Tuberculosis (TB) remains a major communicable disease requiring prolonged multidrug therapy. Although antitubercular therapy (ATT) is essential for achieving microbiological cure and preventing disease transmission, some drugs used particularly in drug-resistant TB regimens may produce clinically important adverse effects. Ototoxicity is among the most disabling complications associated with injectable aminoglycosides, including amikacin, kanamycin and streptomycin. These agents may accumulate within the inner ear and damage cochlear hair cells, producing bilateral, symmetrical and predominantly sensorineural hearing loss. The impairment usually begins at higher frequencies and may progressively involve speech frequencies with continued exposure. It is often irreversible and may be accompanied by tinnitus, vertigo or impaired balance.^[1,2] The risk of ototoxicity may be influenced by the specific medication, cumulative dose, duration of exposure, increasing age, renal impairment, concurrent use of other ototoxic drugs, previous noise exposure and pre-existing auditory dysfunction. A systematic review by Dillard et al. reported a pooled prevalence of approximately 41% for aminoglycoside-induced hearing loss among patients treated for drug-resistant TB, demonstrating the considerable magnitude of this preventable treatment-related disability.^[3] Wangchuk et al. similarly documented hearing loss in 45.2% of patients undergoing multidrug-resistant TB treatment, highlighting the importance of systematic audiological surveillance.^[4] Pure-tone audiometry is a widely accepted method for detecting and quantifying hearing impairment. Conventional audiometry evaluates frequencies important for speech perception, while extended high-frequency audiometry, where available, may identify cochlear changes before conventional frequencies are affected. Otoscopy and impedance audiometry are also useful for excluding external- and middle-ear disorders that may cause conductive hearing loss. Baseline testing before initiating an ototoxic medicine, followed by periodic monitoring during treatment, enables early recognition of threshold deterioration and facilitates clinical decisions regarding dose modification, drug substitution and auditory rehabilitation.^[2,5] Current WHO recommendations favour effective all-oral regimens for eligible drug-resistant TB patients, thereby reducing exposure to injectable ototoxic medicines; nevertheless, patients with current or previous exposure to such drugs remain at risk.^[1] Hearing impairment can adversely affect communication, employment, social participation, mental health and adherence to treatment. Assessment of its prevalence, severity and associated clinical factors among patients receiving ATT is therefore important for strengthening pharmacovigilance and integrating audiological monitoring into TB services. The present study was undertaken to assess hearing impairment among patients receiving antitubercular therapy and to examine its relationship with treatment and patient-related characteristics.

AIM

To assess hearing impairment among patients receiving antitubercular therapy.

OBJECTIVES

1. To determine the prevalence, type and severity of hearing impairment among patients receiving antitubercular therapy.
2. To examine the association of hearing impairment with demographic characteristics, comorbidities, treatment regimen, ototoxic-drug exposure and duration of therapy.

Source of Data

The study participants were recruited from patients with diagnosed tuberculosis who were receiving antitubercular therapy and attending the designated TB centre, respiratory medicine outpatient department or inpatient services of the study institution during the study period. Treatment records, laboratory reports, prescriptions and audiological findings constituted the principal sources of data.

Study Design

A hospital-based cross-sectional analytical study was conducted.

Study Location

The Department of Otorhinolaryngology and Audiology of a tertiary-care teaching hospital.

Study Duration

The study was conducted over a period of 12 months, Participant recruitment, clinical evaluation, review of treatment records and audiological assessment were completed during this period.

Sample Size

A total of 120 eligible patients receiving antitubercular therapy were included. Participants were enrolled consecutively until the required sample size was attained.

Inclusion Criteria

• Patients with microbiologically or clinically diagnosed pulmonary or extrapulmonary TB.
• Patients who were currently receiving ATT for drug-sensitive or drug-resistant TB.
• Patients aged 18 years or older.
• Patients who had received ATT for at least two weeks at the time of assessment.
• Patients who provided written informed consent.
• Patients who were able to understand and respond reliably during audiological testing.

