Elevated bone conduction thresholds across pathogen groups in chronic otitis media: a retrospective comparative study

Chronic otitis media (COM) may be accompanied by elevated bone-conduction thresholds. This study compared audiometric characteristics across pathogen groups and examined the association between pathogen group and elevated bone-conduction pure-tone average (BC-PTA). This retrospective study included 190 patients with COM: 53 in the tuberculous otitis media (TOM) group, 45 in the Gram-negative group, 41 in the Gram-positive group, and 51 in the culture-negative group. Elevated BC-PTA was defined as BC-PTA > 20 dB HL. The primary multivariable model included all 190 patients and adjusted for age, sex, disease duration, cholesteatoma status, pathogen group, and recruitment period. One sensitivity analysis was restricted to 169 patients enrolled during the common recruitment period. A second retained all non-TOM patients and only the 21 TOM patients with lesion-site culture or molecular evidence of M. tuberculosis complex ( n = 158). Elevated BC-PTA was observed in 60.4% of the TOM group, 68.9% of the Gram-negative group, 36.6% of the Gram-positive group, and 31.4% of the culture-negative group (overall P < 0.001). Median BC-PTA was 30.0 and 31.2 dB HL in the TOM and Gram-negative groups, respectively, compared with 18.8 and 16.2 dB HL in the Gram-positive and culture-negative groups. In the primary adjusted analysis, TOM (aOR, 6.616; 95% CI, 2.101–20.831; P = 0.001) and Gram-negative infection (aOR, 3.484; 95% CI, 1.274–9.530; P = 0.015) were associated with higher odds of elevated BC-PTA than Gram-positive infection. The direct comparison between TOM and Gram-negative infection was not statistically significant (aOR, 1.899; 95% CI, 0.615–5.860; P = 0.265). Older age and longer disease duration were also associated with elevated BC-PTA. Both sensitivity analyses yielded adjusted odds ratios similar in direction and magnitude to those of the primary analysis. TOM and Gram-negative infection were associated with higher odds of elevated BC-PTA than Gram-positive infection, whereas the difference between TOM and Gram-negative infection remained uncertain. These findings may support careful etiological assessment and comprehensive audiometric evaluation in patients with COM.