Cochlear implantation under local anesthetic: A systematic review and meta-analysis

2.1 Study registration

This study was registered on the PROSPERO database of systematic reviews (reference number CRD42021251617). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed.¹¹

2.2 Inclusion criteria

Studies for inclusion were clinical studies of cochlear implantation under LA where audiological, surgical, postoperative, or quality of life outcomes at any postoperative time were reported. Studies included were experiential or observational of human subjects. Animal or laboratory studies were excluded, as were studies focusing on surgical techniques with no human subjects.

2.3 Study search strategy

BW performed the comprehensive searches using the search terms for MEDLINE as shown in Figure 1. Search terms were adapted for PubMed, CINAHL, and EMBASE. Searches for these databases were conducted using the National Institute for Health and Care Excellence Healthcare Databases Advanced Search (Health Education England). Web of Science Core Collection, the Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Trial, and ClinicalTrials.gov were searched individually and exact search terminology for each database can be found in Appendix. There were no restrictions on language or date. These searches were rechecked by a clinical effectiveness librarian to ensure complete coverage of all available literature.

2.4 Study selection

Two reviewers (BW/PG) independently screened titles and abstracts. Full texts of studies potentially meeting the inclusion criteria were then reviewed by BW/PG to assess eligibility for inclusion. Where a full text could not be found, authors were contacted directly via email, if they remained unavailable, the study was excluded. Conflicts between reviewers were resolved by discussion. Reference lists of screened full texts were examined and reviewed for potential additional studies which met the inclusion criteria. The primary outcome measures were surgical technique, postoperative complications audiological outcomes. Secondary outcome measures were quality of life, comorbidities precluding GA, postoperative inpatient stay, and preoperative measures to aid postoperative recovery.

2.5 Data extraction

Data were independently extracted and checked by both BW and PG using a predetermined spreadsheet (Microsoft Excel).

2.6 Risk of bias analysis

Risk of bias of selected studies was completed using the Brazelli risk of bias tool¹² for nonrandomized studies and the Oxford Centre for Evidence-Based Medicine (OCEBM) grading system.¹³ Biases were independently assessed and compared by both BW and PG using a spreadsheet (Microsoft Excel), with conflicts resolved by discussion. Where conflicts arose which could not be resolved by discussion, a senior reviewer (J. M.) provided a final decision.

3.1 Search results

Searches were performed on April 25, 2021 and results of screening are shown in Figure 2. The initial search identified 254 studies. After removing duplicates, 132 full titles and abstracts were screened, of which 24 articles were shortlisted for full-text review. Six articles did not meet the inclusion criteria and were therefore excluded, leaving 18 articles for qualitative analysis. The heterogenicity of study design and outcome reporting precluded a quantitative analysis for outcomes other than operative time.

3.2 Characteristics of included studies

Study characteristics are displayed in Table 1. Included studies consisted of eight case series, five case controls and five case reports from 1998 to 2021. There were a total of 450 participants, of which 262 (58%) received cochlear implants under LA and 188 (42%) received cochlear implantation under GA. From available demographic data, there were 241 males and 209 females (with a male predominance of 1:1.5). The average age of implantation under LA was 59.9 years (range: 13–93). A single study reviewed cochlear implantation in children, others compared outcomes between adults. The implant models used were: Med-El (Synchrony, Flex 28, Concerto, Sonata, Form 24); Advanced Bionics (HiFocus 1J, HiFocus Mid-Scala); and Cochlear (CI532, Freedom Contour Advance, Freedom, N512, CI632, N22, CI422, CI24M, Contour advance CI24RE, Nucleus CI522). Full electrode insertion was achieved by the majority of authors, with only three implants by two authors experiencing partial electrode implantation.^{14, 15}

3.3 Comorbidity grading

A total of 242 participants across seven articles had an American Society of Anesthesiologists (ASA) Physical Status grade identified. Of 161 with reported ASA grades who received cochlear implantation under LA, the majority were ASA II (n = 82) and III (n = 64), followed by ASA IV (n = 9) and I (n = 6). Of the 81 receiving cochlear implantation under GA with an identified ASA grade, none were ASA IV, and the majority were ASA II (n = 48) and III (n = 31), with ASA I accounting for only two participants. In cases where a range of ASA grading was reported, the lowest ASA grading was taken.

