Evaluation of Diagnostic Accuracy of Wideband Tympanometry Absorbance for Otosclerosis

Abstract

Otosclerosis is a common cause of hearing loss, yet diagnosing it remains a challenge. Wideband Tympanometry Absorbance (WTA) has been proposed as a potential tool for improving diagnostic accuracy. The aim of study was to investigate the diagnostic value of WTA in diagnosing otosclerosis by comparing its results in patients with clinically diagnosed otosclerosis and a normal control group. This prospective study was conducted in Mashhad, Iran, from 2022 to 2023, involving 64 participants including otosclerosis and control. Inclusion criteria for the otosclerosis group included a clinical diagnosis of otosclerosis confirmed by audiometric tests and conventional tympanometry, and eligibility for surgical intervention. The control group consisted of individuals with normal clinical audiometric and tympanometric results. Both groups underwent conventional and WTA tympanometric assessments. The evaluation of WTA involved 64 participants, divided evenly between otosclerosis a normal ears group. Conventional tympanometry at 226 Hz showed significant differences in compliance between otosclerosis and normal ears (p = 0.02). In contrast, at 1000 Hz did not demonstrate significant differences in compliance (p = 0.2). Also, WTA did not demonstrate significant differences in compliance (p > 0.9). Measurements of gradient and resonance across both 226 and 1000Hz frequencies showed no statistically significant differences. The sensitivity and specificity of WTA, evaluated through compliance, were 72% and 40%, respectively, with an area under the curve of 0.50. The study illustrates that while conventional tympanometry is effective in distinguishing otosclerosis from normal tympanic conditions, WTA shows limited diagnostic efficacy. The moderate sensitivity and specificity of WTA suggest that it should not be relied upon as the sole diagnostic tool.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12070-024-04876-5.

Introduction

Otosclerosis is a primary osteodystrophic disorder affecting the otic capsule and is a leading cause of progressive conductive hearing loss [1, 2]. Despite the, including clinical audiometry and traditional tympanometry [3–5], otosclerosis often eludes early detection due to its overlap with other middle ear pathologies, highlighting the need for more discriminative diagnostic methods.

Advances in audiological diagnostics have yet to simplify the early and accurate identification of otosclerosis. Conventional tympanometry, assessing middle ear function at single or limited frequency ranges, fails to capture the complex impedance variations characteristic of otosclerosis [6, 7]. Precise diagnosis is essential for appropriate and timely therapeutic interventions, which are crucial for improving auditory outcomes.

The sensitivity and specificity of traditional tympanometry are generally high for ruling out middle ear pathologies, but its low sensitivity limits its diagnostic efficacy as a standalone tool for confirming otosclerosis [8–10]. This technique offers enhanced sensitivity and specificity in detecting the subtle alterations associated with early otosclerosis, thus overcoming the limitations of traditional diagnostic approaches.

Validating the diagnostic accuracy of WTA addresses a significant gap in otological research and could potentially revolutionize the approach to diagnosing otosclerosis, facilitating more targeted interventions. This study aims to evaluate the diagnostic utility of WTA for otosclerosis by comparing its performance against that of a normal, age- and sex-matched control group. We hypothesize that WTA will demonstrate superior sensitivity and specificity relative to conventional tympanometric measures, offering a more definitive diagnostic tool for otosclerosis. Additionally, the study will investigate whether WTA can identify distinctive tympanometric patterns indicative of otosclerosis, potentially aiding in its differential diagnosis from other middle ear conditions.

Material and Methods

Study Design and Setting

This prospective, bicentric study was conducted between 2022 and 2023 at the referral Ear, Nose, and Throat (ENT) clinics in Mashhad, Iran.

Participants

The study enrolled patients presenting to the ENT clinics with hearing complaints within the study timeframe. Inclusion criteria included patients diagnosed with otosclerosis based on clinical evaluations, audiometry, and conventional tympanometry, who were candidates for surgery. Additionally, a control group consisted of patients with normal clinical, audiometric, and tympanometric results, matched for age and sex with the otosclerosis group. Exclusion criteria included non-cooperation in completing further tests and lack of access to complete clinical and paraclinical data.

Among the participants, 32 had clinically diagnosed otosclerosis, and 32 had normal ears. It is important to note that within the otosclerosis group, some participants had only one ear affected by otosclerosis while the other ear was normal, and some had both ears affected.

