Abstract

Although clinical benefits for aminoglycoside dosing services have been suggested, this has not been clearly documented in a prospective fashion. Therefore, we randomly assigned patients to be dosed (1) by their physician (Physician dosing method), (2) by predicting an initial dosage (Predictive dosing method), or (3) by calculating an initial dosing regimen by measuring the pharmacokinetics for the individual patients after a loading dose (Individual dosing method). The patients' clinical response and nephrotoxicity were then evaluated. The individual dosing method resulted in erratic aminoglycoside levels, necessitating its elimination from the study. This group was not included in the final analysis. Of the 164 patients entering the study, 41 had a documented gram-negative infection, received aminoglycosides for more than 2 days, and had serum aminoglycoside levels measured. The predictive dosing method in these 41 patients produced statistically significant higher peak and lower trough levels, but there was no difference in the incidence of nephrotoxicity or clinical response. The 95% confidence intervals precluded any major clinical benefit in these patients with documented gram-negative infections. We question the previous findings of increased efficacy and decreased nephrotoxicity with the use of an aminoglycoside dosing service and suggest that larger studies be done.