Exclusion Criteria

• Patients with documented congenital or hereditary hearing loss.
• Patients with active otitis externa, acute otitis media, tympanic-membrane perforation or another condition producing temporary conductive hearing loss.
• Patients with a history of major ear surgery, head injury involving the temporal bone or acoustic trauma.
• Patients with occupational noise exposure severe enough to represent an alternative primary cause of hearing loss.
• Patients receiving chemotherapy or other established ototoxic medicines unrelated to ATT, where their effect could not be separated from ATT exposure.
• Patients who were critically ill, uncooperative or unable to complete pure-tone audiometry.
• Patients who declined consent.

Procedure and Methodology

Approval was obtained from the Institutional Ethics Committee before commencing the study. Eligible patients were approached consecutively, informed about the study and enrolled after written informed consent had been obtained. Each participant was interviewed using a predesigned and pretested data-collection form.

Information was obtained regarding age, sex, residence, occupation, socioeconomic characteristics, smoking, alcohol consumption, previous noise exposure, history of ear disease, tinnitus, vertigo, subjective hearing difficulty and previous exposure to ototoxic medicines. Clinical information included type and site of TB, drug-susceptibility status, category of treatment, current ATT regimen, individual drugs received, duration of treatment, injectable-drug exposure, cumulative duration of such exposure, previous TB treatment and relevant comorbidities such as diabetes mellitus, hypertension, HIV infection and renal disease.

Treatment details were verified from treatment cards, prescriptions and hospital records. Participants were classified according to whether they had received an ototoxic injectable antitubercular drug such as amikacin, kanamycin or streptomycin. Current all-oral and injectable-containing regimens were analysed separately.

A general clinical and otorhinolaryngological examination was performed. Otoscopy was undertaken to assess the external auditory canal and tympanic membrane. Tuning-fork tests, including Rinne and Weber tests, were performed as preliminary assessments. Tympanometry was undertaken where indicated to exclude middle-ear pathology.

Pure-tone audiometry was conducted by a trained audiologist in a sound-treated room using a calibrated clinical audiometer. Air-conduction thresholds were determined separately for each ear at 250, 500, 1,000, 2,000, 4,000 and 8,000 Hz. Bone-conduction thresholds were measured at 500-4,000 Hz whenever air-conduction thresholds were elevated. Extended high-frequency thresholds above 8,000 Hz were assessed where the equipment permitted.

The pure-tone average was calculated using hearing thresholds at 500, 1,000, 2,000 and 4,000 Hz. Hearing impairment was defined as a pure-tone average of 20 dB hearing level or greater in the better-hearing ear. It was classified as sensorineural when air- and bone-conduction thresholds were elevated without a clinically significant air-bone gap, conductive when an air-bone gap was present with relatively preserved bone conduction and mixed when both components occurred. The degree of impairment was categorized as mild, moderate, moderately severe, severe or profound according to the audiological classification adopted by the study institution.

Because this was a cross-sectional assessment, an ATT-attributable threshold shift could not be established in participants without pretreatment audiograms. Therefore, the primary outcome was reported as prevalent hearing impairment during ATT, while its possible association with ototoxic exposure was assessed analytically without assuming causation.

Sample Processing

No separate biological specimen was collected specifically for the study. Recent renal-function reports, including serum creatinine and estimated glomerular filtration rate, were retrieved from the patient records because impaired renal clearance could increase aminoglycoside exposure. When a recent report was unavailable and testing was clinically indicated, a venous blood sample was collected under aseptic precautions, transported to the biochemistry laboratory and analysed according to the institution’s standard operating procedures. Audiological data were recorded directly from the calibrated audiometer and entered into the study form.

Data Collection

Data were collected using a predesigned, structured and pretested case-record form. The form included:

• Sociodemographic characteristics.
• TB type, site and drug-resistance status.
• Current and previous ATT regimens.
• Exposure to individual potentially ototoxic medicines.
• Treatment duration and relevant cumulative exposure.
• Comorbidities and other risk factors for hearing loss.
• Ear-related symptoms and otoscopic findings.
• Tympanometry and pure-tone audiometry results.
• Type, laterality and severity of hearing impairment.

