For most readers this lives in the background — you'll never get cisplatin, you'll never spend two weeks on IV gentamicin. For the readers it does touch, the difference between a baseline hearing test and no baseline, or speaking up about a new ringing ear on day three versus day thirty, is often the difference between full recovery and permanent loss. The questions to ask your prescriber are short and free. The downside of asking is nothing; the downside of not asking, for the unlucky ones, is permanent.
Sound gets turned into nerve signal by about 16,000 hair cells lining a coiled tube in your inner ear, the cochlea. Each cell is tuned to one pitch — the cells near the tube's entrance handle high frequencies, the ones deep inside handle low ones. Most ototoxic drugs come for the high-pitch cells first. The early sign is specific: you stop hearing the consonants in a conversation across a noisy room, or the smoke-alarm chirp from the floor below, before any standard hearing test would catch it.
Aminoglycoside antibiotics — gentamicin, tobramycin, amikacin, streptomycin, neomycin — and the chemotherapy drug cisplatin get pulled into those hair cells, poison the cells' energy machinery, and kill them outright (Huth et al. 2011). Human cochlear hair cells don't grow back. The damage often keeps progressing for weeks after the last dose because the drug lingers in inner-ear fluid long after it has cleared from your blood.
Loop diuretics like furosemide work by a different route. They knock out the chemical battery powering the cochlea — disrupting the ion gradient hair cells need to fire — without killing the cells (Ding et al. 2016). When the drug clears, hearing usually comes back. High-dose aspirin sticks to a motor protein in the same cells and stops them amplifying sound (Oliver et al. 2001); the cells stay alive, and any tinnitus from it fades within a day or two of stopping.
Two different mechanisms, two very different outcomes. That cellular difference is the dividing line between drugs you recover from and drugs you don't.
What the trials actually show
Aminoglycoside antibiotics cause measurable hearing loss in roughly 10–25% of patients who get a single course, climbing past 50% in people on repeated courses for cystic fibrosis or drug-resistant tuberculosis (Huth et al. 2011). About 1 in 500 people carry a mitochondrial variant called m.1555A>G that makes a single dose of gentamicin enough to render them profoundly deaf for life (McDermott et al. 2022). If anyone in your family lost their hearing after a hospital antibiotic, that's the variant worth getting tested for.
Cisplatin chemotherapy causes permanent hearing loss in 40–80% of treated adults and at least half of treated children. The protective drug story is the one bright spot in this whole entry.
The over-the-counter end of the list is more surprising. Two long prospective studies tracked about 27,000 men and 69,000 women for two decades each, watching who developed hearing loss or persistent tinnitus and who didn't. Regular users — taking aspirin, ibuprofen or naproxen, or acetaminophen twice a week or more — had roughly a 20% higher rate of new hearing loss and a similar bump in new persistent tinnitus among the women (Curhan et al. 2010) (Curhan et al. 2022). Low-dose aspirin (the cardiovascular dose, around 81 mg) did not show the signal. The signal is in the chronic-use pattern, not the occasional headache pill.
Loop diuretics (furosemide, bumetanide), high-dose IV vancomycin in kidney failure, macrolide antibiotics (erythromycin, azithromycin), and the quinine-family drugs (chloroquine, hydroxychloroquine) round out the list. In most of these cases the damage is reversible if you stop the drug early (Ding et al. 2016) (Ikeda et al. 2018) (Vanoverschelde et al. 2021). Chronic, multi-year hydroxychloroquine is the partial exception — case reports of permanent loss exist with cumulative high-dose use.
What missing it looks like
For the reader on a 10-day hospital course of IV gentamicin who never gets a baseline hearing test, the pattern often goes like this. Week one: you feel slightly off in the hallway, but you're sick in a hospital, so you write it off. Week two: you go home; conversation sounds fine, so you don't mention anything. Six months later you realise you can't follow your kid's voice in a busy restaurant. The audiologist says it's age and offers hearing aids. It isn't age, and the damage will not improve.
The version of this you really don't want to miss is the balance one. Gentamicin sometimes goes after the motion sensors in the inner ear instead of, or in addition to, the hearing cells. The symptom isn't vertigo — you don't spin. The world bounces when you turn your head. The ground feels uncertain on stairs. You start avoiding walks in the dark because you can't trust your footing. In the largest case series of this condition, 90% of patients had completely normal hearing tests, and the prescriber missed the diagnosis in nearly every case — patients showed up to a neurologist years later still wondering what had happened to them (Ahmed et al. 2012). Of 21 people who told the prescriber about imbalance during treatment, 20 were dismissed.
