Abstract

Objective:

This systematic review and meta-analysis aimed to identify prognostic factors influencing hearing recovery in children with sudden sensorineural hearing loss (SSNHL), thereby providing a reference for developing prevention strategies and optimizing therapeutic interventions.

Methods:

We conducted a comprehensive search of PubMed, Web of Science, Cochrane Library, Embase, CNKI, and SinoMed databases to identify relevant studies published prior to 2024. Inclusion criteria encompassed pediatric patients with SSNHL (<18 years old) who had clearly documented audiological outcomes. We extracted data on demographic characteristics, clinical features, and treatment regimens to investigate the association between prognostic factors and hearing recovery.

Results:

This meta-analysis included 1,006 patients from 13 studies. The combined rates for partial and complete hearing recovery were 19.6% [95% confidence interval (CI): 13.6–25.7%, I² = 80.63%] and 18.6% (95% CI: 12.2–25.1%, I² = 85.99%), respectively. Better prognosis was associated with unilateral hearing loss [odds ratio (OR): 2.741, P = 0.004], age >12 years (OR: 1.911, P = 0.034), treatment delay ≤ 14 days (OR: 6.402, P < 0.001), and an ascending audiogram curve (OR: 6.910, P < 0.001). Conversely, poorer outcomes were associated with profound hearing loss curves (OR: 0.376, P < 0.001) and initial pure-tone average (PTA) thresholds >80 dB HL (OR: 0.451, P = 0.005).

Conclusion:

Early intervention (≤14 days) and targeted treatment for profound hearing loss (PTA > 80 dB HL) hold promise for improving pediatric outcomes. Systemic inflammatory markers (neutrophil-to-lymphocyte ratio) and age-specific pathophysiological mechanisms warrant further investigation. Conducting multicenter prospective studies is essential to validate these findings and establish standardized diagnostic and treatment protocols for pediatric SSNHL.

Systematic Review Registration:

https://www.crd.york.ac.uk/PROSPERO/view/CRD42024612409, CRD42024612409.

Introduction

The clinical diagnostic criteria for sudden sensorineural hearing loss (SSNHL) vary slightly across different national guidelines, summarized as follows: (1) The German guidelines define SSNHL as “sudden onset (typically unilateral) sensorineural hearing loss,” while the American and Chinese guidelines specify “sudden sensorineural hearing loss occurring within 3 days.” (2) According to the Chinese guidelines, hearing loss is defined as ≥20 dB HL in at least two adjacent frequencies, while the American guidelines define it as ≥30 dB affecting at least three consecutive frequencies. SSNHL is predominantly unilateral, although bilateral cases may occur simultaneously or sequentially, as noted in German, American, and Chinese guidelines. (3) All three guidelines agree that SSNHL has no identifiable systemic or local etiology. (4) SSNHL may be accompanied by tinnitus, ear fullness, vertigo, nausea, or vomiting (German, American, and Chinese guidelines). (5) The German guidelines also note that hyperacusis, hyperesthesia, and percutaneous sensory abnormalities may be present (39). According to the United Nations Convention on the Rights of the Child, a child refers to any person under the age of 18. Therefore, SSNHL in children refers to sudden hearing loss occurring in patients aged 0–18 years (2). In the Aachen region of Germany, among a population of approximately 250,000, the incidence rate of idiopathic SSNHL in children is one in 10,000. This rate is only one-tenth to one-twentieth of the incidence rate observed among adults during the same period, which is significantly lower than that observed in adults (3). According to data from the U.S. Medical and Pharmaceutical Claims Database, those under 18 years account for 6.6% of SSNHL cases, those under 14 years account for 3.5%, and those under 9 years account for only 1.2% (4). This is a rare but serious condition that may adversely affect the social adaptation, behavior, and psychological development of children during their formative years. At present, no specific epidemiological studies exist for SSNHL in the pediatric population. Despite the significant impact of this disorder, its rarity and the limitations of existing research have resulted in an insufficient description of prognostic factors for SSNHL in children.

