Audiometry-Confirmed Sudden Sensorineural Hearing Loss Incidence among COVID-19 Patients and BNT162b2 Vaccine Recipients

Objective

To compare sudden sensorineural hearing loss (SSNHL) incidence rates over the coronavirus disease 2019 (COVID-19) outbreak and the COVID-19 vaccination campaign periods to pre-COVID-19 periods.

Study Design

Retrospective cohort.

Setting

Secondary hospital.

Patients

Patients >12 years with auditory-confirmed SSNHL were enrolled. COVID-19 status and BNT162 inoculation records ≤28 days before SSNHL diagnosis were retrieved. Patients were categorized according to their date of presentation over four equal periods: 1) July 2018–April 2019 (first prepandemic period), 2) May 2019–February 2020 (second prepandemic period), 3) March 2020–December 2020 (COVID-19 outbreak), and 4) January 2021–October 2021 (BNT162b2 vaccinations campaign).

Interventions

Pre- and post-COVID-19 emergence; BNT162b2 vaccine.

Main Outcome Measures

Incidence rate ratios (IRRs) were calculated to compare SSNHL cases during the COVID-19 and vaccination periods with pre-COVID-19 periods.

Results

Of the 100 patients with SSNHL over the four periods, 1 had COVID-19 and 8 were vaccinated. The annual SSNHL incidence was 12.87, 12.28, 13.45, and 19.89 per 100,000 over periods 1 to 4, respectively. SSNHL incidence over the third period was not significantly different than the first/second periods (IRR = 1.045, 95% confidence interval [CI] = 0.629–1.85, ρ = 0.788, and IRR = 1.095, 95% CI = 0.651–1.936, ρ = 0.683, respectively), whereas SSNHL incidence rate over the fourth period was higher (IRR = 1.545, 95% CI = 0.967–2.607, ρ = 0.068, and IRR = 1.619, 95% CI = 1–2.73, ρ = 0.05, respectively). SSNHL incidence in vaccine recipients was lower than prepandemic unvaccinated patients (IRR = 0.584, 95% CI =0.464–1.67, ρ = 0.984, and IRR = 0.612, 95% CI =0.48–1.744, ρ = 0.92, respectively).

Conclusion

There were fewer SSNHL cases during the first COVID-19 months. Although the SSNHL rate over the COVID-19 vaccination campaign increased, it was not higher for patients who received the BNT162b2 vaccine.

INTRODUCTION

Sudden sensorineural hearing loss (SSNHL) is an otological emergency that needs to be confirmed by audiometry and treated with steroids within a limited time frame after the presentation. It is commonly defined as a hearing threshold decrease of ≥30 dB at three or more consecutive frequencies. Numerous etiologies for SSNHL have been proposed including viral agents (e.g., herpes simplex/zoster), neoplasms, trauma, ototoxicity, autoimmune diseases, vascular/ischemic pathologies, and idiopathic. The annual SSNHL incidence rate in the U.S. ranged between 5 and 20 cases per 100,000 persons in the 1970s and 1980s (1). More recent statistics from the 2000s indicate that the annual rate was as high as 27 cases per 100,000 persons (2).

The coronavirus disease 2019 (COVID-19) is a respiratory disease caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). First identified in December 2019 (3), COVID-19 rapidly spread on a global scale within several weeks. The World Health Organization declared COVID-19 a global pandemic on March 11, 2020.

COVID-19 entails an array of symptoms, most are respiratory (e.g., cough, pneumonia). No hearing-related signs or symptoms associated with COVID-19 were reported in large-scale studies (4). Limited research suggesting an association between COVID-19 and SSNHL was inconsistent. Although some supported a temporal association between acute COVID-19 and SSNHL (5), others have failed to support or even rejected any association (6,7). A large-scale Israeli study showed a significant decrease in non-audiometry-confirmed SSNHL cases over the first COVID-19 year compared with two prepandemic years (2018 and 2019), without examining their COVID-19 status (8).