For the reader who takes two ibuprofen a day for chronic joint pain across a decade, the picture is quieter and more cumulative. A few percentage points of added hearing-loss and tinnitus risk every year. The kind of risk that's hard to feel in the moment — until the ringing starts and doesn't stop, and the audiogram shows the high-frequency notch that wasn't there a decade ago.
The downstream cost is the part nobody warns you about. Once tinnitus is permanent, it sits with you through every quiet moment — falling asleep takes longer, the alarm wakes you tired, you stop noticing songs because there's always a higher tone laid over them. High-frequency hearing loss makes conversations across a noisy room exhausting in a way that quietly drains your attention; people you used to enjoy talking with become harder to follow, and over years that shifts who you sit next to and how often you go out. The link between acquired hearing loss, tinnitus, and depression is real and well-documented. The drug-driven slice of all of that is the slice you can avoid.
The common thread is that the damage isn't dramatic in the way "drug allergy" or "side effect" usually means. There's no rash, no nausea, no obvious moment. It's just that one morning, after months or years, you notice the world sounds quieter, or moves differently, and you can't undo it.
What to actually ask for
If you're being prescribed any of the drugs above — or you spot one on a hospital chart — three short conversations make most of the difference.
For the over-the-counter side of the list, the action is different. Low-dose aspirin (the 81 mg cardiovascular dose) is fine — the cohort data is reassuring on this (Curhan et al. 2022). Daily ibuprofen, naproxen, or acetaminophen for chronic pain is the use pattern that picked up the signal. If that's you, the question for your doctor is whether there's a non-drug pain plan, a topical NSAID, or a different controlled-dose strategy that gets you the same relief.
What most guides get wrong
Over-the-counter doesn't mean ear-safe. Two ibuprofen a day for years is the picture that picked up the signal in the Curhan studies — not the occasional headache pill (Curhan et al. 2010). The risk is in the chronic pattern. Low-dose aspirin for heart-attack prevention is exempt; moderate-dose aspirin, ibuprofen, naproxen and acetaminophen are not.
Gentamicin damage doesn't always show on a hearing test. Most people with gentamicin damage to the balance organs have completely normal audiograms. The symptom is unsteady walking and the world bouncing when they turn their head, not deafness — and it's missed by almost everyone the first time (Ahmed et al. 2012). If you've had IV gentamicin and your balance feels different afterwards, ask specifically about bilateral vestibular hypofunction (Strupp et al. 2017).
Antibiotic ear drops are mostly a different story. Drops put into a healthy outer ear with an intact eardrum barely reach the inner ear and carry very low risk. The picture changes with a perforated eardrum, after ear surgery, or with tympanostomy tubes — in those cases an aminoglycoside drop can reach the inner ear directly, and a non-ototoxic alternative (a fluoroquinolone drop) is preferred.
Stopping the drug doesn't always reverse the damage. True for high-dose aspirin and most macrolides; usually true for loop diuretics. Not true for aminoglycosides — they continue killing hair cells for weeks after the last dose because the drug lingers in inner-ear fluid. Not true for cisplatin damage either, which is essentially permanent unless sodium thiosulfate was used (Brock et al. 2018).
Who carries the most risk
A few populations sit far above the median reader's risk and should treat the action list above as mandatory rather than optional.
- Anyone scheduled for cisplatin or carboplatin chemotherapy. Ask whether sodium thiosulfate is an option for your tumour type. Pediatric oncology has a clear path; adult oncology does not yet, but the question is still worth asking.
- People with cystic fibrosis, drug-resistant tuberculosis, or recurrent gram-negative infections. Anyone likely to receive repeated aminoglycoside courses across a lifetime — the cumulative dose is what drives the risk, not the individual course. Track yours.
- People with reduced kidney function on loop diuretics or vancomycin. Both clear through the kidneys; when clearance drops, levels in the inner ear rise.
- Older adults on long-term loop diuretics. Watch for new tinnitus after a dose increase or when an antibiotic gets added during a hospital admission.
- People on long-term hydroxychloroquine for lupus or rheumatoid arthritis. Permanent damage is rare but documented with cumulative high-dose use. A baseline hearing test at the start of therapy is reasonable, especially alongside the routine eye exams these patients already get.
- Families with unexplained deafness after a hospital antibiotic. The m.1555A>G mitochondrial variant runs through families. A one-time genetic test gives a lifetime answer for everyone who inherited it.