Each completed form was checked for accuracy and completeness. Participants were assigned unique identification numbers, and personally identifiable information was kept confidential. Data were entered into an electronic database and verified before analysis.

Statistical Methods

Data were analysed using SPSS/R or an equivalent statistical package. Categorical variables were summarized as frequencies and percentages, while continuous variables were expressed as mean with standard deviation or median with interquartile range according to their distribution. The prevalence of hearing impairment was reported with a 95% confidence interval.

The association of hearing impairment with categorical predictors was assessed using the chi-square test or Fisher’s exact test. Continuous variables were compared using the independent-samples t test or Mann-Whitney U test, as appropriate. Effect estimates were presented as odds ratios with 95% confidence intervals. Variables that were clinically important or showed a univariable association at p<0.20 were considered for multivariable binary logistic-regression analysis to identify factors independently associated with hearing impairment. Age, sex, ototoxic-drug exposure, treatment duration, previous ATT, renal dysfunction, diabetes and other relevant confounders were considered in the adjusted model. Multicollinearity and model fit were assessed before interpreting the model. All tests were two-tailed, and p<0.05 was considered statistically significant.

Table 1: Overall assessment of hearing status among patients receiving antitubercular therapy (N=120) Variable Total (N=120), Mean (SD) or n (%) Hearing impairment (n=43) No hearing impairment (n=77) Effect estimate (95% CI) Test of significance P value Age, years 43.7 (13.2) 49.1 (12.6) 40.7 (12.7) MD=8.4 years (3.6-13.2) Independent t=3.48 0.001* Male sex 76 (63.3) 29 (67.4) 47 (61.0) OR=1.32 (0.60-2.90) χ²=0.49 0.485 Duration of ATT, months 5.8 (3.4) 7.0 (3.5) 5.1 (3.1) MD=1.9 months (0.7-3.1) Independent t=3.07 0.003* Drug-resistant tuberculosis 44 (36.7) 26 (60.5) 18 (23.4) OR=5.01 (2.24-11.24) χ²=16.34 <0.001* Exposure to an ototoxic antitubercular drug 51 (42.5) 31 (72.1) 20 (26.0) OR=7.36 (3.18-17.03) χ²=24.01 <0.001* Subjective hearing difficulty 37 (30.8) 33 (76.7) 4 (5.2) OR=60.23 (18.47-196.43) χ²=67.18 <0.001* Tinnitus 29 (24.2) 21 (48.8) 8 (10.4) OR=8.24 (3.22-21.09) χ²=23.07 <0.001* Vertigo or imbalance 11 (9.2) 7 (16.3) 4 (5.2) OR=3.55 (0.97-12.96) Fisher’s exact test 0.052 Pure-tone average in better ear, dB HL 24.6 (17.8) 47.8 (14.3) 11.6 (4.7) MD=36.2 dB (31.7-40.7) Welch’s t=15.97 <0.001* Audiometrically detected hearing impairment 43 (35.8) [95% CI: 27.8%-44.7%] 43 (100.0) 0 (0.0) Prevalence=35.8% (27.8%-44.7%) Descriptive outcome - Among the 120 patients receiving antitubercular therapy, the mean age was 43.7±13.2 years, and 76 (63.3%) were male. Audiometrically detected hearing impairment was present in 43 patients, giving an overall prevalence of 35.8% (95% CI: 27.8%-44.7%). Patients with hearing impairment were significantly older than those without impairment (49.1±12.6 versus 40.7±12.7 years), with a mean difference of 8.4 years (95% CI: 3.6-13.2; t=3.48, p=0.001). Male sex was slightly more common in the hearing-impaired group than in the unaffected group (67.4% versus 61.0%); however, the association was not statistically significant (OR=1.32, 95% CI: 0.60-2.90; p=0.485). The mean duration of ATT was significantly longer among patients with hearing impairment (7.0±3.5 months) than among those without impairment (5.1±3.1 months), with a mean difference of 1.9 months (95% CI: 0.7-3.1; t=3.07, p=0.003). Drug-resistant