Five-year surveillance of aminoglycoside usage in a university hospital.

Because an increased incidence of nosocomial bacteremia due to gentamicin-resistant organisms occurred in our hospital, amikacin was substituted for gentamicin as the primary empiric aminoglycoside. Prospective surveillance of aminoglycoside use and of bacterial resistance to aminoglycosides was done before and after the substitution. We compared the baseline period when gentamicin was the primary aminoglycoside with the subsequent period when amikacin accounted for 81% of aminoglycoside used. During the two periods the patient population did not differ with regard to indications for aminoglycoside therapy. Among the gram-negative organisms isolated, rates of amikacin resistance during the baseline (1.1%) and usage (1.05%) periods were not significantly different (p greater than 0.6). Resistance to gentamicin and tobramycin during the period of amikacin use showed a statistically significant decrease (p greater than 0.001) due to decreased resistance to gentamicin of Providencia, Serratia, indole-positive Proteus, and decreased resistance to tobramycin of Pseudomonas aeruginosa. Unrestricted use of amikacin does not necessarily lead to increase in amikacin resistance but may lead to a decrease in resistance to gentamicin and tobramycin among gram-negative organisms.