Related ground worth a look: noise-induced hearing loss (which stacks with drug damage in a hospital ICU), age-related hearing loss (which masks early ototoxic change), tinnitus management when it's already chronic, and Ménière's disease — where gentamicin injected directly into the ear is used deliberately as a treatment, not a side effect.
- — Ear-toxic drugs stack on top of age-related loss, pushing hearing past the line sooner.
- — New ringing in the ears is a classic sign a medication is ototoxic — worth reviewing the list.
- — A baseline audiogram before a risky drug, and another during it, is how the damage gets caught while it's still reversible.
- — An annual med review is where an ear-toxic drug you no longer need gets caught and dropped before it touches your hearing.
- — Fluoroquinolone antibiotics are the other drug class worth questioning — same move: ask whether something safer would do the job.
- — Like loud noise, these drugs kill hair cells that don't grow back — both are one-way damage worth preventing.
- — A drug hitting the inner ear can show up as hearing that drops fast — treat it as the same emergency until proven otherwise.
Substance + claimed effects
Ototoxic medications are prescribed and over-the-counter drugs with documented capacity to damage the cochlea, vestibular system, or both. The substance is not a single drug but a heterogeneous group spanning multiple therapeutic classes: aminoglycoside antibiotics (gentamicin, tobramycin, amikacin, streptomycin, neomycin, kanamycin), platinum-based chemotherapy (cisplatin, carboplatin at high dose), loop diuretics (furosemide, bumetanide, ethacrynic acid), salicylates and NSAIDs (aspirin, ibuprofen, naproxen), macrolide antibiotics (erythromycin, azithromycin, clarithromycin), glycopeptide antibiotics (vancomycin), and quinine derivatives (quinine, chloroquine, hydroxychloroquine). The claimed effects, scored across the entry's relevant dimensions, are: elevation of pure-tone hearing thresholds (most often beginning above 8 kHz and progressing downward); new-onset or potentiated tinnitus; bilateral vestibular hypofunction with oscillopsia and gait imbalance; and downstream effects on mood, sleep, and cognitive load that follow from chronic tinnitus or sensorineural hearing loss. Reversibility ranges from full resolution within hours of dose cessation (salicylates) to permanent, irreversible loss of hair cells and vestibular function (aminoglycosides, cisplatin). The entry covers the substance class holistically and the decision-points that affect whether reader-relevant damage occurs: drug choice, dose, monitoring, and substitution.
Evidence by addressing question
Mechanism
The mechanisms diverge by class and predict the reversibility profile.
Aminoglycosides are actively trafficked into outer hair cells (OHCs) of the cochlea and into vestibular hair cells via mechanotransduction channels and endocytosis, where they bind to mitochondrial 12S rRNA, suppress mitochondrial protein synthesis, generate reactive oxygen species, and trigger apoptosis (Huth et al. 2011). Damage starts in the basal turn of the cochlea — the high-frequency region — and progresses apically with continued exposure. Drug persists in inner-ear fluids for months after serum clearance, producing delayed-onset hearing loss up to 6 months post-treatment. Gentamicin and tobramycin show preferential vestibulotoxicity; amikacin and kanamycin are more cochleotoxic. Because human cochlear hair cells do not regenerate, loss is permanent.
Cisplatin enters OHCs via copper transporter Ctr1 and organic cation transporter OCT2, generates ROS, depletes glutathione, and triggers apoptosis preferentially in the basal turn. Damage is dose-cumulative and largely irreversible. Sodium thiosulfate scavenges cisplatin in plasma before it reaches the cochlea when administered 6 hours after the cisplatin infusion (Brock et al. 2018).
Loop diuretics inhibit the Na-K-2Cl cotransporter (NKCC1) in the stria vascularis, abolishing the endocochlear potential and causing stria edema. The mechanism is functional rather than cytotoxic, which explains the predominantly reversible profile; permanent loss occurs in renal failure (impaired clearance, accumulation), rapid IV bolus, and co-administration with aminoglycosides (Ding et al. 2016).
Salicylates act as competitive inhibitors of chloride at the anion-binding site of prestin, the OHC motor protein, abolishing electromotility without killing the cell (Oliver et al. 2001). Effect is dose-dependent, produces a flat ~20–40 dB sensorineural loss and tinnitus, and resolves within 24–72 hours of discontinuation.
NSAIDs are hypothesized to reduce cochlear blood flow via prostaglandin inhibition; the mechanism is less well-characterized than salicylates'.
Macrolides have unclear ototoxic mechanism; cases include both cochlear and central auditory effects, mostly reversible.
Quinine derivatives bind melanin (concentrated in stria vascularis) and persist; mechanism involves OHC dysfunction and spiral-ganglion sodium channel inhibition.