Four studies (n = 43 participants)^{14, 16-18} used the Portsmouth Physiological and Operative Severity Score for the enumeration of Mortality and morbidity (P-POSSUM) scoring tool, a tool which was developed by Copeland et al.¹⁹ which predicts mortality in patients undergoing different types of surgery. A combined 32 participants in the LA groups of these studies had an average predicted morbidity of 46.38% (10.2–93.6%) and average predicted mortality of 6.76% (0.4–36%), compared with 39.16% (18.5–70.3) and 2.94% (0.8–7.1%) of 11 participants who underwent CI under general anesthesia (GA), respectively. A single case report utilizing the surgical outcome risk tool calculated their patient's predicted risk of 1%, with a frailty score of four.¹⁴

3.4 Audiological outcomes

Audiological outcomes were reported in nine studies (113 patients) and are presented in Table 2. Overall, there was a trend toward benefit from cochlear implantation under LA.

Preoperative data were available for seven studies, of which every patient demonstrated severe to profound sensorineural hearing loss through different audiological outcome measures. Pure tone average (PTA) was the most used (n = 6) method of assessing audiological function. Speech discrimination was assessed in seven studies, with Bamford–Kowal–Bench (BKB) sentences and Arizona State University (AzBio) sentences commonly used, in two instances best-aided conditions were used. Otoacoustic emission, Hearing In Noise Test (HINT) scoring, Categories of Auditory Performance (CAP) scoring, listening progress profiles and telephone profile testing were each utilized preoperatively once.

Of the eight studies where postoperative audiological function testing is recorded, only six compare these with preoperative results. Of these six articles, all patients with implants implanted under LA showed improvement from preoperative audiological function testing. Similar to preoperative testing, multiple heterogenous methodologies were employed to assess audiological function postoperatively, as identified in Table 2.

3.5 Surgical techniques and outcomes

Surgical incision and approach was reported in 14 studies, a post auricular (standard) approach was used in 12 studies,^{10, 14-16, 18, 20-26} with a posterior suprameatal approach used in two further studies.^{21, 27} Facial nerve monitoring was only reported for one patient undergoing local anesthetic-conscious sedation (LA-CS).²⁰ Procedures under LA were largely successful, with only one patient requiring conversion to GA,²⁸ although the reasons for converting from LA-CS to GA were not specified. Sedation agents varied greatly and were reported in 16 studies.^{10, 14-18, 20-29} Methods of sedation and nausea control used were propofol (n = 5), dexmedetomidine (n = 6), fentanyl (n = 7), midazolam (n = 11), remifentanil (n = 4), sufentanil (n = 2), hydroxyzine (n = 1), meperidine (n = 2), and clonidine (n = 2). The majority of the 16 reported LA regimens used lidocaine (n = 14), with all except one regimen also using adrenaline/epinephrine.²⁵ Alternative agents to lidocaine included bupivacaine (n = 2) and levobupivacaine (n = 1).

Of six studies which compared operative times between LS-CS and GA,^{10, 15, 18, 23, 28, 29} four examined the statistical relationship between duration of both procedures.^{10, 15, 23, 29} Of these, three found a statistically significant difference with a p < .05, and the other by Hamerschmidt et al.¹⁰ approached significance. Surgery times for LA-CS compared with GA were: 61.09 compare 69.05 (p = .046),²⁹ 54.53 compare 79.50 (p = .001),²³ and 86.7 compare 122.3 min (p = .00002).¹⁵ In a study by Hamerschmidt in 2012, a Wilcoxon test was performed to compare time of surgery, with a W result of 65.5 this was not significant, but did approach significance, at a p value of .05.¹⁰ Meta-analysis using a random effects model showed the mean duration of surgery to be shorter in the LA group by 22.59 min (CI −30.80 to −14.37, p < .00001). The forest plot for this is displayed in Figure 3 and the funnel plot in Figure 4.

Adverse events during or in the postoperative period following cochlear implantation under LA were rare. There were no major adverse events. Of 33 instances of minor adverse events, the most common experienced was vertigo (n = 17), followed by postoperative wound infection (n = 6), temporary facial nerve paralysis (n = 4), confusion (n = 4), residual/new tinnitus (n2). There were no instances of taste disturbance, and no patients required a repeat or revision operation.^{10, 15, 22, 23, 28-30}

Postoperative inpatient hospitalization was compared between groups in six studies and is shown in Table 3. Five studies statistically analyzed differences in inpatient hospitalization following participants procedure, of which three were statistically significant differences.

3.6 Risk of bias of included studies

All studies were retrospective with all studies either single case reports, case series or case–control studies with OCEBM grade IV-V. Significant heterogenicity existed between study populations and reporting of primary and secondary outcome measures. Risk of bias assessment is presented in Figure 5. The majority of studies had a high risk of bias. No studies presented a representative sample from a relevant population, the majority of studies did not state clear inclusion or exclusion criteria for entry into their study, with appropriate prognostic and comorbid factors identified. In the majority of studies, patients were not selected consecutively and data collection was not undertaken prospectively, or this was unclear. Most studies did not disclose the qualifications, experience or facilities. In the majority of instances, follow-up was minimal or length of follow-up was unspecified.