While the precise etiology remains elusive, viral infections, problems with blood flow in the inner ear, and immune system issues are considered predominant causes of pediatric SSNHL based on existing evidence (5, 6). Systemic glucocorticoids are predominant, with intratympanic steroid injection (ITS) used as salvage therapy (7). Bilateral SSNHL is more often associated with suspected etiologies, such as inner ear malformations and viral infections (8).

Due to the low incidence of SSNHL in children, it is challenging to conduct prospective clinical studies to identify its prognostic factors. In addition, most studies investigating prognostic factors for SSNHL are retrospective in nature, characterized by small sample sizes and varying definitions of hearing recovery. Previous meta-analyses have been conducted on SSNHL in children, but the included studies were limited to English-language literature, and some research was outdated (9). Therefore, we conducted a systematic review and meta-analysis specifically focused on prognostic factors for SSNHL in children. This study included relevant literature published in Chinese and English journals, aiming to systematically review related research and identify factors significantly associated with hearing recovery in children with SSNHL.

Materials and methodsSearch strategy

We conducted a systematic search of articles published prior to 2024 in PubMed, Web of Science, Cochrane Library, Embase, CNKI, and SinoMed databases, and screened the reference lists of relevant articles. The included studies were case series and cohort studies focusing on SSNHL in patients under 18 years of age. Search strategies were developed for each database incorporating keywords and Medical Subject Headings (MeSH) terms related to “sudden deafness,” “sudden sensorineural Hearing Loss,” “deafness, sudden,” “children,” and “child”. The detailed PubMed search strategy is provided in Figure 1. Two reviewers (ZW and PZ) conducted an initial assessment of eligible studies through a double-blind independent review process. Any discrepancies were resolved through discussion or consultation with a third reviewer (JS or MD).

PRISMA flow diagram of literature screening and inclusion process.

Data collection

Inclusion criteria: Articles were required to meet all of the following conditions:

Document audiological characteristics;

Report clinical outcomes in pediatric and adolescent populations (0–18 years);

Focus on patients diagnosed with SSNHL;

Be peer-reviewed and published in 2015 or later (the year the Chinese Society of Otolaryngology-Head and Neck Surgery revised its clinical guidelines for SSNHL).

Exclusion criteria: Studies were excluded if they met any of the following considerations:

Non-research publications, including case reports, reviews, and editorial commentaries;

Inadequate methodological quality according to the MINORS criteria;

Failure to provide stratified data analysis for pediatric and adult populations;

Conference abstracts, unpublished data, and non-peer-reviewed materials.

Conflict resolution and screening process:

Any disagreements in study selection were resolved through consensus discussion.

The reference lists of included articles underwent manual cross-checking.

Only full-text, peer-reviewed journal articles were ultimately considered for inclusion.

The systematic review extracted the following variables from eligible studies:

Study methodology

Publication details (year, country);

Study design (prospective/retrospective, controlled/uncontrolled);

Inclusion and exclusion criteria;

Outcome classification system (hearing recovery grades: no recovery, partial recovery, complete recovery);

Sample size (number of participants and/or affected ears).

Demographic characteristics

Gender distribution;

Age at diagnosis.

Audiological parameters

Unilateral vs. bilateral involvement;

Baseline hearing thresholds (pretreatment);

Associated symptoms (tinnitus, vertigo, ear fullness);

Audiogram configuration (specific frequency hearing patterns).

Treatment characteristics

Use of systemic corticosteroids, intratympanic injections, or combination therapy;

Time interval from symptom onset to treatment initiation.

Outcomes assessment

The primary outcomes were categorized as no recovery, partial recovery, or complete recovery. The methodological quality (risk of bias) of included observational studies was assessed using the Methodological Index for Non-Randomized Studies (MINORS) instrument (11). Following the approach described in a previous study (12), we categorized quality as follows: non-comparative studies—very low (0–4), low (5–8), moderate (9–12), and high (13–16); comparative studies—very low (0–6), low (7–12), moderate (13–18), and high (19–24) This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (13).