Evidence
Aminoglycoside cochleotoxicity is among the best-documented adverse drug reactions in medicine. Reported incidence varies widely (2–25% for hearing loss, up to 50% in cystic fibrosis and TB patients with repeated courses) due to differences in detection methods (Huth et al. 2011). With high-frequency audiometry (extended above 8 kHz), incidence rises to ~47%; with treatment courses of 6–12 months, approaches 100%. Approximately 10 million doses are consumed annually in the United States. The mitochondrial variant m.1555A>G (~1 in 500 of the general population) confers susceptibility to profound, sometimes single-dose, irreversible deafness (McDermott et al. 2022).
Gentamicin vestibulotoxicity is documented in a 23-year case series of 103 patients (later expanded to 127) presenting with imbalance and/or oscillopsia after gentamicin therapy. None had vertigo; 90% had no hearing loss; in 20 of 21 patients with intratreatment imbalance, the symptom was dismissed by the prescriber (Ahmed et al. 2012). Diagnosis depends on bilateral video head-impulse test and Bárány Society criteria for bilateral vestibulopathy (Strupp et al. 2017).
Cisplatin hearing loss occurs in 40–80% of treated adults and ≥50% of children, often progressing into the speech-frequency range with cumulative doses. The SIOPEL-6 phase 3 RCT (n=109) in pediatric hepatoblastoma demonstrated that sodium thiosulfate (20 g/m², infused 6 h after cisplatin) reduced grade ≥1 hearing loss from 63% to 33% (RR 0.52, 95% CI 0.33–0.81, p=0.002) with no detriment to overall or event-free survival at 3 years (Brock et al. 2018). The earlier ACCL0431 phase 3 RCT (n=125, COG) showed similar relative-risk reduction (Freyer et al. 2017). FDA approved sodium thiosulfate (Pedmark) on 20 Sept 2022 for pediatric localized non-metastatic solid tumors (FDA 2022).
NSAID and acetaminophen association with hearing loss and tinnitus rests on two large prospective cohorts by the Curhan group. In men (n=26,917, follow-up 18 years, 3,488 incident cases), regular (≥2×/week) use raised hearing-loss hazard by 12% for aspirin, 21% for NSAIDs, 22% for acetaminophen; the effect was substantially stronger in men under 50 (HR 1.33, 1.61, 1.99 respectively) and dose-duration dependent (Curhan et al. 2010). In women (NHSII, n=69,455, follow-up 22 years, 10,452 incident persistent tinnitus cases), frequent NSAID or acetaminophen use was associated with ~20% increased tinnitus risk; low-dose aspirin (cardiovascular dose) showed no elevated risk; moderate-dose aspirin was associated with tinnitus only in women under 60 (Curhan et al. 2022).
Salicylate ototoxicity is reversible: bilateral sensorineural threshold elevation of ~20–40 dB and tinnitus with several grams/day, resolving within 24–72 hours of discontinuation. Low-dose (≤100 mg/day cardiovascular) and standard analgesic doses (≤1–2 g/day) rarely produce audible effect; threshold for clinically apparent symptoms is typically >4 g/day in adults.
Loop diuretic ototoxicity is dose- and infusion-rate-dependent, presenting with tinnitus, hearing loss, or vertigo, and is usually reversible. Permanent loss has been reported with rapid IV bolus, renal failure, and concomitant aminoglycoside use (Ding et al. 2016).
Macrolides: A systematic review identified 78 audiometry-documented cases of macrolide-associated sensorineural hearing loss across 44 studies; in 92.3% hearing loss was reversible after cessation (Ikeda et al. 2018). Population data show a ~25% increased prevalent tinnitus risk associated with macrolide use (Vanoverschelde et al. 2021). Erythromycin shows the strongest signal at high IV doses in renal impairment.
Vancomycin alone at therapeutic levels has no conclusive monotherapy ototoxicity signal in modern data; ototoxicity is observed predominantly with high serum levels in renal failure or with concomitant aminoglycoside.
Quinine, chloroquine, hydroxychloroquine cause cinchonism — reversible high-frequency SNHL, tinnitus, and disequilibrium — at therapeutic doses; chronic high-cumulative-dose hydroxychloroquine for SLE/RA can produce irreversible loss in case reports, though large cohort data show no excess hearing loss in SLE attributable specifically to antimalarial use.