TB was present in 60.5% of patients with hearing impairment compared with 23.4% of those without impairment and was associated with fivefold higher odds of hearing impairment (OR=5.01, 95% CI: 2.24-11.24; p<0.001). Similarly, 72.1% of affected patients had been exposed to an ototoxic antitubercular drug compared with 26.0% of unaffected patients, resulting in significantly higher odds of impairment (OR=7.36, 95% CI: 3.18-17.03; p<0.001). Subjective hearing difficulty was reported by 37 (30.8%) patients and demonstrated a very strong association with audiometrically confirmed impairment (OR=60.23, 95% CI: 18.47-196.43; p<0.001). Tinnitus was reported by 29 (24.2%) patients and was significantly more frequent among those with hearing impairment than among those with normal hearing (48.8% versus 10.4%; OR=8.24, 95% CI: 3.22-21.09; p<0.001). Vertigo or imbalance was reported by 11 (9.2%) patients and was more frequent in the hearing-impaired group (16.3% versus 5.2%), although the association narrowly missed statistical significance (OR=3.55, 95% CI: 0.97-12.96; p=0.052). The mean pure-tone average in the better ear was significantly higher among patients with hearing impairment (47.8±14.3 dB HL) than among those without impairment (11.6±4.7 dB HL), with a mean difference of 36.2 dB (95% CI: 31.7-40.7; Welch’s t=15.97, p<0.001). Table 2: Prevalence, type and severity of hearing impairment among patients receiving antitubercular therapy (N=120) Audiological outcome n (%) 95% CI Test of significance P value Overall hearing status Normal hearing 77 (64.2) 55.3%-72.2% - - Hearing impairment 43 (35.8) 27.8%-44.7% - - Laterality among participants with hearing impairment (n=43) Bilateral impairment 31 (72.1) 57.3%-83.3% χ²=8.40† 0.004* Unilateral impairment 12 (27.9) 16.7%-42.7% Type of impairment among affected participants (n=43) Sensorineural hearing loss 34 (79.1) 64.8%-88.6% χ²=40.79‡ <0.001* Conductive hearing loss 6 (14.0) 6.6%-27.3% Mixed hearing loss 3 (7.0) 2.4%-18.6% Severity among affected participants (n=43) Mild 18 (41.9) 28.4%-56.7% χ²=22.00§ <0.001* Moderate 13 (30.2) 18.6%-45.1% Moderately severe 7 (16.3) 8.1%-30.0% Severe 4 (9.3) 3.7%-21.6% Profound 1 (2.3) 0.4%-12.1% Frequency pattern among affected participants (n=43) Predominantly high-frequency loss 27 (62.8) 47.9%-75.7% χ²=5.63† 0.018* Flat or speech-frequency involvement 16 (37.2) 24.3%-52.1% †Chi-square goodness-of-fit test comparing two categories with equal expected proportions. ‡Goodness-of-fit test comparing the three types with equal expected proportions. §Goodness-of-fit test comparing the five severity categories with equal expected proportions. Of the 120 patients, 77 (64.2%; 95% CI: 55.3%-72.2%) had normal hearing, whereas 43 (35.8%; 95% CI: 27.8%-44.7%) had audiometrically detected hearing impairment. Among the 43 affected patients, bilateral hearing impairment was observed in 31 (72.1%; 95% CI: 57.3%-83.3%), while unilateral impairment was found in 12 (27.9%; 95% CI: 16.7%-42.7%). Bilateral impairment was significantly more frequent than unilateral impairment (χ²=8.40, p=0.004). Sensorineural hearing loss was the predominant type, affecting 34 (79.1%; 95% CI: 64.8%-88.6%) patients. Conductive and mixed hearing loss were identified in 6 (14.0%; 95% CI: 6.6%-27.3%) and 3 (7.0%; 95% CI: 2.4%-18.6%) patients, respectively. The distribution of the three types differed significantly, with a clear predominance of sensorineural impairment (χ²=40.79, p<0.001). Regarding severity, 18 (41.9%) patients had mild, 13 (30.2%) had moderate, 7 (16.3%) had moderately severe, 4 (9.3%) had severe and 1 (2.3%) had profound hearing impairment. The