The first COVID-19-approved vaccine was the mRNA technology-based BNT162b2 (Pfizer-BioNTech). Emergency authorization for ages 16+ in December 2020 initiated massive vaccination campaigns worldwide (9). Adverse effects of the BNT162b2 vaccine include myocarditis, lymphadenopathy, appendicitis, and herpes zoster reinfection, and an insignificant increase in facial palsy was also cited (10). As vaccination campaigns progressed, large population-based studies have been published, reporting on SSNHL as a possible adverse event. Findings from the United States, United Kingdom, and Italy suggest that observed SSNHL incidence in vaccine recipients was similar, if not lower, than unvaccinated populations (11,12). However, many studies exploring the associations between SSNHL and COVID-19 patients and vaccine status relied on self-reports and questionnaire responses without any objective form of proof (audiometric test results), anonymous and incomplete reports to the Centers for Disease Control and Prevention Vaccine Adverse Event Reporting System (13), and calculating SSNHL incidence rate trends by linking hearing loss administrative codes with recent steroid prescription in large deidentified population databases (14). Thus, these studies suffer from recall bias, relying on subjective reports and considerable inaccuracy in (over) identifying SSNHL cases (15).

We sought to study audiometry-confirmed SSNHL cases incidence rates before and during the COVID-19 pandemic, and following the licensure of the BNT162b2 vaccine.

MATERIALS AND METHODS

The study was approved by an Institutional Review Board (blinded for review).

First 2 Years of COVID-19 in Israel

The first case of COVID-19 in Israel was confirmed on February 21, 2020. Because of the rapid spread of COVID-19 cases and the global pandemic declaration by the World Health Organization, Israel began imposing social distancing measures to limit disease spread as of March 11, 2020. Three national lockdowns were imposed during the first pandemic year: March 19, 2020, to May 7, 2020; September 18, 2020, to October 18, 2020; and December 27, 2020, to February 7, 2021. The first lockdown was the strictest, as patients were advised to seek medical attention only in cases of emergency. On December 20, 2020, Israel launched a national vaccination campaign exclusively with the BNT162b2 vaccine to all citizens 16+ years of age (10).

The second COVID-19 year was characterized by a massive nationwide vaccination campaign, but without strict lockdown measures. More than half of the Israeli population had been vaccinated by March 2021, with the COVID-19 incidence rate dropping from 900 cases per million per day (January 2021) to fewer than 2 cases by June 2021. However, after the emergence of new variants (e.g., delta), resurgence of infection and illness occurred around August 2021. On June 21, 2021, the Ministry of Health approved the vaccination for children 12+ years of age and later even for children >5 years. On July 12, 2021, the Israeli authorities approved the administration of a third booster inoculation (16).

Study Hypothesis

SSNHL incidence rates over the COVID-19 outbreak and the COVID-19 vaccination campaign periods were similar to the pre-COVID-19 periods.

Design and Study Sample

Data were retrospectively obtained from the electronic patient records of Samson Assuta Ashdod University Hospital, a secondary care public hospital in the coastal city of Ashdod, in southern Israel. It serves some ~255,000 residents as of April 2021. Of them, 205,000 were 12+ years of age. There have been no substantial changes in the population size during the study period.

SSNHL cases were identified using code 388.2 of the International Classification of Diseases, Ninth Revision (ICD-9), i.e., sudden hearing loss, confirmed by audiometry. To minimize the possibility of exclusion of SSNHL cases because of incorrect coding, we identified and included patients whose clinical history and audiometry supported an SSNHL diagnosis with other ICD-9 codes (i.e., V19.2, deafness or hearing loss; 389, hearing loss; 389.1X, sensorineural/sensory/neural hearing loss, unspecified/unilateral/bilateral/asymmetrical; and 389.2X, mixed hearing loss, unspecified/unilateral/bilateral). We collected age, sex, COVID-19 status, BNT162b2 inoculation status, time from positive SARS-CoV-2 PCR test or BNT162b2 inoculation dates relative to SSNHL diagnosis, treatment, and hearing outcomes after treatment.

We excluded patients with missing data (e.g., audiometry, vaccination dose, or date), patients under 12 years of age, patients with recent acoustic trauma, and any documented previous sensorineural hearing loss.

Based on presentation date, patients were categorized into one of four study groups of roughly 304 days (range, 303–305): first pre-COVID-19 control period (July 1, 2018–April 30, 2019); second pre-COVID-19 control period (May 1, 2019–February 29, 2020); COVID-19 pandemic year, prevaccination (March 1, 2020–December 31, 2020); and COVID-19 pandemic year, vaccination campaign (January 1, 2021–October 31, 2021).