Protocol — monitoring and prescribing decisions
ASHA (1994) and AAA (2009) guidelines define ototoxicity for monitoring purposes as: 20+ dB threshold shift at one frequency, 10+ dB at two adjacent frequencies, or loss of response at three previously responsive frequencies (ASHA 1994). Recommended schedule: baseline audiogram within 72 hours of starting aminoglycoside (24 hours for cisplatin), then every 2–3 days or weekly during treatment, with follow-up up to 6 months after cessation because of delayed-onset effects. High-frequency audiometry (above 8 kHz, up to 16–20 kHz) detects damage before it reaches speech frequencies. Distortion-product otoacoustic emissions (DPOAEs) are an OHC-specific objective measure useful when behavioral audiometry is infeasible (NICU, sedated patients). Compliance with monitoring is poor in practice: a recent QI review found that fewer than half of high-risk aminoglycoside courses are paired with audiometric monitoring.
For pediatric cisplatin: NCCN guidelines recommend sodium thiosulfate co-administration in eligible (localized, non-metastatic) solid tumors (FDA 2022). For neonates requiring empiric aminoglycoside in suspected sepsis: the Genedrive MT-RNR1 point-of-care test (26-minute turnaround on a buccal swab) detects m.1555A>G and allows substitution of a cephalosporin without delay; NICE recommends conditional NHS adoption (NICE 2022). The PALOH implementation trial demonstrated 100% sensitivity and specificity preclinically and successful integration into emergency NICU workflows without delaying time-to-antibiotic (McDermott et al. 2022). WHO 2020 MDR-TB guidelines deprioritized kanamycin and capreomycin and restricted amikacin and streptomycin to second-line use after other options exhausted, citing ototoxicity (WHO 2020).
Contraindications and risk-stratification
Absolute or near-absolute contraindications to aminoglycoside therapy: confirmed maternal/personal m.1555A>G or m.1494C>T variant (single dose can cause profound deafness); pre-existing severe-to-profound sensorineural hearing loss in the better-hearing ear (cannot afford further loss); active vestibular disease with bilateral compromise. Relative contraindications and dose-modification triggers: renal impairment (CrCl <60), volume depletion, concomitant ototoxic exposure (loop diuretic, cisplatin, vancomycin, NSAIDs), age >65, prior aminoglycoside courses, cumulative dose >10 prior IV courses, family history of aminoglycoside-induced deafness, low birth weight, concurrent loud-noise exposure (synergistic damage). For cisplatin: pre-existing hearing loss should trigger discussion of carboplatin substitution where oncologically equivalent, dose reduction, or sodium thiosulfate eligibility.
Misconceptions
The dominant misconception is that over-the-counter ≠ safe for hearing. Aspirin, ibuprofen, naproxen, and acetaminophen are not benign at the dose-and-frequency levels reported in the Curhan cohorts: 2+ days per week, sustained over years, raises hearing-loss risk by ~20% and tinnitus risk by ~20% in women (Curhan et al. 2010, Curhan et al. 2022). A second is that gentamicin damage is universally announced by hearing loss — in fact 90% of gentamicin vestibulotoxicity patients have normal audiograms but bilateral vestibular failure that goes undiagnosed for years (Ahmed et al. 2012). A third is that topical ear drops are dangerous: aminoglycoside ototopical drops with an intact tympanic membrane have very low systemic absorption and minimal ototoxicity risk; risk arises with tympanic membrane perforation or tympanostomy tubes. A fourth is the assumption that "if hearing loss starts, you stop the drug and it reverses" — true for salicylates and most macrolides, but aminoglycosides continue to damage cells for weeks after cessation, and cisplatin damage is essentially permanent.
Audience subgroups
The highest-stakes populations: pediatric oncology patients on cisplatin-containing regimens; neonates and infants in intensive care receiving empiric aminoglycosides for sepsis; adults on long-term IV antibiotic courses for cystic fibrosis, MDR-TB, complicated infective endocarditis, or recurrent gram-negative bacteremia; dialysis and CKD patients on intermittent aminoglycoside or vancomycin; ICU patients on simultaneous loop diuretic + aminoglycoside; older adults on long-term loop diuretics with declining renal function; chronic-pain and rheumatology patients on long-term NSAIDs or hydroxychloroquine; patients with carrier status for m.1555A>G (often unknown unless tested).
