From Persistent Otorrhea to Hearing Loss: Clinical Predictors of Hearing Impairment in Chronic Suppurative Otitis Media : A Systematic Review of Randomized Controlled Trials and Primary Studies

Abstract

Introduction: Chronic suppurative otitis media (CSOM) is the single most important preventable cause of acquired hearing impairment worldwide, yet the transition from persistent otorrhea to clinically meaningful hearing loss is heterogeneous, and clinicians lack a consolidated evidence-based hierarchy of bedside predictors. This systematic review was undertaken to identify, appraise, and synthesise the clinical, otoscopic, audiometric, and intraoperative predictors of hearing impairment in CSOM.

Methods: The study strictly adhered to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) 2020 guidelines. Eligible designs: RCTs, cohort, case-control, cross-sectional, and observational studies. Systematic reviews and meta-analyses excluded from the characteristics table. Two reviewers independently screened, extracted data, and appraised risk of bias (JBI, QUADAS-2). Certainty graded with GRADE. Due to substantial heterogeneity, a structured narrative synthesis with tabulated effect estimates and reconstructed ROC analysis was performed.

Results: Twenty-six primary studies comprising 5,507 participants and at least 5,668 ears met eligibility. Hearing impairment was near-universal in CSOM, with 50–78% prevalence in community series and 91% of children with CSOM developing permanent threshold shift >15 dB HL. Conductive loss predominated (63–98%), while a sensorineural component was documented in 14.7–52% of series. Six predictors were consistently and independently associated with hearing loss magnitude: (i) tympanic membrane perforation size with a reproducible dose–response gradient (mean hearing level 35.3 dB for small, 39.7 dB for medium, 45.9 dB for large; mean ABG 9.12, 16.13, and 26.43 dB); (ii) posterior-quadrant and malleolar perforation site (posterosuperior 23.9 dB versus anterosuperior 29.30 dB in small perforations; malleolar 29.26 dB versus non-malleolar 18.25 dB); (iii) duration of otorrhea and disease — the strongest single predictor in an African series (p<0.0001) with Spearman ρ=0.689; (iv) persistent or high-frequency active discharge; (v) squamosal disease and cholesteatoma, with ossicular erosion in 82% of cholesteatomatous ears versus 12.7% of non-cholesteatomatous ears; and (vi) ossicular chain status, where discontinuity raised mean PTA from 43.62 to 58.15 dB HL and ABG from 27.89 to 38.88 dB. In multivariable models, co-existing CSOM conferred OR 16.70 for permanent hearing loss; cholesteatoma (HR 3.24), preoperative PTA >50 dB (HR 2.89), and disease duration >5 years (HR 2.52) independently predicted unfavourable outcome. The high-frequency conductive hearing loss test at a 10 dB threshold identified incomplete ossicular discontinuity with 83% sensitivity and 92% specificity (reconstructed AUC 0.95); optimal ABG cut-offs for discontinuity and fixation were 28.4 dB and 32.8 dB.

Discussion: The synthesised evidence supports a coherent two-compartment model: a mechanical compartment (loss of tympanic membrane vibrating area, loss of round-window acoustic shielding, progressive ossicular erosion) explains the conductive component, while a cytotoxic compartment (transmission of bacterial exotoxins and inflammatory mediators across the round-window membrane) explains the frequency-dependent sensorineural component that accumulates with disease duration and discharge activity. The two compartments are not independent — duration of otorrhea acts as a shared upstream driver. Clinically, this justifies a stratified surveillance pathway in which large, posterior, or malleolar perforations, squamosal disease, and duration >5 years are prioritised for early audiometric assessment and definitive surgery.

Conclusion: Hearing impairment in CSOM is predictable rather than random. Perforation size, posterior and malleolar site, duration and activity of otorrhea, squamosal or cholesteatomatous disease, and ossicular chain integrity constitute a reproducible predictor set assessable at the bedside and refined audiometrically. Preoperative pure-tone audiometry, particularly frequency-specific ABG and the high-frequency conductive hearing loss test, offers clinically useful discrimination and should be incorporated into routine counselling and surgical planning. Prospective multicentre cohorts using standardised CSOM definitions are required before a validated risk-prediction instrument can be recommended.