All participants underwent Wideband Tympanometry Absorbance (WTA) testing using the Interacoustics Titan device, which also included capabilities for conventional tympanometry at 226 Hz and 1000 Hz. The device was operated by an experienced audiologist.

Ethics

Informed consent was obtained from all participants prior to their inclusion in the study. Data were collected and analyzed anonymously, using coded identifiers to ensure participant confidentiality. The study protocol was reviewed and approved by the Ethics Committee of Mashhad University of Medical Sciences (Approval No. IR.MUMS.MEDICAL.REC.1401.375).

Variables

Study variables included: Quantitative assessments of air and bone conduction thresholds across frequencies ranging from 250 to 8000 Hz, with specific attention to the differential thresholds between air conduction (AC) and bone conduction (BC) as indicators of conductive hearing loss. Tympanometric evaluations at 226 Hz, 1000 Hz, and WTA measuring ear canal volume, middle ear pressure, compliance, and gradient. Parameters were assessed both quantitatively and qualitatively based on established norms and thresholds.

Data Analysis

Variables were described using means, standard deviations, frequencies, and percentages. Normality of quantitative data was tested using the Shapiro–Wilk test. Depending on normality, either the independent two-sample t-test or the Wilcoxon rank sum test was used for analysis. Qualitative variable distributions were compared using the Chi-squared test or Fisher’s exact test as needed. All statistical analyses performed with “gtsummary” package in R Program. Also, receiver operating characteristic (ROC) analysis performed with “pROC” package in R program. For all statistical analyses, significance set at p < 0.05.

Results

Demographics Characteristics

In the study, a total of 64 participants were evaluated. The sample comprised 24 men (37.5%) and 40 women (62.5%), ensuring a balanced gender distribution. The median age of participants was 38.5 years.

Tympanometry Findings

According to audiometric test results and clinical evaluation, 48 ears were diagnosed with otosclerosis among the 128 ears evaluated. Therefore, tympanometric and WTA assessment findings were compared between normal and otosclerotic ears.

Tympanometric findings revealed significant variations in compliance between normal and otosclerotic ears. At 226 Hz, compliance in the normal subgroup averaged 0.3 mL (0.27–0.6 mL), compared to 0.3 mL (0.19–0.4 mL) in the otosclerosis subgroup, with these differences reaching statistical significance (p = 0.020).

At a frequency of 1000 Hz, compliance measurements showed no significant difference between the groups, with a median value of 1.2 mL (0.73–1.7 mL) for the normal subgroup and 1.3 mL (0.88–1.9 mL) for the otosclerosis subgroup (p = 0.2).

At WTA, compliance measurements showed no significant difference between the groups, with a median value of 45.0 percent (38.00–55.5) for the normal subgroup and 46.0 percent (39.75–53.0) for the otosclerosis subgroup (p > 0.9).

Gradient measurements at 226 Hz demonstrated median values of 86.0 (70.00–100.3) for the normal subgroup compared to 82.0 (68.50–102.0) for the otosclerosis subgroup, with no significant differences (p = 0.8). Similar non-significant results were observed at 1000 Hz, with the normal subgroup showing a gradient median of 97.0 (82.00–116.0) and the otosclerosis subgroup at 96.5 (85.25–111.5) (p = 0.8).

Resonance measures in the wideband frequency analysis showed that the normal subgroup had a median resonance of 1014.0 (range: 809.75 to 1062.8), while the otosclerosis subgroup showed higher values at 1040.5 (range: 901.25 to 1168.3), with these differences approaching statistical significance (p = 0.081).

The pattern in peak pressure showed significant differences between the groups, with 63 (98%) normal ears not showing similar patterns compared to 49 (77%) otosclerotic ears (p < 0.01). Only 1 (1.6%) normal ear showed a similar pattern compared to 15 (23%) otosclerotic ears.

Similarly, the pattern in room pressure also showed significant differences, with 62 (97%) normal ears not showing similar patterns compared to 48 (75%) otosclerotic ears (p < 0.01). Only 2 (3.1%) normal ears showed a similar pattern compared to 16 (25%) otosclerotic ears (Table 1).

Table 1.