Alternatives
By indication: for neonatal/community-acquired sepsis, cephalosporins are first-line substitutes for aminoglycosides when m.1555A>G is detected. For routine UTI and uncomplicated gram-negative infection, fluoroquinolones, cephalosporins, and beta-lactam/beta-lactamase inhibitors are alternatives. For chronic pain, the trade-off space includes acetaminophen at controlled dose, COX-2 inhibitors (incompletely studied for ototoxicity), topical NSAIDs (low systemic exposure), and non-pharmacologic options. For loop-diuretic-resistant heart failure or edema, thiazide diuretics, mineralocorticoid receptor antagonists, and SGLT2 inhibitors are non-ototoxic alternatives in some settings. For SLE/RA flare prevention, methotrexate and biologics carry their own risk profiles but no signal for ototoxicity.
Failure modes
The most common failure mode is not monitoring. Even when guidelines specify pre-, intra-, and post-treatment audiometry, fewer than half of qualifying aminoglycoside courses in published QI audits actually receive monitoring, attributable to lack of sound-booth access, no audiologist on call, and clinician unfamiliarity with the schedule. The second failure mode is dismissing tinnitus or imbalance as nonspecific: in the Ahmed gentamicin series, 20/21 patients reporting in-treatment imbalance were ignored by prescribers; symptoms attributable to other causes (anxiety, infection, fatigue) delay drug-stoppage decisions (Ahmed et al. 2012). A third is cumulative-dose blindness: each individual course may be modest, but cumulative aminoglycoside courses >10 strongly predict ototoxicity. A fourth is polypharmacy stacking — co-administration of aminoglycoside + loop diuretic + NSAID is common in ICU and produces synergistic toxicity that monotherapy data underestimates.
Practicalities
High-frequency audiometry equipment (extended to 16–20 kHz) costs roughly $15–30k for a clinic; sound booths $10–20k. In the US, ototoxicity monitoring services are often billed under audiology evaluation codes; coverage is variable. The Genedrive MT-RNR1 test runs ~£100/test in the NHS; cost-effectiveness analysis supports adoption in the neonatal sepsis pathway (NICE 2022). Sodium thiosulfate (Pedmark) is FDA-approved for pediatric oncology and is reimbursed under standard cancer-care benefits in the US.
Stakes and reversibility
Reversibility map: salicylates, low-dose NSAIDs, macrolides (most cases), loop diuretics (most cases), quinine cinchonism — typically reversible within hours to weeks of discontinuation. Aminoglycosides — typically irreversible cochlear and vestibular damage; vestibular function can occasionally show partial recovery through central compensation but hair cells do not regenerate. Cisplatin — essentially always permanent. Chronic high-cumulative-dose hydroxychloroquine — case reports of irreversible damage. The reversibility profile is the single most important fact for the reader: a reversible drug causing tinnitus is a signal to discuss the drug; an irreversible drug causing tinnitus is a signal that damage has already accrued.
Out-of-scope
Recreational drugs (alcohol, MDMA, recreational opioids — separate ototoxicity profile); occupational and recreational noise exposure (a separate substance even though it interacts synergistically with aminoglycosides); age-related and noise-induced hearing loss; idiopathic sudden sensorineural hearing loss; Ménière's disease (intratympanic gentamicin is a treatment, not a side-effect, in this entry's scope); cochlear implant indications.
The credibility range
The optimist case. The major classes of ototoxic medications are extensively characterized; for the high-stakes drugs (aminoglycosides, cisplatin), monitoring guidelines have existed since 1994 (ASHA 1994) and the mechanism is understood to the molecular level. Genetic susceptibility (m.1555A>G) can now be screened in 26 minutes at the bedside, and FDA-approved otoprotection (sodium thiosulfate) cuts pediatric cisplatin hearing loss by ~50% without compromising survival (Brock et al. 2018, FDA 2022). For OTC analgesics, the absolute risk increase is modest (~20% relative, on a low base rate) and the protective alternative — low-dose aspirin, controlled-dose acetaminophen, non-drug analgesics — is accessible. Most ototoxic exposures the average reader encounters (acetaminophen for headache, occasional ibuprofen, a 5-day course of macrolide for community-acquired infection) carry negligible permanent-damage risk. The strong case is that ototoxicity is a solved problem in principle and the gap is implementation: ask the right questions, monitor when indicated, substitute when possible.
The skeptic case. Monitoring guideline compliance is low in practice. Adults on intensive-care aminoglycosides typically do not receive baseline audiograms. Gentamicin vestibulotoxicity is missed by prescribers in >90% of cases per the largest case series, with patients receiving the correct diagnosis years later from a neurologist (Ahmed et al. 2012). The m.1555A>G point-of-care test is available only in pilot NHS neonatal units; in most settings worldwide, the genetic risk is unscreened. Sodium thiosulfate is FDA-approved only for a narrow pediatric indication; adult oncology patients have no comparable otoprotection. NSAID/acetaminophen ototoxicity has been demonstrated in large prospective cohorts but mechanism is poorly characterized and absolute risk is small; the population implication of dialing back chronic-use analgesics across the catalogue is unclear. The skeptic position: the problem is named and the science exists, but the system fails to act on it, and most readers cannot get monitoring even if they ask for it.
