Methods: Electronic medical record data was obtained from the University of Kentucky Center for Clinical and Translational Science Enterprise Data Trust from 9/1/2007 to 10/1/2015. Adult patients receiving NAF or CFZ for greater than 48 hours were included. Patients with renal dysfunction or vancomycin therapy were excluded. AKI was assessed with the RIFLE criteria. Patients were matched on baseline renal function, concomitant nephrotoxins, and comorbidities predisposing to AKI. Comparative statistics and multivariate logistic regression were performed.
Results: In total, 3951 patients were evaluated, with 518 receiving NAF and 3433 receiving CFZ. At baseline, the NAF group was younger (49.6±16.1 v 55.7±16.3 years old, p <0.0001), more likely to be male (62 v 57%, p=0.046) and had higher baseline renal function (101±41 v 92±36 mL/min, p<0.0001). AKI occurred in 29.2% of NAF patients compared to 14.8% of CFZ patients (p<0.0001). NAF patients were more likely experience in-hospital mortality (14.3% v 4.9%, p<0.0001); however, when stratified by endocarditis, osteomyelitis, or bacteremia, this trend was not found. Following matching, the AKI rate for NAF patients remained 29.2%, and the rate in CFZ patients was 21.6% (p=0.007). Receipt of NAF was associated with an adjusted OR of 1.59 (95% CI 1.19-2.14) for AKI compared to CFZ. Other factors associated with increased odds of AKI were exposure to loop diuretics (OR 2.38, 95% CI 1.76-3.23) or hypotension (OR 2.17, 95% CI 1.6-2.97). Mortality remained higher in the NAF group (14.3% v 8.7%, p = 0.006); however, no difference was noted in patients with endocarditis, osteomyelitis, or bacteremia. Similar numbers of patients in both groups were discharged healthy to home. Length of hospitalization was similar between groups.
Conclusion: NAF was associated with more AKI and in-hospital mortality compared to patients treated with CFZ, which suggests CFZ may be a reasonable alternative therapy.
W. C. Rutter,