One hundred thirty-eight patient courses of tobramycin (tobra) and gentamicin (genta) were prospectively monitored for ototoxicity using weekly audiograms and electronystag-mograms. Twice weekly drug serum levels and kidney function tests were determined. A pre, during, and post-therapy history was obtained and the results were analyzed to determine significant parameters of ototoxicity.
Statistical analysis of the data was performed using SPSS on an IBM 370/3033 computer. Tobra showed less toxicity than genta but only the difference in vestibular toxicity was statistically significant. Significant associations with toxicity included the patient developing a high temperature, total dose, low hematocrit for tobra, high hematocrit for genta, high creatinine clearance with cochlear toxicity, high creatinine with nephrotoxicity, poor condition or critically ill, and duration of therapy greater than 10 days. Non-significant parameters included dose rate (mg/kg), serum levels, age, prior noise exposure, prior aminoglycosides, prior ear infections, non-aminoglycoside ototoxic drugs, underlying disease, or total number of high risks present. Ototoxicity was independent of nephrotoxicity.