Prospective Evaluation of the Effect of an Aminoglycoside Dosing Regimen on Rates of Observed Nephrotoxicity and Ototoxicity

The nephrotoxicity and ototoxicity associated with once-daily versus twice-daily administration of aminoglycosides was assessed in patients with suspected or proven gram-negative bacterial infections in a randomized, double-blind clinical trial. Patients who received therapy for >/=72 h were evaluated for toxicity. Patients also received concomitant antibiotics as deemed necessary for treatment of their infection. Plasma aminoglycoside concentrations, prospective aminoglycoside dosage adjustment, and serial audiologic and renal status evaluations were performed. The probability of occurrence of a nephrotoxic event and its relationship to doses and daily aminoglycoside exposure served as the main outcome measurement. One hundred twenty-three patients were enrolled in the study, with 83 patients receiving therapy for at least 72 h. For 74 patients plasma aminoglycoside concentrations were available for analysis, and the patients formed the group evaluable for toxicity. The primary infectious diagnosis for the patients who were enrolled in the study were bacteremia or sepsis, respiratory infections, skin and soft tissue infections, or urosepsis or pyelonephritis. Of the 74 patients evaluable for toxicity, 39 received doses twice daily and 35 received doses once daily and a placebo 12 h later. Nephrotoxicity occurred in 6 of 39 (15.4%) patients who received aminoglycosides twice daily and 0 of 35 patients who received aminoglycosides once daily. The schedule of aminoglycoside administration, concomitant use of vancomycin, and daily area under the plasma concentration-time curve (AUC) for the aminoglycosides were found to be significant predictors of nephrotoxicity by multivariate logistic regression analysis (P </= 0.001). The time to a nephrotoxic event was significantly influenced by vancomycin use and the schedule of administration, as assessed by Cox proportional hazards modeling (P </= 0.002). The results of the multivariate logistic regression analysis and the Cox proportional hazards modeling demonstrate that both the probability of occurrence and the time to occurrence of aminoglycoside nephrotoxicity are influenced by the schedule on which the aminoglycoside is administered as well as by the concomitant use of vancomycin. Furthermore, this risk of occurrence is modulated by the daily AUC for aminoglycoside exposure. These data suggest that once-daily administration of aminoglycosides has a predictably lower probability of causing nephrotoxicity than twice-daily administration.