Additional data regarding SARS-CoV-2 cases and BNT162b2 vaccine recipients in Ashdod were obtained from the Ministry of Health’s COVID-19 data center. According to official numbers, there were 14,544 and 26,203 PCR-positive SARS-CoV-2 cases in Ashdod during the third and fourth periods, respectively (overall 40,747 cases). Also, 127,543 persons in total were inoculated with the BNT162b2 vaccine over the fourth period in Ashdod (first, second, and third dose).

SSNHL cases of post-BNT162b2 vaccine were included if occurred within 28 days after inoculation.

Statistical Analysis

Data collected were summarized using frequency tables, descriptive statistics, incidence rate ratios (IRRs), 95% confidence intervals (CIs), and two-tailed mid-p values. A p value ≤0.05 was considered to be statistically significant. Analyses comparing total cases of audiometry-confirmed SSNHL (binary-dependent variable) over each period (nominal nondependent variable) were conducted using annual adjusted IRRs. Incidence rates for COVID-19 vaccine recipient (binary-independent variable) SSNHL cases (binary-dependent variable) during the fourth period to the incidence of SSNHL were compared with Ashdod’s population 12+ years of age over both pre-COVID-19 periods (binary-dependent variable control). Additional covariates, such as age, sex, laterality, treatment, and outcome, were also obtained and are shown in the results.

All statistical analyses were performed using the Statistical Package for the Social Sciences (version 28.0; IBM Corp., Armonk, NY) and WINPEPI (17).

RESULTS

We identified 100 SSNHL patients over periods 1 to 4 (22, 21, 23, and 34, respectively). Demographic statistics are reported in Table 1. Of them, only 1 tested positive for COVID-19 (third period), and 8 were vaccinated ≤28 days before their presentation (fourth period; 2 presented after the first dose, 5 after the second dose, and 1 after the third dose; mean day after inoculation, 13). Over periods 1 to 4, annual adjusted SSNHL incidence rates were 12.87, 12,38, 13.45, and 19.89 per 100,000 persons, respectively (of 205,123 residents 12+ years). For BNT162b2 vaccine recipients, the annual adjusted SSNHL incidence rate was 7.52 per 100,000 vaccine recipients in our study population (of 127,543 residents 12+ years vaccinated over the fourth period only).

TABLE 1.

Clinical and demographic data for patients presenting with SSNHL

	Prepandemic Period 1 (07/2018–04/2019) (N = 22)	Prepandemic Period 2 (05/2019–02/2020) (N = 21)	Postpandemic Period 3 (03/2020–12/2020) (N = 23)	Postpandemic Period 4 (01/2021–10/2021) (N = 34)
Age, years
Mean ± SD	54.90 ± 17.57	48.90 ± 19.12	55.95 ± 17.15	49.76 ± 15.91
Median (range)	52 (23–81)	47.5 (20–75)	51 (20–82)	49.5 (17–82)
Sex
Male, n (%)	6 (27.3)	10 (47.6)	11 (47.8)	15 (44.1)
Female, n (%)	16 (72.7)	11 (52.4)	12 (52.2)	19 (55.9)
Laterality
Left, n (%)	10 (45.5)	11 (52.4)	8 (34.8)	19 (55.9)
Right, n (%)	9 (40.9)	9 (42.9)	14 (60.9)	14 (41.2)
Bilateral, n (%)	3 (13.6)	1 (4.8)	1 (4.3)	1 (2.9)

Table 2 shows adjusted annual SSNHL IRRs per study periods. Although the annual SSNHL incidence rate during the fourth period was significantly higher than the first and second periods (IRR = 1.54, 95% CI = 0.96–2.60, ρ = 0.06, and IRR = 1.61, 95% CI = 1–2.73, ρ = 0.05, respectively), the annual SSNHL incidence rate during the third period was not significantly higher than either pre-COVID-19 period (IRR = 1.04, 95% CI = 0.62–1.85, ρ = 0.78, and IRR = 1.09, 95% CI = 0.65–1.93, ρ = 0.68, respectively).