Statistical analysis

We conducted a meta-analysis using the SPSS AU online platform. For categorical variables, pooled rates with 95% confidence intervals (CI) were reported. For continuous variables, means with 95% confidence intervals were calculated. Heterogeneity was assessed using the I² statistic. When I² = 0, a fixed-effects model was applied; when I² > 0, a random-effects model was employed (10).

ResultsStudy characteristics

Figure 1 (PRISMA flow diagram) illustrates the literature screening process, with a total of 13 studies meeting the inclusion criteria. One study, a doctoral dissertation, was excluded because part of its data had already been published in 2018. Another study was excluded for failing to adequately stratify data between children and adults. Of the included studies, nine were case series (rated as Level 4 evidence according to the Oxford Centre for Evidence-Based Medicine grading system) and four studies were cohort studies (rated as Level 3 evidence according to the Oxford Centre for Evidence-Based Medicine grading system). The MINORS scores for the included studies are presented in Table 1. The ideal MINORS score is 16 points for case series and 24 points for cohort studies; however, none of the included studies achieved the maximum score. These 13 included studies collectively enrolled 1,022 patients. However, three studies did not report the number of ears with hearing loss, making it impossible to calculate the exact total number of ears included. Five studies explicitly specified the duration of follow-up.

Author and yearCountry Study typePatientsTotal ears analyzedFollow-up earsFollow-up (months)No recovery (ears), n (%)Partial recovery (ears), n (%)Complete recovery (ears), n (%)MINORS scoreXu and Fu, 2024 (17)ChinaCase series109NRNRNR51 (51.4)43 (39.4)10 (9.2)6Xiao et al., 2023 (21)ChinaCase series251317317NR222 (70.03)54 (17.04)41 (12.93)9Lu et al., 2019 (26)ChinaCohort252626NR9 (34.6)10 (38.4)7 (26.9)13Ha et al., 2019South KoreaCohort42NRNRNR24 (57.1)11 (26.2)7 (16.7)17Qian et al., 2018 (16)ChinaCase series757878NR22 (29.3)21 (28.0)32 (42.7)14Kim et al., 2018 (20)South KoreaCase series6767672.5 (1–3)30 (44.8)13 (19.4)24 (35.8)13Chen et al., 2018 (8)ChinaCase series1011291297.4 (4–12)95 (73.6)13 (10.1)21 (16.3)10Li et al., 2016 (22)ChinaCase series136151151NR94 (62.2)43 (28.5)14 (9.3)7Kizilay and Koca Ç, 2016 (23)TurkeyCase series1416166.8 (1–20)13 (81.25)0 (0.00)3 (18.75)7Dedhia and Chi, 2016 (1)USACase series66627 (1–83)NRNR0 (0.0)6Li et al., 2015 (15)ChinaCase series1011131130.61 (1–6)72 (63.7)30 (26.6)11 (9.7)8Chung et al., 2015 (18)South KoreaCohort373737NR10 (27.0)9 (24.3)18 (48.6)11Ao et al., 2015 (14)ChinaCohort444444NR17 (38.6)9 (20.4)14 (31.8)11

Characteristics of included studies.

MINORS, methodological index for non-randomized studies; NR, not reported.

