-MRSA therapies (vancomycin, linezolid), 0.20 to 0.75; broad spectrum beta-lactams (e.g., cefepime, pip/tazo), 0.40 to 0.81; CAP therapies (e.g., ceftriaxone, levofloxacin), 0.48 to 1.00 (Figure). Median inpatient duration of therapy was 5 days (IQR 3-8; range 1 to 68). Total length of therapy was median 6 days (IQR 4-10; range 1 to 68).
Figure. Proportions of Patients Prescribed Antimicrobial Categories of Interest During the First 48 Hours of Admissions for Pneumonia. Legend: Anti-MRSA = vancomycin or linezolid; HAP abx = cefepime, piperacillin/tazobactam, ceftazidime, meropenem; CAP = ceftriaxone, azithromycin, ampicillin/sulbactam, amoxicillin/clavulanate, cefdinir, levofloxacin.
Automated reports and visual tools can provide actionable insights for ASP practice. From this dashboard, we identified variable but high rates of anti-MRSA and broad-spectrum beta-lactam use within the first 48 hours of admission. The median inpatient and total length of therapy of 5 and 6 days, respectively, were similar to guideline-recommended durations. The up-front cost for building analytical tools can be substantial, but can be viewed as an investment if the metrics and methods are carefully selected.
All Authors: No reported disclosures
Comments