Author's call. The evidence base on aminoglycoside, cisplatin, and salicylate ototoxicity is mature and high-grade — this is an evidence-4 substance, not contested. Controversy is moderate: the field debates monitoring frequency, NSAID-attributable risk at typical OTC doses, and the threshold for substituting alternatives. The reader's relevant action is awareness: most readers won't get cisplatin, but many will encounter loop diuretics, NSAIDs, or aminoglycoside antibiotics over a lifetime, and the difference between irreversible damage and full recovery often comes down to recognizing in-treatment tinnitus and pushing on the prescriber. The entry should treat the genetic, pediatric, and high-dose long-course populations as the small-but-life-altering tail and the OTC analgesic population as the broader, lower-grade signal worth knowing about.
Stakeholder + incentive map
- Pharmaceutical manufacturers (aminoglycoside, cisplatin, loop diuretic generics): low-margin generics; little incentive to fund post-marketing safety surveillance. Sodium thiosulfate (Fennec) has strong incentive to expand label.
- Hospital pharmacy and antimicrobial stewardship: aligned with substitution where evidence supports (NICE Genedrive, MDR-TB second-line restriction).
- Audiology profession (ASHA, AAA): aligned with formal monitoring schedules; capacity-constrained.
- Oncology (NCCN, COG, SIOPEL): driving force behind cisplatin otoprotection trials; aligned with sodium thiosulfate in pediatrics.
- Tuberculosis programs: historically forced into aminoglycoside use for MDR-TB; WHO 2020 update aligned with substitution where bedaquiline/linezolid available (WHO 2020).
- OTC analgesic industry: commercial counter-incentive to publicize cumulative-use risk; labels do not warn of hearing-loss signal at chronic-use doses.
- Patient advocacy (cystic fibrosis, hereditary deafness communities, Hearing Loss Association of America): push for genetic screening, monitoring access, and informed-consent improvement.
Population variability
- Genetic: ~1 in 500 carry m.1555A>G mitochondrial variant; penetrance varies with heteroplasmy and modifier variants (m.4394C>T enhances). Some carriers develop profound deafness from a single dose; others tolerate aminoglycosides without effect.
- Age: neonates have higher cochlear and renal vulnerability; older adults have reduced renal clearance and baseline hearing loss that masks early changes; pediatric oncology patients face longer remaining-lifespan disability.
- Sex: NHSII showed female susceptibility to NSAID tinnitus; Curhan men's cohort showed strongest analgesic-hearing-loss association under age 50.
- Renal function: CrCl <60 substantially elevates aminoglycoside, loop diuretic, vancomycin, and macrolide ototoxicity through delayed clearance and accumulation.
- Baseline hearing: any pre-existing SNHL elevates both the absolute disability from added loss and the difficulty of detecting incremental change.
- Disease-specific risk: cystic fibrosis, MDR-TB, recurrent bacteremia, and standard-risk hepatoblastoma populations face cumulative exposure orders of magnitude above the median patient.
- Co-exposure: concurrent noise exposure synergistically increases aminoglycoside uptake into hair cells; combining loop diuretic + aminoglycoside multiplies risk.
Knowledge gaps
- Adult cisplatin otoprotection: sodium thiosulfate is approved only for localized pediatric tumors; adult metastatic populations have no equivalent.
- Mechanism and absolute risk of OTC NSAID/acetaminophen ototoxicity at typical use frequencies — Curhan cohorts establish association but RCT data are absent.
- Whether universal m.1555A>G screening before any aminoglycoside exposure (not just neonatal) is cost-effective in adult populations.
- Vancomycin monotherapy ototoxicity at modern therapeutic drug monitoring targets — older case series may not reflect current dosing.
- Long-term hydroxychloroquine ototoxicity dose-response in modern rheumatology cohorts.
- Whether extended high-frequency audiometry should be standard before any course of loop diuretic in CKD patients.
- Reliable biomarkers for vestibular damage during treatment (analogous to OAEs for cochlear).