Comparison of conventional and wideband tympanometry findings between otosclerosis and control ears

Parameter	Frequency	Normal, N = 80	Otosclerosis, N = 48	p-value*
Compliance (mL)	226 Hz	0.3 (0.27, 0.6)	0.3 (0.19, 0.4)	0.020
	1000 Hz	1.2 (0.73, 1.7)	1.3 (0.88, 1.9)	0.2
	Wideband	45.0 (38.00, 55.5)	46.0 (39.75, 53.0)	> 0.9
Gradient	226 Hz	86.0 (70.00, 100.3)	82.0 (68.50, 102.0)	0.8
	1000 Hz	97.0 (82.00, 116.0)	96.5 (85.25, 111.5)	0.8
Resonance	Wideband	1014.0 (809.75, 1062.8)	1040.5 (901.25, 1168.3)	0.081
Pattern in peak pressure	Not similar**	63 (98%)	49 (77%)	< 0.01
	Similar **	1 (1.6%)	15 (23%)
Pattern in room pressure	Not similar**	62 (97%)	48 (75%)	< 0.01
	Similar**	2 (3.1%)	16 (25%)

^*Wilcoxon rank sum test

**Similarity to proposed osteograph

Diagnostic Performance of Tympanometric Indices at Optimal Cutoff Values

The sensitivity and specificity of the tympanometric indices were determined using receiver operating characteristic (ROC) curves at their optimal cutoff values. The areas under the curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were evaluated for each parameter (Table 2).

Table 2.

Diagnostic values of Conventional and Wideband Tympanometry parameters

Variable		Value at Best Cutoff	AUC	Sensitivity	Specificity	PPV	NPV
Compliance (mL)	226 Hz	0.23	0.59	0.32	0.88	0.72	0.56
	1000 Hz	1.85	0.51	0.27	0.83	0.61	0.53
	Wideband	40.50	0.50	0.72	0.40	0.54	0.59
Gradient	226 Hz	77.50	0.59	0.52	0.68	0.62	0.59
	1000 Hz	94.50	0.52	0.48	0.62	0.56	0.54
Resonance	Wideband	1067.00	0.57	0.39	0.80	0.66	0.57

For Compliance at 226 Hz, the optimal cutoff was 0.23 mL. This index demonstrated an AUC of 0.59. It achieved a sensitivity of 32% and a specificity of 88% (Fig. 1A).

Figure1.

Receiver Operating Characteristic (ROC) Curves for Tympanometric Indices. A ROC curve for Compliance at 226 Hz. This curve demonstrates the diagnostic ability of compliance at 226 Hz, with an Area Under the Curve (AUC) of 0.588. B ROC curve for Gradient at 226 Hz. The curve shows the diagnostic performance of gradient at this frequency, with an AUC of 0.587. C ROC curve for Compliance at 1000 Hz. This panel illustrates the diagnostic capability of compliance at 1000 Hz, with an AUC of 0.512. D ROC curve for Gradient at 1000 Hz. The curve depicts the diagnostic utility of gradient at 1000 Hz, with an AUC of 0.517

At 1000 Hz, the optimal cutoff for compliance increased to 1.85 mL. This showed a lower AUC of 0.51, with a sensitivity of 27% and a specificity of 83% (Fig. 1C).

At WTA, the optimal cutoff for compliance increased to 40.50 percent. This showed a lower AUC of 0.50, with a sensitivity of 72% and a specificity of 40% (Fig. 2A).

Fig. 2.

Receiver Operating Characteristic (ROC) Curves for Wideband Indices. A ROC curve for compliance at Wideband. This curve demonstrates the diagnostic ability with an Area Under the Curve (AUC) of 0.497. B ROC curve for Resonance at Wideband. This curve demonstrates the diagnostic ability with an Area Under the Curve (AUC) of 0.571

The Gradient at 226 Hz had an optimal cutoff of 77.50, with an AUC of 0.59. It demonstrated a sensitivity of 52% and a specificity of 68% (Fig. 1B).

At 1000 Hz, the Gradient's optimal cutoff was 94.50, with a slightly lower AUC of 0.52. It had a sensitivity of 48% and a specificity of 62% (Fig. 1D).

Resonance, measured at the optimal cutoff of 1067.00, yielded an AUC of 0.57. It showed a sensitivity of 39% and a specificity of 80% (Fig. 2B).

Discussion

The aim of this study was to compare the results of 226 Hz, 1000 Hz and wideband tympanometry in patients with otosclerosis and determine the diagnostic value of wideband tympanometry. The findings reveal that conventional tympanometry at 226 Hz generally offers higher specificity, particularly in the measure of compliance, which may be useful for ruling out otosclerosis in a clinical setting. However, its low sensitivity limits its diagnostic efficacy as a standalone tool. Wideband tympanometry measures and resonance do not significantly enhance the diagnostic capability for otosclerosis, as indicated by generally lower or comparable AUC values and sensitivities.