severity categories showed a statistically significant unequal distribution (χ²=22.00, p<0.001), with mild and moderate impairment together accounting for 72.1% of cases. A predominantly high-frequency pattern was found in 27 (62.8%; 95% CI: 47.9%-75.7%) affected patients, whereas 16 (37.2%; 95% CI: 24.3%-52.1%) had flat or speech-frequency involvement. High-frequency hearing loss was significantly more frequent (χ²=5.63, p=0.018), supporting the characteristic audiological pattern of drug-related ototoxicity. Table 3: Factors associated with hearing impairment among patients receiving antitubercular therapy (N=120) Factor Category Hearing impairment (n=43), n (%) No hearing impairment (n=77), n (%) Odds ratio (95% CI) Test of significance P value Age ≥50 years 24 (55.8) 22 (28.6) 3.16 (1.45-6.88) χ²=8.66 0.003* <50 years 19 (44.2) 55 (71.4) Reference Sex Male 29 (67.4) 47 (61.0) 1.32 (0.60-2.90) χ²=0.49 0.485 Female 14 (32.6) 30 (39.0) Reference Diabetes mellitus Present 16 (37.2) 14 (18.2) 2.67 (1.14-6.22) χ²=5.33 0.021* Absent 27 (62.8) 63 (81.8) Reference Renal dysfunction Present 9 (20.9) 6 (7.8) 3.13 (1.03-9.51) χ²=4.35 0.037* Absent 34 (79.1) 71 (92.2) Reference Treatment category Drug-resistant TB 26 (60.5) 18 (23.4) 5.01 (2.24-11.24) χ²=16.34 <0.001* Drug-sensitive TB 17 (39.5) 59 (76.6) Reference Ototoxic-drug exposure Exposed 31 (72.1) 20 (26.0) 7.36 (3.18-17.03) χ²=24.01 <0.001* Not exposed 12 (27.9) 57 (74.0) Reference Duration of ATT ≥6 months 28 (65.1) 27 (35.1) 3.46 (1.58-7.56) χ²=10.04 0.002* <6 months 15 (34.9) 50 (64.9) Reference Previous ATT Present 21 (48.8) 19 (24.7) 2.91 (1.32-6.43) χ²=7.25 0.007* Absent 22 (51.2) 58 (75.3) Reference Statistically significant at p<0.05. Percentages in the hearing-status columns were calculated using their respective column totals. Hearing impairment was significantly associated with several demographic, clinical and treatment-related factors. Among affected patients, 55.8% were aged ≥50 years compared with 28.6% of those without impairment. Patients aged ≥50 years had more than threefold higher odds of hearing impairment than younger patients (OR=3.16, 95% CI: 1.45-6.88; χ²=8.66, p=0.003). Male sex was not significantly associated with hearing impairment (OR=1.32, 95% CI: 0.60-2.90; p=0.485). Diabetes mellitus was present in 37.2% of affected patients compared with 18.2% of unaffected patients and was associated with significantly increased odds of impairment (OR=2.67, 95% CI: 1.14-6.22; p=0.021). Renal dysfunction was also more frequent in the hearing-impaired group (20.9% versus 7.8%) and was associated with approximately threefold higher odds of hearing impairment (OR=3.13, 95% CI: 1.03-9.51; p=0.037). Drug-resistant TB was present in 60.5% of patients with hearing impairment compared with 23.4% of patients without impairment. It was associated with approximately fivefold greater odds of hearing impairment (OR=5.01, 95% CI: 2.24-11.24; χ²=16.34, p<0.001). Exposure to an ototoxic antitubercular drug showed the strongest association: 72.1% of affected patients had such exposure compared with 26.0% of unaffected patients, corresponding to more than sevenfold higher odds of impairment (OR=7.36, 95% CI: 3.18-17.03; χ²=24.01, p<0.001). An ATT duration of ≥6 months was found in 65.1% of patients with hearing impairment compared with 35.1% of those without impairment and was associated with significantly increased odds (OR=3.46, 95% CI: 1.58-7.56; p=0.002). Previous ATT was documented in 48.8% of affected patients compared with 24.7% of unaffected patients and was associated with nearly threefold higher odds of hearing impairment (OR=2.91, 95% CI: 1.32-6.43; p=0.007).

