This study adhered to the principles of the Declaration of Helsinki (“World Medical Association Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects,” 2013), and complied with national and institutional ethical guidelines for human research. Ethical approval was granted by the Institutional Review Board of the Policlinico Hospital Consortium in Bari, Italy (approval number 10336/2023CE). All participants provided written informed consent prior to inclusion.

A retrospective case-control study of 34 cochlear implantation (CI) adults (> 18 years) treated by the senior author between 2018 and 2022 was conducted. The review included age, gender, indications, audiological results, and post-operative complications.

Participants were divided into two groups. The COM group included 17 consecutive subjects with chronic otitis media (COM) who underwent cochlear implantation simultaneously with SP for eradication of inflammatory disease. The CI group involved 17 age and sex matched patients without middle ear disease who underwent CI alone for severe to profound bilateral sensorineural hearing loss (SNHL). Diagnosis of COM was based on symptoms persisting for more than 3 months, otomicroscopy findings, and temporal bone computed tomography (CT). Patients with adhesive otitis media, atelectasis, chronic suppurative otitis media (COM), COM with cholesteatoma (CCOM) and those who had previously undergone canal wall-down (CWD) mastoidectomy were included. In the COM group, the middle ear inflammatory status was classified as follows:

Exclusion criteria included age under 18 years, incomplete medical, or audiological records, evidence of inner ear malformations, or retrocochlear pathology on preoperative imaging, and a postoperative follow-up period of less than 12 months.

Cochlear implantation was indicated in all cases based on the current standard audiological criteria. Pre-surgery high-resolution computed tomography (CT) and magnetic resonance imaging (MRI) scans showed no evidence of inner ear malformations in all cases. In the CI group, electrode insertion was achieved through a classical transmastoid posterior tympanotomy approach. In the COM group, cochlear implantation was performed in a single stage, along with subtotal petrosectomy, to create a safe, closed, and dry cavity before electrode insertion. In every case, it consisted of a complete exenteration of all pneumatic cells of the mastoid, removal of the canal wall and middle ear structures, and double-layer blind-sac closure of the EAC. The Eustachian tube was then subjected to a systematic inspection. In the absence of cerebrospinal fluid leakage, a perilymph gusher, or a significant perceived risk of ascending infection from the nasopharynx, the tube was left patent. In only two cases, obliteration with muscular grafts and connective tissue was performed selectively for the indications mentioned above, with the middle ear and mastoid cavity obliterated with abdominal fat. In the remaining 15 subjects, the middle ear and the mastoid cavity were not routinely obliterated, as would typically be expected. All subjects were implanted with a Nucleus multichannel device (Cochlear LTD, Sydney, Australia) using the extended round window (RW) approach and perimodiolar array. A complete array insertion into the cochlea was confirmed intraoperatively via neural response telemetry. In all cases, the round window niche was obliterated with autologous muscle to protect the cochlea and prevent perilymph leakage. An experienced cochlear implant (CI) audiologist programmed the speech processor and optimized the parameters after initial activation to ensure optimal speech understanding in everyday situations. All patients used a cochlear implant with the same Advanced Combination Encoder (ACE)-based strategy. None of the patients discontinued CI use with an average daily use of 15.4 ± 1.5 h.

Unaided and post-implantation pure-tone audiometry was performed in all subjects. The pure-tone average across five frequencies (PTA₅: 0.25, 0.5, 1, 2, and 4 kHz) was calculated to assess hearing thresholds, with all measurements expressed in decibels hearing level (dB HL). PTA₅ was selected for its broader representation of the speech-frequency range compared to PTA₄ (0.5, 1, 2, and 4 kHz), thereby providing a more comprehensive measure of auditory function in quiet conditions.

Speech recognition in quiet conditions was evaluated in a sound field using pre-recorded speech stimuli (60 disyllabic words) administered at 65 dB SPL at 0° azimuth. Scores were reported as the percentage of correctly repeated words.

Speech recognition in noise was measured under best-aided conditions in a sound field. Tests were conducted in a sound-treated chamber using a loudspeaker positioned one meter away from the subject's head (S0N0). Stimuli consisted of taped disyllabic words presented at 65 dB SPL with a fixed signal-to-noise ratio (SNR) of + 5 dB (cocktail party noise at 60 dB SPL).

Two lists of 50 words (Turrini et al., 1993) were used and speech perception scores were calculated as the percentage of correctly identified words. The randomization of speech lists was implemented across all participants, with the exclusion of repetition within the same session.

Speech comprehension in noise was further examined using the Italian Oldenburg Sentence Test (OLSA) (Puglisi et al., 2015). The evaluation was conducted in an open field with speech and noise presented from the front (S₀N₀). Background noise was set at 65 dB SPL, and the speech level was adaptively adjusted according to each participant's responses to determine the SNR corresponding to 50%-word recognition (critical SNR). Two lists of 30 sentences were used following one training list.

