Methods: A retrospective chart review of inpatients over age 18 receiving VP or VC for ≥48 hours was conducted at Rochester General Hospital from 1/1/14 to 8/30/14. Patients were excluded if a study antibiotic was received in the past 7 days, study antibiotics started more than 48 hours apart, undergoing dialysis, serum creatinine>1.5 times baseline or creatinine clearance <30mL/min at therapy initiation. We estimated that 43 patients were required per group for 80% power to detect a 25% difference in acute kidney injury (AKI) by AKIN criteria. Concomitant nephrotoxin use and Charlson Comorbidity Index (CCI) scores were also recorded. Χ2 tests were used to evaluate for AKI. Univariate analysis was conducted for AKI risk factors, followed by multivariate analysis.
Results: AKI occurred in 36/111 (32%) of VP courses versus 10/66 (15%) of VC courses, p=0.011, OR 2.13. There was no difference in days of therapy, length of stay, CCI score, or median and maximal vancomycin levels between groups. Most patients were classified as AKIN stage 1 (VP 69%, VC 80%, p=0.7). AKI developed later in VC than VP (6.5 versus 3 days, p=0.028), however AKI duration was similar. Dialysis was more often needed with VP (8% vs 0%, p=0.6). More VC patients received concomitant chemotherapy (3% vs 11%, p=0.028) and acyclovir (0% vs 11%, p=0.001), while ACE/ARB therapy was more common in VP cases (35% vs 17%, p=0.008). By multivariate analysis these variables were not independently associated with AKI.
Conclusion: The higher incidence of AKI during VP therapy corresponds to the incidence reported in prior studies. The mechanism of AKI is not known, but it is likely that the increased risk arises from the combination, since none of these antibiotics used alone have high AKI risk. Clinicians should consider this differential risk of nephrotoxicity when initiating empiric antibiotics in hospitalized patients.
J. K. Hix, None
M. Laguio, None