The present study evaluated hearing status among 120 patients receiving antitubercular therapy and found audiometrically confirmed hearing impairment in 43 patients, giving a prevalence of 35.8% (95% CI: 27.8%-44.7%). This represents a substantial burden because the study population included patients receiving treatment for both drug-sensitive and drug-resistant tuberculosis, rather than only patients exposed to second-line injectable drugs. Dillard et al. (2021)^[1], in a systematic review of 18 studies, reported a pooled prevalence of aminoglycoside-induced hearing loss of approximately 40.6% among patients treated for drug-resistant TB. The prevalence in the present study was slightly lower, probably because only 42.5% of participants had documented exposure to an ototoxic antitubercular drug. In contrast, Faye et al. (2025)^[2] reported a closely comparable prevalence of 37.2% among 438 patients treated for drug-resistant TB in South Africa. The systematic review and recent South African study therefore support the magnitude observed in the present investigation.

Wangchuk et al. (2021)^[3] detected hearing loss in 45.2% of patients receiving multidrug-resistant TB treatment in Bhutan, whereas Sharma et al. (2016)^[4] demonstrated progressive hearing-threshold deterioration among Indian patients receiving kanamycin. Ghafari et al. (2020)^[5] reported a markedly higher incidence of 82.4% among kanamycin-treated MDR-TB patients. Their use of serial audiometry extending to ultra-high frequencies of 16 kHz probably identified early cochlear damage that would have been missed by conventional pure-tone audiometry. Differences between studies may also reflect variations in baseline hearing status, diagnostic criteria, aminoglycoside type, cumulative exposure, HIV prevalence, monitoring intervals and inclusion of extended high-frequency thresholds.

Patients with hearing impairment in the present study were significantly older than those with normal hearing, with mean ages of 49.1 and 40.7 years, respectively. Age ≥50 years was associated with more than threefold higher odds of impairment (OR=3.16, 95% CI: 1.45-6.88). Faye et al. (2025)^[2] similarly found that hearing impairment increased with advancing age, particularly among older DR-TB patients. Hong et al. (2020)^[6] reported a substantial prevalence of pre-existing hearing loss among patients about to begin DR-TB treatment, emphasizing that age-related or previously acquired hearing loss may coexist with subsequent drug-related ototoxicity. Consequently, the higher prevalence among older patients in the present study may represent the combined effects of presbycusis, comorbid disease, previous treatment exposure and reduced cochlear reserve.

Although 63.3% of the participants were male, sex was not significantly associated with hearing impairment (OR=1.32, p=0.485). Ghafari et al. (2020)^[5] also failed to demonstrate a significant association between sex and kanamycin-induced hearing loss. However, Faye et al. (2025)^[2] reported male sex as a predictor of hearing impairment. This difference may be attributable to occupational noise exposure, smoking and substance use among men in particular populations rather than to an independent biological effect of sex.