The Speech, Spatial, and Qualities of Hearing Scale (SSQ) (Gatehouse and Noble, 2004), was administered pre- and post-implantation to assess subjective hearing performance. The SSQ comprises three sections: 14 items assessing speech perception, 17 items evaluating spatial hearing, and 18 items concerning the quality of hearing. Each item is rated on a 0–10 scale, with 0 representing ‘unable to hear' and 10 representing ‘perfect hearing'. All participants were asked to consider their daily listening conditions in answering an expert physician's question.

All patients with COM underwent clinical follow-up at 1 week, 1 month, 3 months, 6 months after surgery, and then yearly. Patients at risk for cholesteatoma recurrence, specifically those with a previous history of cholesteatoma surgery or with intraoperative suspicion of residual disease, underwent annual CT follow-up.

Audiological and patient-reported outcome measures were analyzed at the last available follow-up visit after CI activation, with comparable follow-up duration between groups and a mean duration of approximately 36 months.

Local complications of CI were assessed using a score of 0–3 (0, no retraction; 1, retraction with subcutaneous protrusion of the receiver; 2, protrusion of the implant and array; and 3, extrusion of some part of the implant), as reported by Di Bari et al. (2024).

Statistical analysis was conducted using Microsoft Excel for Microsoft 365 MSO (Version 2306), MedCalc software version 22.009 by MedCalc Software Ltd, and the web application Statistics Kingdom 2017. Data distribution was assessed with the Shapiro–Wilk normality test. The mean and standard deviation (SD) were calculated for quantitative variables with a normal distribution. For variables without a normal distribution, the median and 95% confidence interval for the median were determined. The means or medians of independent samples were compared using the Student's t-test or the Mann-Whitney U test. The statistical significance was set at p<0.05.

Demographics and audiological characteristics of both groups are reported in Table 1.

Mean age at implantation was 62.36 years (±14.91) in the CI group and 63.28 years (± 15.65) in the COM group. Mean follow-up time was 36 months in both groups (SD ± 0.52 in the CI group and ± 0.81 in the COM group). In all patients of the COM group, hearing loss in the implanted ear was progressive. In the CI group, hearing loss was progressive in 16 patients (94.2%) and sudden in one (5.8%), with variable etiologies, including Ménière's disease (n = 1), otosclerosis (n = 4), and idiopathic cases (n = 11). The unaided PTA₅ in the implanted ear was 107.5 ± 8.6 dB HL in the CI group and 111.83 ± 7.5 dB HL in the COM group. The PTA₅ in the better ear was 92.9 ± 8.1 dB HL in the CI group and 92.6 ± 7.3 dB HL in the COM group. The mean duration of hearing loss (HL), calculated from the patient-reported onset of subjective hearing impairment, was 12.7 years (SD ± 4.6) in the COM group, and 12.9 years (SD ± 3.5) in the CI group.

Overall, no statistically significant differences were observed between the COM and CI groups in terms of age at implantation, duration of hearing loss, unaided PTA5 in the implanted and contralateral ears, or follow-up duration (all p > 0.05), confirming baseline comparability between the two cohorts.

The characteristics of the COM group are summarized in Table 2.

Nine patients were initially affected by COM without cholesteatoma and 8 patients by CCOM. Two patients had not undergone previous surgery, while the remaining 15 had been previously treated with CWUT in 10 cases and CWDT in the other 5. During the surgical procedure, the middle ear mucosa appeared quiescent in the majority of cases while an active middle ear disease was identified in three patients. All patients underwent subtotal petrosectomy (SP) with blind sac closure of the EAC, followed by cochlear implantation as a single-stage procedure. The mastoid cavity was obliterated with abdominal fat and the Eustachian tube sealed with muscular grafts and connective tissue in only two cases (Table 2).

One patient, operated on during an active stage of COM without cholesteatoma, experienced a postoperative local complication characterized by cavity infection and failure of the EAC closure 3 months after surgery. The condition was managed with revision surgery and broad-spectrum antibiotic therapy, without the need for CI removal. Additionally, one case of abdominal subcutaneous hematomas was drained under local anesthesia 2 days after surgery. At the last follow-up, only one patient had a grade 1 post-auricular retraction (Di Bari et al., 2024). No recurrences of COM, and CCOM were observed on periodic CT-scan.

Audiological outcomes were assessed at the last follow-up post-CI activation.

Overall, auditory performance with CI improved in both groups. The aided PTA₅ was 36. 6 dB (± 2.5) in the COM group and 35, 6 (± 1.05) in the CI group with no significant difference (p = 0.09).

