Audiometric test criteria in the detection of cisplatin ototoxicity.

Audiometric criteria establishing the ototoxic effects of cisplatin (CDDP) vary widely in the literature. Normal test-retest variability in audiometric thresholds resulting from serial audiometry is likely a confounding variable in the over-reporting of ototoxic effects. This study investigated the variability of serial audiometry for a group of 21 head and neck cancer patients receiving no known ototoxic drugs in an effort to establish guidelines for determining reasonable audiometric criteria for CDDP therapy. Comparison of this group to a similar group of 31 patients receiving high dose (100+ mg/m2) bolus administration of CDDP suggests that audiometric shift criteria developed from Receiver Operating Characteristic analysis and employing multiple frequency averaging above 1000 Hz may lead to more universally accepted audiometric criteria for ototoxicity, which should in turn lead to more reliable reports of disease incidence.