The mean duration of ATT was significantly longer among patients with hearing impairment than among those with normal hearing (7.0 versus 5.1 months, p=0.003). Treatment for ≥6 months was associated with 3.46 times greater odds of impairment. Aminoglycoside ototoxicity is cumulative and may progress from high-frequency cochlear damage to involvement of frequencies required for speech perception. Ghafari et al. (2020)^[5] found that higher kanamycin exposure, expressed as the area under the concentration-time curve, was significantly associated with hearing loss. Van Altena et al. (2017)^[7] demonstrated that therapeutic drug monitoring of aminoglycosides was associated with a lower probability of hearing loss during MDR-TB treatment. These findings support careful control of cumulative exposure and periodic monitoring during prolonged treatment.

Drug-resistant TB was independently important at the unadjusted level, with 60.5% of hearing-impaired patients having DR-TB compared with 23.4% of those with normal hearing. DR-TB was associated with approximately fivefold greater odds of hearing impairment. This association is clinically plausible because older DR-TB regimens frequently included injectable aminoglycosides for extended periods. Khoza-Shangase et al. (2021)^[8] showed that patients treated with aminoglycoside-containing regimens experienced poorer hearing outcomes than those receiving bedaquiline-based treatment. Souleymane et al. (2021)^[9] reported that baseline and monthly audiometry, followed by replacement of injectable drugs with linezolid when hearing abnormalities emerged, prevented severe hearing loss while maintaining favourable treatment outcomes. These observations support the transition toward effective all-oral DR-TB regimens.

Exposure to an ototoxic antitubercular drug showed the strongest treatment-related association in the present study. Hearing impairment was observed in 72.1% of exposed patients compared with 26.0% of unexposed patients, giving an OR of 7.36 (95% CI: 3.18-17.03). Dillard et al. (2021)^[1] found pooled hearing-loss estimates of approximately 49.7% for kanamycin and 38.9% for amikacin, confirming considerable variation between injectable drugs. Sabur et al. (2021)^[10] found that low-dose amikacin remained clinically useful in selected MDR-TB patients but required careful therapeutic and audiological monitoring. Lindeborg et al. (2022)^[11] consequently recommended avoidance of irreversible ototoxic medicines whenever effective alternatives are available and structured monitoring when their use cannot be avoided.

Previous ATT was associated with nearly threefold higher odds of hearing impairment (OR=2.91, p=0.007). Repeated courses may result in cumulative aminoglycoside exposure or leave residual cochlear injury from earlier treatment. In addition, patients who previously received ATT are more likely to have recurrent or drug-resistant disease and consequently receive longer and more toxic regimens. Hong et al. (2020)^[6] emphasized that baseline hearing assessment is essential because pre-existing impairment may otherwise be incorrectly classified as incident ototoxicity.

Diabetes mellitus was associated with 2.67 times higher odds of hearing impairment, while renal dysfunction was associated with approximately threefold higher odds. Diabetes may contribute through microangiopathy, oxidative stress and neuropathic injury involving the auditory system. Renal dysfunction is particularly important because impaired aminoglycoside clearance can increase systemic and cochlear drug exposure. Lindeborg et al. (2022)^[11] identified underlying renal disease as a reason for intensified ototoxicity monitoring, while Ghafari et al. (2020)^[5] highlighted the role of pharmacokinetic exposure in kanamycin-related cochleotoxicity. Although these associations are biologically plausible, the relatively wide confidence interval for renal dysfunction indicates limited precision and warrants confirmation in a larger sample.

Subjective hearing difficulty was strongly associated with audiometrically detected impairment (OR=60.23), and tinnitus was associated with more than eightfold higher odds. Nevertheless, 10 patients with audiometrically confirmed impairment did not report subjective hearing difficulty, indicating that symptom-based surveillance alone would have missed almost one-quarter of affected patients. Ganesan et al. (2018)^[12] noted that ototoxicity commonly begins at frequencies above the speech range and may remain unnoticed until cochlear damage progresses. Konrad-Martin et al. (2018)^[13] similarly recommended pretreatment counselling, baseline testing and serial objective audiological monitoring rather than relying solely on symptoms. Tinnitus should therefore be treated as an important warning symptom, but its absence should not be interpreted as evidence of normal hearing.