TABLE 2.

IRRs for SSNHL, per study period

	Observed Cases	Expected Cases per Yeara	Incidence per 100,000 people per entire year	Compared with First Interval			Compared with Second Interval
				Incidence Rate Ratio	95% CI	p	Incidence Rate Ratio	95% CI	p
Period 1 (07/2018–04/2019)	22	26.4	12.87	Reference
Period 2 (05/2019–02/2020)	21	25.2	12.28
Period 3 (03/2020–12/2020)	23	27.6	13.45	1.045	0.62–1.85	0.78	1.09	0.65–1.93	0.68
Period 4 (01/2021–10/2021)	34	40.8	19.89	1.54	0.96–2.60	0.06	1.61	1–2.73	0.05

Bolded with p value ≤0.05 was considered to be statistically significant.

^aObserved cases per 10-month interval were multiplied by 12 and divided by 10 for the annually expected case rate.

Table 3 shows adjusted annual SSNHL IRRs between unvaccinated patients and those vaccinated ≤28 days before their presentation. The annual adjusted SSNHL incidence rate for BNT162b2 vaccine recipients of 7.52 cases per 100,000 persons was insignificantly lower than unvaccinated patients over the first and second periods of 12.87 and 12.28 cases per 100,000 persons (IRR = 0.58, 95% CI = 0.46–1.67, ρ = 0.98, and IRR = 0.61, 95% CI = 0.48–1.74, ρ = 0.92, respectively).

TABLE 3.

IRRs for SSNHL, vaccinated versus unvaccinated subjects

	Observed Cases	Expected Cases per Yeara	Incidence per 100,000 People per Year	Compared with First Interval			Compared with Second Interval
				Incidence Rate Ratio	95% CI	Two-Tailed Mid-p	Incidence Rate Ratio	95% CI	Two-Tailed Mid-p
Unvaccinated, Period 1 (07/2018–04/2019)	22	26.4	12.87	Reference
Unvaccinated, Period 2 (05/2019–02/2020)	21	25.2	12.28
Vaccinated,b Period 4 (01/2021–10/2021)	8	9.6	7.52	0.58	0.46–1.67	0.98	0.61	0.48–1.74	0.92

^aObserved cases per 10-month interval were multiplied by 12 and divided by 10 for the annually expected case rate.

^bVaccinated ≤28 days before SSNHL diagnosis.

When treatment for SSNHL was considered, 81 charts were available (19 with missing data). Of them, 61 (75.3%) were treated with oral steroids, 2 (2.5%) with intratympanic (IT) steroids, 16 (19.8%) with oral and IT steroids, and 2 (2.5%) with oral steroids, IT steroids, and hyperbaric oxygen treatment. Treatment for SSNHL was consistent over the third and fourth periods compared with the first and second periods. We were able to analyze the outcome of 80 patients (1 with missing data). Of them, 38 patients (47.5%) had subjectively and objectively favorable outcome (i.e., bone conduction thresholds), 8 patients (10%) had subjectively favorable outcome without audiometric validation, and 34 patients (42.5%) had no subjective nor objective improvement. No major changes were observed in treatment outcomes for SSNHL patients over all time periods.

DISCUSSION

We found no significant change in SSNHL incidence over the first 10 months of the COVID-19 pandemic compared with the years immediately prior, but there was a significant increase in SSNHL incidence during the COVID-19 vaccine campaign period. These findings are in line with existing research conducted in Israel (8,14). Only one patient with documented SSNHL tested positive for COVID-19 of 40,747 known cases in the region over the first 10 pandemic months. SSNHL affected only 8 of 127,543 recipients of the BNT162b2 vaccine, who presented on the median 13th day postinoculation; this is very low and thus unrelated to the vaccine.

There were few SSNHL cases over the first 10 months of the pandemic (study period 3). This period was characterized by strict lockdowns; calls to stay at home in noncritical cases; increased use of telehealth consultations, which were preferred over frontal visits; and closure of ambulatory services (e.g., hearing institutes). All of these objective measures may have delayed the identification, timely treatment, and diagnosis of SSNHL on time. Hearing tests were not routinely conducted, and confirmed COVID-19 patients were even less likely to be sent to undergo audiometric studies. At-home and mobile SSNHL tests have been developed and seem to be reliable but could not eventually replace an objective audiometry (18).

