Ototoxicity Monitoring Threshold Shift Interpreter
Ototoxicity Monitoring Threshold Shift Interpreter
Apply ASHA’s three published limbs for a significant change in hearing during cochleotoxic drug therapy — 20 dB at one frequency, 10 dB at two adjacent frequencies, or loss of response at three consecutive frequencies — with the confirmation ASHA requires.
ASHA significant change
ASHA 1994, three limbsLargest single-frequency decrease 25 dB, worst adjacent pair 5 dB, no loss of response at three consecutive frequencies, change confirmed on retest
ASHA’s three limbs, and what each one is for
| Limb | ASHA’s wording | Why it is written that way |
|---|---|---|
| Single frequency | “20 dB decrease at any one test frequency” | A change large enough to stand on its own against test-retest variability |
| Two adjacent frequencies | “10 dB decrease at any two adjacent test frequencies” | Trades magnitude for corroboration: two neighbouring frequencies moving together is unlikely to be noise |
| Three consecutive frequencies | “loss of response at three consecutive test frequencies where responses were previously obtained” | For damage that has run past the audiometer’s output. ASHA notes it “refers specifically to the highest frequencies tested” |
| Confirmation, on every limb | “change must be confirmed by repeat testing”; retest “within 24 hours” | A single unreliable test can produce any of the three limbs on its own |
Baseline timing and test range, as ASHA specifies them
| Element | What ASHA asks for |
|---|---|
| Baseline, aminoglycosides | “prior to or within 72 hours of initial drug administration” |
| Baseline, cisplatin or carboplatin | “within 1 week prior to, and no later than 24 hours after, the initial treatment” |
| Minimum baseline test range | “bilateral pure-tone air-conduction thresholds at audiometric test frequencies from 0.25 to 8 kHz”, including 3 and 6 kHz |
| Above 8 kHz | “if possible, thresholds at frequencies above 8 kHz should also be measured”; the set “will normally include 9, 10, 11, 12, 14, 16, 18, and 20 kHz” |
| Why go above 8 kHz | “the available evidence indicates that high-frequency audiometry is the method of choice”; 9 to 20 kHz “provides maximum hearing sensitivity information”, though it “can be a lengthy procedure” |
Change from a baseline, not a grade — and the baseline is the problem
Ototoxicity monitoring asks a different question from every other page in this category. It is not interested in how much hearing a patient has; it is interested in whether it has changed since a reference point. ASHA’s 1994 guideline makes that explicit — change is “always computed relative to baseline measures” — and defines three limbs, any one of which counts as significant: a 20 dB decrease at one test frequency, a 10 dB decrease at two adjacent test frequencies, or loss of response at three consecutive test frequencies where responses were previously obtained. Change must then be confirmed by repeat testing.
The three limbs are not three ways of saying the same thing, and the second is the one most often forgotten. A 15 dB drop at one frequency meets nothing. A 10 dB drop at two neighbouring frequencies does. The criterion is trading magnitude against corroboration: one frequency moving a long way and two frequencies moving together are both hard to produce by chance, and a smaller change at two adjacent frequencies is stronger evidence than a middling change at one.
Two practicalities decide whether monitoring works at all. The first is the baseline, and it has to predate the drug: ASHA asks for it before or within 72 hours of the first aminoglycoside dose, and within a week before or 24 hours after the first cisplatin or carboplatin treatment. A baseline obtained in the second week of treatment has already absorbed part of the change it was created to detect, and the monitoring then measures the wrong interval. The second is the frequency range. ASHA states that high-frequency audiometry is the method of choice for early detection and that testing from 9 to 20 kHz gives the most information, while conceding that it is lengthy. Ototoxic damage starts at the top of the cochlea, so a conventional 0.25 to 8 kHz audiogram can be flat and normal while extended high-frequency thresholds have already shifted — and the loss-of-response limb cannot fire at all for frequencies that were never tested.
This is the point where audiometry meets the drug chart, which is why aminoglycosides and the platinums are the bridge between this category and the laboratory. Serum concentrations and cochlear concentrations do not move together: aminoglycosides are cleared from cochlear and vestibular tissue far more slowly than from plasma, so an acceptable trough does not mean the cochlea has cleared the drug, and cumulative exposure and treatment duration carry more of the risk than any single measurement. A criterion met is not a diagnosis of ototoxicity either: it is a change of a stated size, and middle ear disease, noise, wax, an unreliable test or the illness itself will produce the same arithmetic. Audiometric results are equipment- and calibration-dependent: the dB HL scale is defined against the reference equivalent threshold levels of the ISO 389 series for a particular transducer and coupler, so the same ear can read differently on two audiometers, and the BSA recommended procedure puts the uncertainty of any threshold measurement at no better than plus or minus 5 dB. Your own audiogram and the calibration standard it was recorded against govern, not this arithmetic.
Frequently asked questions
What are the ASHA criteria for a significant threshold shift?
A 20 dB decrease at any one test frequency, a 10 dB decrease at any two adjacent test frequencies, or loss of response at three consecutive test frequencies where responses were previously obtained — each computed relative to the patient’s own baseline, and each requiring confirmation by repeat testing.
Is a 15 dB drop at one frequency a significant shift?
No, under ASHA’s criteria. The single-frequency limb is 20 dB, and the 10 dB limb needs two adjacent frequencies to have both deteriorated. A 15 dB change at one frequency alone falls between the two limbs and meets neither, which is the commonest misreading of these rules.
Why does the baseline have to be before the drug?
Because the criteria measure change from it. ASHA asks for a baseline before or within 72 hours of the first aminoglycoside dose, and within a week before or 24 hours after the first cisplatin or carboplatin treatment. A later baseline has already absorbed some of the change it exists to detect, so the monitoring silently measures a shorter interval than it appears to.
Why test above 8000 Hz?
Because ototoxic change appears there first. ASHA states that high-frequency audiometry is the method of choice for early detection and that including 9 to 20 kHz gives the most information on hearing change, while noting it can be a lengthy procedure. A standard audiogram can be unchanged while extended high-frequency thresholds have already moved.
Does a confirmed shift mean the drug caused it?
No. The criteria describe a change of a stated size relative to a baseline and say nothing about cause. Middle ear disease, a wax plug, noise exposure, an unreliable test and the course of the underlying illness all produce the same numbers, which is part of why confirmation on retest is required.
Related calculators
References
- American Speech-Language-Hearing Association. Guidelines for the Audiologic Management of Individuals Receiving Cochleotoxic Drug Therapy, GL1994-00003. Rockville, MD: ASHA, 1994.
- Occupational Safety and Health Administration. 29 CFR 1910.95 — Occupational noise exposure. Washington, DC: US Department of Labor.
- British Society of Audiology. Recommended Procedure: Pure-tone air-conduction and bone-conduction threshold audiometry with and without masking, OD104-32. Reading: BSA, 2018.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/