Vertigo or imbalance was more frequent among patients with hearing impairment, although the association narrowly failed to reach statistical significance (OR=3.55, p=0.052). Aminoglycosides can affect both cochlear and vestibular hair cells, but vestibular toxicity may be under-recognized because symptoms can be nonspecific and partially compensated by visual and proprioceptive mechanisms. Krause et al. (2016)^[14] described both cochlear and vestibular toxicity as important dose-limiting adverse effects of aminoglycosides. The borderline result in the present study may reflect the small number of patients reporting vestibular symptoms and the absence of formal vestibular-function testing.

Among the 43 affected participants, bilateral hearing impairment was significantly more common than unilateral impairment (72.1% versus 27.9%, p=0.004). Sensorineural hearing loss was the predominant type, accounting for 79.1% of cases, whereas conductive and mixed losses accounted for 14.0% and 7.0%, respectively. This distribution is characteristic of aminoglycoside cochleotoxicity, which generally causes bilateral sensorineural damage. Wangchuk et al. (2021)^[3] likewise documented predominantly sensorineural impairment during MDR-TB treatment. Conductive and mixed losses in the present study may have resulted from unrelated external- or middle-ear disease and should not automatically be attributed to ATT.

Mild and moderate hearing impairment together accounted for 72.1% of affected patients, while severe and profound impairment were comparatively uncommon. This may suggest that most cases were detected before extensive involvement of speech frequencies. Stevenson et al. (2022)^[15] nevertheless demonstrated longitudinal deterioration during kanamycin treatment, showing that mild high-frequency changes can progress if exposure continues. Audiological abnormalities should therefore trigger prompt clinical review even when functional communication remains relatively preserved.

Predominantly high-frequency hearing loss was found in 62.8% of affected participants and was significantly more common than flat or speech-frequency involvement. This pattern supports a possible ototoxic mechanism because cochlear injury from aminoglycosides typically begins in the basal turn, which is responsible for high-frequency perception, and subsequently progresses toward lower frequencies. Hollander et al. (2020)^[16] highlighted the value of standardized ototoxicity grading for detecting early high-frequency changes and guiding clinical action. Conventional audiometry alone may underestimate early toxicity; extended high-frequency audiometry and otoacoustic emissions may provide greater sensitivity where resources permit.

Hearing impairment was a clinically important adverse outcome among patients receiving antitubercular therapy, affecting 35.8% of the study population. Bilateral sensorineural hearing loss was the predominant presentation, and most affected patients had mild-to-moderate impairment with predominantly high-frequency involvement. Older age, diabetes mellitus, renal dysfunction, drug-resistant tuberculosis, previous ATT, longer treatment duration and exposure to ototoxic antitubercular drugs were significantly associated with hearing impairment. Ototoxic-drug exposure demonstrated the strongest association. Baseline and periodic audiological assessments, particularly among high-risk patients, may facilitate early detection and help prevent progression to irreversible speech-frequency hearing loss. Whenever clinically feasible, safer all-oral regimens should be preferred, while patients with established impairment should receive appropriate audiological counselling and rehabilitation.

LIMITATIONS OF STUDY

The study had several limitations. Its cross-sectional design established associations but could not determine temporal or causal relationships between ATT exposure and hearing impairment. Pretreatment audiograms were unavailable for some participants; therefore, pre-existing hearing loss could not always be distinguished from treatment-related ototoxicity. The single-centre setting and relatively small sample size limited the generalizability of the findings. Participants receiving drug-sensitive and drug-resistant TB regimens were included together, resulting in variations in drug exposure and treatment duration. Detailed cumulative doses, serum aminoglycoside concentrations and adherence to individual medicines could not be assessed uniformly. Extended high-frequency audiometry, otoacoustic-emission testing and serial vestibular assessment were not performed in every patient. Residual confounding from previous noise exposure, HIV infection, nutritional status, genetic susceptibility and concomitant ototoxic medicines might also have influenced the findings. Finally, the relatively small numbers in some subgroups resulted in wide confidence intervals for certain associations.