Scope vs. brief. The brief named NSAIDs, loop diuretics, and aminoglycosides explicitly. The article covers all three and additionally treats cisplatin (the most-studied ototoxic and the one with an FDA-approved otoprotectant), salicylates (the mechanism story that makes "reversible" make sense), macrolides, vancomycin, and the quinine family. The brief's "including" was read as non-exhaustive. Splitting any of these into their own entry would fragment a class that readers need to think about as a class — the dividing line in the article (reversible vs irreversible) only works when you can compare drugs side by side.
Combined addressing sections. Collapsed protocol and contraindications into one section because the reader's action list integrates both ("ask, monitor, report, plus the hard contraindication is m.1555A>G"). Splitting them would have created two short sections with overlapping action callouts.
Rating difficulty — the know problem. This is an awareness entry, not a habit. The meta dimensions reflect what the awareness produces for the affected subset (catching reversible damage early, avoiding chronic-OTC tinnitus pattern, preventing missed gentamicin vestibulotoxicity). Most readers will never encounter cisplatin or IV gentamicin; for the small subset that does, the value is enormous. Settled on modest 1s and 2s across health_short_term, focus, sleep, mood rather than higher scores — the impact is concentrated in a tail, not the median reader.
Evidence score held at 4, not 5. Aminoglycoside/cisplatin/salicylate data is RCT- and guideline-grade. The OTC NSAID/acetaminophen association rests on two very large prospective cohorts (Curhan 2010; Curhan 2022) with consistent direction but no RCT confirmation and small absolute effect. Treating that as "level 5 settled" would overstate the OTC piece.
What was excluded and why.
- Recreational drugs (alcohol, MDMA, recreational opioids) — separate substance and category (mental, mindset).
- Occupational and recreational noise — its own entry; flagged in
out-of-scope. - Intratympanic gentamicin for Ménière's disease — the same drug is used deliberately as a treatment; belongs in a Ménière's entry.
- Sudden idiopathic SNHL, age-related hearing loss, tinnitus management as a standalone topic — separate entries.
- Adult cisplatin otoprotection — no approved option yet; mentioned in passing but not built out, because there's nothing actionable to tell adults yet beyond "ask anyway."
Future links to wire when entries exist. Tinnitus (standalone management), noise-induced hearing loss, presbycusis / age-related hearing loss, audiometric screening, hearing aids, Ménière's disease, cystic fibrosis care, MDR-TB treatment regimens.
Separate-entry candidates surfaced while writing. "Bilateral vestibular hypofunction" deserves its own short entry under hearing or msk-conditions — it's underdiagnosed enough that pointing readers to it directly would help, and the gentamicin pathway is only one cause. "m.1555A>G pharmacogenetic testing" likewise — the test exists, costs little, has lifetime value, and currently lives buried inside this entry.
Hard editorial calls. Whether to lead with cisplatin (most dramatic, narrowest reader applicability) or with NSAIDs (less dramatic, broadest reader applicability). Chose the cellular dividing line — reversible vs irreversible — as the spine, because that's the fact a reader actually uses when deciding whether a new ringing matters. Cisplatin and NSAIDs both fall out of that frame naturally.
Ototoxic Medications
Reader's action is asking the prescriber two or three questions, requesting a baseline audiogram before a high-risk course, and reporting in-treatment tinnitus or imbalance. Time cost is minutes per encounter.
Aminoglycoside, cisplatin, and salicylate ototoxicity are characterised at the molecular level with longstanding monitoring guidelines (ASHA 1994; AAA 2009) and pivotal RCTs (Brock et al. 2018 NEJM; Freyer et al. 2017 Lancet Oncol). NSAID/acetaminophen association is established in large prospective cohorts (Curhan 2010, 2022) but lacks RCT confirmation, holding the entry below 5.
For readers prescribed high-risk regimens (aminoglycosides, cisplatin, high-dose loop diuretics), awareness enables early symptom reporting and dose adjustment that prevents permanent hair-cell loss (Brock et al. 2018; ASHA 1994). For chronic OTC analgesic users, the Curhan prospective cohorts (2010, 2022) suggest modest absolute risk reduction available by controlling use frequency.
Chronic tinnitus and high-frequency sensorineural hearing loss impose measurable cognitive and attentional load; preventing drug-related onset preserves bandwidth. Meaningful only for the subset of readers actually exposed to relevant agents, hence a 1.
Tinnitus is among the most-cited drug side-effect contributors to insomnia (Curhan et al. 2022); preventing or reversing drug-induced tinnitus removes a known sleep disruptor. Scope is limited to the exposed subset.
Acquired tinnitus and SNHL show consistent associations with depression, anxiety, and social isolation in cohort data. The drug-induced slice is small in absolute terms but real for affected readers.