The results of speech in quiet at 65 dB (Figure 1) were not different between the two groups (p-value = 0.07): the COM patients recognized 74, 17% (± 20.65) and CI group 85% (± 13.22) of disyllabic words.

Speech comprehension in noise (Figure 2), measured through the OLSA test, was similar between groups, with no significant differences (p = 0.57). The COM group reached 50% of the speech recognition threshold (SRT) at an SNR of 2.73 dB (± 2.33), while the CI group had an SNR of 0.73 dB (± 0.83).

Figure 3 shows the comparative analysis between the post-implantation ratings for the two groups across the three domains of the Speech, Spatial, and Qualities of Hearing Scale (SSQ). In the Speech section, the COM group obtained a mean score of 4.73 (± 2.7), vs. 4.80 (± 2.7) of the control group, with no statistically significant difference observed (p = 0.69). In the Spatial section, the COM group showed a mean score of 4.96 (± 2.5) compared to the control group's mean score of 4.22 (± 2.3), again with no statistically significant difference (p = 0.50). In the Qualities Section, the COM group attained a mean score of 5, 17 (± 2, 3), while the control group had a mean score of 5,25 (± 1.9), once more without statistical significance (p = 0.87).

From a surgical standpoint, the low complication rate in the COM cohort is encouraging and aligns with current evidence supporting the safety of cochlear implantation following subtotal petrosectomy, with low overall complication rates, most of which are minor (e.g., wound dehiscence, seroma, or hematoma) (Canzi et al., 2023; D'Angelo et al., 2020; Morelli et al., 2025; Szymański et al., 2016; Yan et al., 2020). In this contex, a pivotal technical issue concerns the management of the mastoid cavity, and Eustachian tube, which remain a topic of ongoing debate (Bannister et al., 2020). Traditionally, cavity obliteration combined with secure external auditory canal closure (EAC) has been advocated to achieve a sealed and sterile environment, thereby reducing the risk of postoperative infection and cholesteatoma recurrence (Arístegui et al., 2022). Indeed, Vincenti et al. (2014) reported a 66% (2/3) extrusion rate in cases of cochlear implantation in radical cavity vs. 5.9% (1/17) in subjects with SP and cavity obliteration, confirming the need for SP in patients with SNHL and a canal wall down mastoidectomy.

However, alternative strategies have been proposed (El-Kashlan et al., 2002, 2003) questioned the routine need for middle ear and mastoid cavity obliteration, highlighting potential advantages of a non-obliterative approach, including improved postoperative radiological surveillance for cholesteatoma recurrence and easier access in case of revision surgery. In our series, 15 of 17 patients underwent SP without mastoid obliteration and Eustachian tube closure and no cases of device extrusion or recurrent cholesteatoma were observed during follow-up. These findings in line with previous reports (Hernández et al., 2020; Yaşar et al., 2021), suggest that when meticulous disease eradication and robust multilayered EAC closure are achieved, cochlear implantation can remain safe and effective even without routine obliteration. The only case of cavity infection and EAC closure failure was observed in a chronic suppurative ear in active phase of disease, highlighting the importance of proper eradication of the disease and staging the procedure (Szymański et al., 2016). The absence of COM or cholesteatoma recurrence on long-term imaging supports the efficacy of SP in establishing a stable, dry ear, which is a fundamental prerequisite for successful CI in this population (Zhang et al., 2021). Similar results are reported by Grinblat et al. (2020) with recurrence rates of 0%.

The 17 patients of the COM group received a simultaneous procedure of SP and CI. As reported by Lee et al. (2020), the primary advantage of a simultaneous procedure is that it eliminates the need for multiple surgeries to treat one disease. This significantly reduces the financial cost of multiple surgeries and hospitalizations. Simultaneous procedures also allow for earlier auditory rehabilitation, leading to an earlier improvement in quality of life (QOL). However, when chronic ear disease is active, several authors have highlighted a higher risk of postoperative complications if cochlear implantation is performed in a single stage. The selection of an optimal surgical technique for cochlear implantation in the context of mastoid cavity disease remains a subject of ongoing debate. Although there is no definitive consensus, a two-stage procedure is generally regarded as appropriate in the presence of active infection (Issing et al., 1996; Donnelly et al., 1995; Postelmans et al., 2009). Issing et al. (1998) recommended a two-stage approach to mitigate postoperative complications in cases with existing inflammatory conditions within the surgical cavity. Their findings noted that two of fourteen cases experienced complications, specifically retroauricular fistula, following a single-stage procedure on infected mastoid cavities. Similarly, Leung and Briggs (2007) documented two cases of active infection necessitating revision of the obliteration prior to cochlear implantation in a cohort of 10 patients who underwent staged surgery. In line with these observations, in the present series 1 of 3 patients who underwent single-stage implantation in the setting of active disease developed a postoperative complication, whereas no such events occurred in ears implanted after disease quiescence. Although the small sample size precludes formal statistical testing, the convergence between our findings and the previously published data supports a cautious recommendation in favor of staged surgery in cases of active disease.