The definitions of hearing recovery adopted in the included studies are summarized in Table 2. Among these, four studies used the criteria from the Chinese Medical Association's Guidelines for the Diagnosis and Treatment of SSNHL, defining hearing recovery as follows: no recovery (hearing improvement <15 dB HL), partial recovery (hearing improvement >15 dB HL), and complete recovery (return to normal or pretreatment levels) (14–17). Six studies defined hearing recovery using Siegel criteria as follows: no recovery [pure-tone average (PTA) > 45 dB HL], partial recovery (PTA between 25 and 45 dB HL with ≥15 dB improvement), and complete recovery (PTA < 25 dB HL) (8, 18–21). One study defined hearing recovery as follows: no recovery (hearing improvement <15 dB HL), partial recovery (hearing improvement >15 dB HL), or complete recovery (PTA<25 dB HL) (22). Another study categorized hearing recovery as follows: no recovery (no change), partial recovery (undefined), or complete recovery (restoration to the level of hearing in the unaffected ear) (23). One of the studies did not report specific details on the definition of hearing recovery (1).

Author and yearNo recoveryPartial recoveryComplete recoveryXu and Fu, 2024 (17)<15 dB gain>15 gainNormal/pretreatment levelXiao et al., 2023 (21)PTA > 45 dB HLPTA 25–45 dB HL with ≥15 dB gainPTA < 25 dB HLLu et al., 2019 (26) PTA > 45 dB HLPTA 25–45 dB HL with ≥15 dB gainPTA < 25 dB HLHa et al., 2019 (19)PTA > 45 dB HLPTA 25–45 dB HL with ≥15 dB gainPTA < 25 dB HLQian et al., 2018 (16)<15 dB gain>15 dB gainNormal/pretreatment levelKim et al., 2018 (20)PTA > 45 dB HLPTA 25–45 dB HL with ≥ 15 dB gainPTA < 25 dB HLChen et al., 2018 (8)PTA > 45 dB HLPTA 25–45 dB HL with ≥15 dB gainPTA < 25 dB HLLi et al., 2016 (22)<15 dB gain>15 dB gainPTA < 25 dB HLKizilay and Koca Ç, 2016 (23)No changeNot reportedRestore unaffected ear levelDedhia and Chi, 2016 (1)Not reportedNot reportedNot reportedLi et al., 2015 (15)<15 dB gain>15 dB gainNormal/pretreatment levelChung et al., 2015 (18)PTA > 45 dB HLPTA 25–45 dB HL with ≥15 dB gainPTA <25 dB HLAo et al., 2015 (14)<15 dB gain>15 dB gainNormal/pretreatment level

Definitions of hearing recovery criteria used in included studies.

A total of 217 unique records were identified for this study. After removing duplicates and screening titles and abstracts, 31 studies were selected for full-text review. Based on predefined inclusion criteria, 18 studies were excluded, including seven studies published before 2015 and one study that did not report separate data for children and adults (24). One doctoral dissertation containing previously published data and nine studies with suboptimal MINORS were also excluded (25).

Demographic

We conducted a meta-analysis of the basic demographic characteristics of SSNHL in children. We analyzed two studies measuring gender distribution by number of affected ears and 11 case-based studies, revealing that 54.3% of hearing loss cases occurred in boys (95% CI: 50.8%–57.8%, I² = 0.00%). Regarding age characteristics, a pooled analysis of four studies categorizing age into ≤12 years versus 13–18 years estimated that 38.2% belonged to the younger age subgroup (95% CI: 8.8%–67.5%, I² = 96.20%), while 61.8% were in the older subgroup (95% CI: 32.5%–91.2%, I² = 96.20%). A continuous variable meta-analysis of five studies yielded an overall mean age of 13.46 years (95% CI: 12.47–14.45, I² = 70.95%). Lateralization analysis from 12 studies indicated unilateral hearing loss in 85.2% of cases (95% CI: 78.2%–92.3%, I² = 94.57%) (Table 3).

