Poor empiric dosing of aminoglycosides continues to be a problem in ma
ny hospitals. Consequently, a two-phase study was undertaken to evalua
te gentamicin and tobramycin (G-T) dosing in our institution and to de
velop practical guidelines to achieve optimum initial G-T serum concen
trations. In the pre-intervention phase (phase 1), approximately one-t
hird of 232 G-T treatment courses were initiated with regimens that re
sulted in low initial serum concentrations. Empiric dosing guidelines
were subsequently developed using Pharmacy Pharmacokinetic Service dat
a and literature recommendations. They were validated with patients in
phase 1. Compared with earlier dosing methods, the guidelines resulte
d in initial serum G-T concentrations within the therapeutic range in
an equivalent proportion of patients. To simplify the procedure, recom
mended dosages were based on age, total weight, and serum creatinine o
nly. The guidelines were tabular in format, and prescriber calculation
s were not required. Pocket and wall chart guidelines were distributed
and promoted. During the post-intervention phase (phase 2), the impac
t of the guidelines was assessed. A review of 232 phase 1 and 203 phas
e 2 treatment courses revealed an increase in the initial mean dose pe
r interval prescribed after the intervention (p<0.01). No change in th
e mean interval was noted. When guidelines were followed, a higher pro
portion of treatment courses achieved early therapeutic serum concentr
ations (CP(max) 5-10 mg/L, 80% vs 54%; p<0.005). The new empiric guide
lines appear to be a practical alternative to existing methods of dosi
ng aminoglycosides.