The controversy persists concerning the management of patients with inactive middle ear disease, with some advocating for a single-stage approach. Proponents such as Hamzavi et al. (2001) reported satisfactory outcomes with no complications in seven patients who underwent subtotal petrosectomy combined with cochlear implantation in clinically stable mastoid cavities. However, a significant risk associated with single-stage procedures is the development of implantation cholesteatoma, which may remain asymptomatic and undetected postoperatively. Moreover, several reports of patients who have undergone single-stage surgery for inactive disease indicate occurrences of device extrusion, often necessitating revision surgery. In conclusion, the choice between single- and two-stage surgical strategies should be individualized, considering the inflammatory status of the cavity. While staged procedures are favored in ears with active or infected middle ear disease to reduce postoperative complications, single-stage surgery may be appropriate in cases of inactive, clinically quiescent disease, provided that vigilant postoperative monitoring is maintained to detect potential complications such as cholesteatoma formation or device extrusion.

Postoperative audiological outcomes revealed a significant and statistically comparable improvement in hearing thresholds post-implantation in both groups. The near-identical aided PTA₅ results (COM: 36.6 dB vs. Control: 35.6 dB, p = 0.09) underscore the capacity of electrical stimulation to restore auditory input, regardless of middle ear status. The speech perception outcomes further confirmed that the two groups performed comparably in terms of auditory ability. Both in quiet and in noise, results showed no significant differences in disyllabic word recognition or in the signal-to-noise ratios (SNRs) required to achieve 50% intelligibility on the Oldenburg Sentence Test (OLSA) in both groups. Since speech-in-noise perception is a key indicator of real-world listening efficacy and a persistent challenge for the hearing-impaired (Wilson and Dorman, 2008), these findings suggest that the central auditory plasticity engendered by CI activation effectively compensates for pre-existing peripheral deficits (Fallon et al., 2008; Glick and Sharma, 2017), enabling a level of auditory rehabilitation comparable to standard candidates.

Our results align with evidence from large series showing that SP combined with CI grants significant hearing gains. Velasco et al. (2025) reported a mean improvement in pure-tone thresholds from 98.7 ± 9.2 dB HL preoperatively to 39.3 ± 9.3 dB HL postoperatively and open-set sentence recognition from 6.8% to 78.4% (p < 0.001) in patient who underwent simultaneous STP and CI. Likewise, Zhang et al. (2021) demonstrated postoperative benefit after SP and CI, with speech recognition thresholds improving from 80 ± 21 dB to 31 ± 9 dB (p < 0.001). AzBio sentence scores improved from 11% to 43% (p = 0.002), while Consonant–Nucleus–Consonant (CNC) word scores increased from 6% to 47% (p < 0.001) and from 15% to 66%, respectively. Similarly, (Morelli et al., 2025) reported significant postoperative improvements in 348 patients undergoing CI with SP, with mean PTA decreasing from 114 ± 11 dB HL to 48 ± 8 dB HL (p = 0.004) and speech discrimination score increasing from 12 ± 4% to 54 ± 14% (p = 0.008) 6 months after activation. Notably, consistent with our findings, these improvements were considerable and comparable to those achieved with standard cochlear implantation in patients without middle ear disease.

The Speech, Spatial, and Qualities of Hearing Scale (SSQ) results further revealed comparable overall satisfaction and listening performance between the two groups. No significant differences were observed across any SSQ subscale post-implantation, suggesting that CI provides comparable subjective benefit in patients with chronic otitis media and in standard CI recipients. These findings align with the meta-analysis by McRackan et al. (2018), which showed significant improvements in hearing and CI-specific quality of life after implantation, regardless of patient variability or speech outcomes, reflecting the strong psychosocial impact of hearing restoration.

Collectively, these findings underscore that optimal outcomes depend on careful patient selection, complete disease eradication, and secure EAC closure, which together minimize postoperative morbidity and guarantee long-term implant viability.

The current research had several limitations, that should be take into account when interpreting the findings. As a retrospective case-control analysis, it is subject to potential biases associated with non-randomized designs, including selection bias and the influence of unmeasured confounders. The relatively modest sample size, while adequate for detecting clinically relevant differences in primary audiological outcomes, may limit the generalizability of results and preclude definitive statistical conclusions regarding rarer surgical complications. The mean follow-up duration of 36 months, though substantial, may be insufficient to capture very late complications or disease recurrence in this chronic condition. The validation of these findings and the refinement of patient selection criteria and surgical strategies will be enabled by future multicentre studies with larger cohorts, standardized surgical protocols, and extended longitudinal follow-up.