FactorNo. studiesMeta-analyzed prevalence or mean95% CII2 (%)Unit of analysisMale110.5430.508, 0.5780.000Per patientFemale110.4570.422, 0.4920.000Per patientAge, year ≤1240.3820.088, 0.67596.20Per patient 13–1840.6180.325, 0.91296.20Per patientUnilateral120.8520.782, 0.92394.57Per earInitial PTA, dB HL <4040.1110.052, 0.17078.98Per ear 41–6040.1360.052, 0.21988.64Per ear 61–8040.1980.140, 0.25661.11Per ear >8040.5450.453, 0.63876.88Per earAudiogram Ascending40.1680.056, 0.28082.43Per patient Descending40.160.081, 0.23962.92Per patient Concave/convex2NoneNoneNone Flat40.2660.158, 0.37472.59Per patient Profound40.2850.190, 0.38061.29Per patientAdditional symptoms Tinnitus80.5670.407, 0.72793.16Per patient Vertigo80.2180.131, 0.30681.94Per patient Fullness40.3100.118, 0.50288.83Per patient Systemic steroid70.9490.906, 0.99292.08Per patient ITS40.2850.141, 0.42982.75Per patientTime to treatment, days ≤730.7890.736, 0.8420.000Per patient Partial improvement100.1960.136, 0.25780.63Per ear Total improvement100.1860.122∼0.25185.99Per earContinuous variables Age, year513.4612.470, 14.45170.95Per patient Initial PTA, dB HL372.15559.324, 84.98783.73Per patient Treatment delay, days35.7523.500∼8.00374.45Per patient

Characteristics of pediatric SSNHL.

Audiological characteristics

We conducted a meta-analysis to evaluate the initial degree of hearing loss, baseline hearing thresholds, configuration of audiometric curves, and associated symptoms. Hearing loss severity was categorized into four levels based on the initial pure-tone average threshold (measured in decibels): <40, 41–60, 61–80, and >80 dB HL. This meta-analysis included four studies. The prevalence of comorbidities among patients with baseline PTA <40 dB HL was 11.1% (95% CI: 5.2%–17.0%; I² = 78.98%), while 13.6% (95% CI: 5.2%–21.9%; I² = 88.64%) had an initial PTA of 41–60 dB HL. Moreover, 19.8% of patients with an initial PTA of 61–80 dB HL (95% CI: 14.0%–25.6%; I² = 61.11%) had this condition, while the highest prevalence was observed in patients with an initial PTA >80 dB HL, reaching 54.5% (95% CI: 45.3%–63.8%) (Table 3).

Three studies reported baseline hearing thresholds expressed as mean and standard deviation. The pooled mean initial hearing threshold across these studies was 72.155 dB HL (95% CI: 59.324–84.987 dB; I² = 83.73%). Hearing curve configurations were classified into five types: ascending, descending, concave/convex, flat, and profound. We conducted a meta-analysis reporting case numbers for each type. Due to insufficient data, the concave/convex type was excluded from analysis, as only two studies described this configuration. For the remaining four curve types, data from four studies were included in this analysis. The pooled prevalence of comorbidity among patients with ascending curves was 16.8% (95% CI: 5.6%–28.0%; I² = 82.43%), while that for descending curves was 16.0% (95% CI: 8.1%–23.9%; I² = 62.92%). Flat hearing curves were observed in 26.6% of patients (95% CI: 15.8%–37.4%; I² = 72.59%), while complete deafness curves were present in 28.5% (95% CI: 19.0%–38.0%; I² = 61.29%) (Table 3).

Symptoms

Eight studies reported the prevalence of tinnitus, with a pooled prevalence of 56.7% (95% CI: 40.7%–72.7%; I² = 93.16%). Similarly, eight studies documented vertigo, yielding a pooled prevalence of 21.8% (95% CI: 13.1%–30.6%; I² = 81.94%). Four studies reported the prevalence of ear fullness, with a pooled prevalence of 31.0% (95% CI: 11.8%–50.2%; I² = 88.83%) (Table 3).

Treatment

When classifying the time to initial diagnosis using categorical variables, studies varied in their categorization of time intervals and measurement units. Only the ≤7-day interval had sufficient data (n = 3) for meta-analysis. The proportion of patients starting treatment within ≤7 days after symptom onset was 78.9% (95% CI: 73.6%–84.2