The upswing in SSNHL incidence over the fourth period (vaccines campaign) might be associated with mass administration of the BNT162b2 vaccine. However, this is not supported by the low incidence of SSNHL among vaccine recipients compared with unvaccinated patients over both prepandemic periods in our study population, which is in line with a recent U.S. publication that did not present an association between COVID-19 vaccination and an increased SSNHL incidence when compared with the general population (13). We explain this increase to the reopening of medical facilities, including hearing institutes, also for recovered COVID-19 patients and diagnosed COVID-19 patients, and a concomitant increase in frontal visits while minimizing telehealth consultations, which could explain the delay in SSNHL diagnoses over the third (prevaccination) period but not over the fourth period.

Several theories regarding the role of SARS-CoV-2 in the pathogenesis of SSNHL have been suggested. First, acute SARS-CoV-2 infection increases hypercoagulability states, leading to the formation of parenchymal macro/micro thromboses, which might result in ischemia and hypoxia within the auditory pathway, and more specifically in the cochlea, which is supplied by an end-artery (19). Second, is the expression of the angiotensin converting enzyme-2 (ACE-2) receptors within the nasopharynx, middle/inner ear, and mastoid that enables SARS-CoV-2 penetration into the middle ear and mastoid, but not the inner ear (20,21). To our knowledge, SARS-CoV-2 has yet to be directly extracted from the inner ear despite the presence of the ACE-2 receptors. Nevertheless, we identified only 1 patient with SSNHL who also tested positive for COVID-19. This finding is in accord with other studies that reported only a few coexisting COVID-19 and SSNHL cases, as shown in Supplemental Tables 1 and 2, http://links.lww.com/MAO/B546, for audiometry-confirmed and non-audiometry-confirmed SSNHL, respectively. Concurrence is infrequent relative to the global COVID-19 pandemic.

COVID-19 vaccines have also been examined relative to SSNHL. Regarding vaccines other than BNT162b2, a Turkish study reported increased hearing loss and tinnitus rates among CoronaVac recipients. However, these complaints were subjective and based on questionnaires distributed to health workers and were not confirmed by audiometry (22). Published SSNHL cases post-COVID-19 inoculations are shown in Supplemental Tables 3 and 4, http://links.lww.com/MAO/B546, for audiometry-confirmed and non-audiometry-confirmed SSNHL, respectively.

Advantages of this study include the following: 1) we identified only audiometry-confirmed SSNHL cases, and 2) we applied rigid inclusion and exclusion criteria to accurately capture only true SSNHL cases. This is reflected in disease incidences within our study population, which are well within the SSNHL incidence’s reported range (5–27/100,000) (1,2). Of note, the SSNHL rates in our study population are lower than other Israeli studies (41.5, 60.7, and 56.2 per 100,000 person-years for the reference years, first and second vaccine doses, respectively) (14), and 3) we systematically calculated incidence from credible resources based on exact population size and medical data, number of COVID-19 patients, and BNT162b2 vaccine recipients.

We acknowledge some limitations: 1) our absolute SSNHL case numbers are low, although reasonable for a secondary care hospital; 2) we also assumed that all Ashdod residents sought medical treatment exclusively in their local hospital for this emergency, although they may have been referred or sought treatment elsewhere. Of course, this would be true for the prepandemic periods as well, and 3) after its emergence, COVID-19 has changed its epidemiological dynamics, most notably during the fourth period with the delta variant as the predominant mutation. Its infectious viral dynamics may have led to an increase in SSNHL cases compared with the third period. This limitation is true for viruses universally, which undergo frequent changes or may be seasonally affected. Moreover, it was not feasible to differentiate the COVID-19 serotype for the purpose of this study.

CONCLUSION

COVID-19 infection was not associated with audiometrically confirmed SSNHL cases. Despite the increase in SSNHL cases during the period when vaccination with BNT162b2 had taken place, there was no increase in audiometrically confirmed SSNHL cases among these vaccine recipients. The association between COVID-19 disease and its vaccination warrants further large-scale, audiometrically confirmed research conducted over many years postpandemic.