Assistant Professor UNC Eshelman School of Pharmacy, United States
Background: Patients receiving outpatient parenteral antimicrobial therapy (OPAT) experience frequent adverse drug events, emergency visits and readmissions. Vancomycin is a common antibiotic given to OPAT patients, but there are growing concerns about its comparative effectiveness and safety. The University of North Carolina Medical Center OPAT program was started in 2015 to provide multidisciplinary monitoring and management of patients discharged on intravenous antimicrobials.
Objectives: To estimate risk differences for the relation between vancomycin use and unplanned hospital readmission or death within 90 days of discharge, compared to other OPAT medications.
Methods: We analyzed electronic health record data for an active-comparator cohort of patients discharged on intravenous antimicrobials from the UNC OPAT program between February 2015 through March 2020. Vancomycin exposure at discharge was a binary variable defined based on whether vancomycin was a component of their discharge regimen. The composite outcome of interest was 90-day unplanned hospital readmission or death. We estimated the 90-day risk difference (RD), using a multivariable linear binomial model to control measured confounding by patient characteristics (gender, calendar time, age) and clinical details (infection diagnosis, gram stain, discharge to home or skilled nursing).
Results: Among 1211 OPAT patients, the most common infections were bone/joint infection (64.8%) and bacteremia (21.5%). 456 patients (37.7%) received vancomycin at discharge. Among the patients discharged on comparator intravenous antimicrobials (besides vancomycin), the 90-day risk of unplanned readmission or death was 18.5% (n=140/755). In the crude and adjusted data, vancomycin was associated an increased absolute risk of the outcome (90-day adjusted RD 4.5%; 95% CI -0.3%, 9.2%). 90-day risk difference estimates appeared similar across strata of gender and insurance status, which were a priori effect measure modifiers of interest.
Conclusions: Vancomycin may increase the risk of 90-day hospital readmission or death among OPAT patients, compared to other OPAT medications. Further research should address potential residual confounding by severity of infection, alternative mechanisms of treatment effect heterogeneity, and heterogeneity of infection types treated with intravenous vancomycin in the outpatient setting.