The overall modest AUC values across all measures suggest that neither conventional nor wideband tympanometry alone is sufficient for accurate diagnosis of otosclerosis. These methods may, therefore, be better used as part of a comprehensive diagnostic approach, which could include additional audiometric tests and clinical evaluation to improve diagnostic accuracy.

In clinical practice, while conventional tympanometry can help exclude other middle ear pathologies with its higher specificity at 226 Hz, its utility in confirming otosclerosis is limited [7, 11]. Wideband tympanometry's advantage might lie in its ability to provide more detailed information about middle ear function across a wider range of frequencies, which could potentially be explored further in research to refine diagnostic thresholds and criteria for otosclerosis.

The capability of tympanometric indices to accurately diagnose otosclerosis appears to be constrained, underlining the need for more sensitive and specific diagnostic technologies or a multimodal diagnostic approach to enhance clinical outcomes for patients suspected of having otosclerosis.

Ogut et al. [12] utilized multiple-frequency tympanometry, measuring middle-ear resonant frequency in a cohort of 25 surgically confirmed otosclerotic ears versus 100 normal ears. They demonstrated a significantly higher mean resonant frequency in the otosclerotic group (1190 Hz) compared to controls (934.6 Hz), achieving a sensitivity of 80% and a specificity of 82% at a resonant frequency cutoff of 1025 Hz. In contrast, our study incorporated both 226 Hz and 1000 Hz tympanometric evaluations revealing generally low sensitivity but high specificity across various tympanometric indices. Notably, at 226 Hz for compliance, our findings showed a sensitivity of 32% and specificity of 88%, indicating a lower diagnostic sensitivity.

Karuppannan and Barman [13], explored the diagnostic potential of Wideband Absorbance Tympanometry (WBA) and reported high sensitivity and specificity, particularly at 1000 Hz where both exceeded 90% for detecting otosclerosis. Their study demonstrated significant differences in WBA values at 1000 Hz between otosclerotic and healthy ears, suggesting a strong diagnostic capability of WBA at this frequency. In contrast, our study, while also examining WBA, found lower diagnostic accuracy. Specifically, our results indicated a sensitivity of 72% and specificity of 40% for WBA, with an area under the curve (AUC) of 0.50, highlighting a more modest diagnostic performance. These differences can be attributed to several factors: sample size variations, differences in participant selection criteria, and methodological differences in pressure conditions during WBA measurement. Additionally, our study evaluated WBA at 226 Hz and 1000 Hz, where conventional tympanometry at 226 Hz showed higher specificity but lower sensitivity, suggesting it may be better suited for ruling out rather than confirming otosclerosis. This discrepancy underscores the necessity of using WBA in conjunction with other diagnostic tools to achieve higher diagnostic accuracy for otosclerosis.

Niemczyk et al. [14] examined absorbance plots in 77 otosclerotic ears, distinguishing five unique absorbance plot types based on detailed characteristics such as resonance frequency, number of peaks, and variations in plot width at specified absorbance heights. The approach by Niemczyk et al. underscores the potential for using detailed absorbance plot analyses to enhance diagnostic accuracy, suggesting that such detailed classifications could enable more tailored diagnostic and therapeutic strategies. Meanwhile, our findings suggest that while WTA shows promise in identifying otosclerosis, its variable sensitivity necessitates its use alongside other diagnostic tools, highlighting the need for a multimodal diagnostic strategy.

Conclusion

In conclusion, while both conventional and wideband tympanometry have their place in the diagnostic process for otosclerosis, they should ideally be part of a broader diagnostic strategy that includes other modalities to overcome their individual limitations. Further studies might focus on how these tools can be integrated effectively into diagnostic protocols or how their sensitivity can be enhanced, possibly through technological advancements or combined methodologies.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Conflict of interest

None.

Informed Consent

Patient were informed cosent about study. Data was collected and analyzed anonymously using codes to maintain confidentiality. The study protocol was approved by the Ethics Committee of Mashhad University of Medical Sciences (Approval No. IR.MUMS.MEDICAL.REC.1401.375).

Research Involving Human Participants and/or Animals

All procedures performed in study involving human participant were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.