Comparison of different oval window sealing materials in stapes surgery: systematic review and meta-analysis

This study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Statement [11], and the study protocol was registered on PROSPERO (reference number CRD42022304958) without subsequent modifications to the drafting available at https://​www.​crd.​york.​ac.​uk/​prospero/​display_​record.​php?​RecordID=​304958

We carried out a literature search on Pubmed/MEDLINE, Cochrane, EMBASE, and OVID databases for articles published from December 2001 to 31st December 2021, in English. We also performed a snowball search to identify additional studies by searching the reference lists of publications eligible uploaded to EndNote (Clarivate Analytics, Philadelphia, PA). The search string was the following: “Stapes Surgery” [Mesh] “Otosclerosis”[Mesh] “Progressive hearing loss stapes fixation” [Supplementary Concept] “stapedectomy” [tw] “stapedotomy” [tw] “Oval Window, Ear”[Mesh], “Tissue Transplantation”[Mesh] “Connective Tissue” [Mesh] “Gelatin Sponge, Absorbable”[Mesh], “Fascia”[Mesh], “Veins”[Mesh] “Hyaluronic Acid”[Mesh] “Seal”[tw] “Sealing”[tw] “Graft”[tw] “Vein”[tw] “Fat”[tw]. Two authors (A.S., P.M.) independently reviewed the titles and abstracts obtained from all the databases. Then, the full texts were assessed. All studies that did not meet the inclusion criteria and duplicates were excluded. Cross-references were verified, and valuable articles were included. In the case of disagreement about the eligibility of a study, a third author (M.R.) decided which articles were included. Data were extracted independently by two authors (A.S., P.M.) and confirmed where necessary by the principal investigator (A.S.). A data extraction spreadsheet was developed and the information from the studies included in the review was extracted and tabulated using an Excel sheet, double-checked for accuracy. Another reviewer (M.R.) checked the data collected.

The study selection was based on the Population, Intervention, Comparison, Outcome, and Study (PICOS) framework [12] (Table 1). The air–bone gap (ABG) closure (post-operative air conduction minus postoperative bone conduction) was defined according to the American Academy of Otolaryngology-Head and Neck Surgery (AAO-HNS) [13]. The calculation was made by using the four-tone average (0.5–1.3, 3, or 4 kHz). Surgical success is usually defined as the post-operative closure of the ABG to ≤ 10 dB, without deterioration of bone conduction (BC) at all frequencies [14].

Studies that considered stapedectomy, ossiculoplasty, or cases of surgical revision were excluded. We excluded reviews, case reports, case series, letters, comments, and editorials. Articles that did not satisfy the primary outcome were excluded. Secondary outcomes included SNHL and postoperative dizziness. All the outcomes were assessed at the longest follow-up time.

Two review authors (AS and PM) independently extracted data and assessed the risk of bias for selected studies including the population demographics, study design, type of surgical procedure, type of prosthesis used, outcome, and follow-up. Discrepancies were identified and resolved through discussion.

The meta-analysis was carried out using the STATA 16 v statistical software, and a p value ≤ 0.05 was considered to be statistically significant. Based on the treatment, we performed the first analysis stratifying the population by the use of fat compared to all other treatments. The second analysis was performed considering, as treatment, the use of the gelfoam. Data were expressed as mean ± standard deviation with a 95% confidence interval. The summary statistics required for each outcome were the number of participants in the treatment and control groups at post-test, and the mean and SD of the ABG closure outcome after the intervention. Random-effects models were prespecified a priori, given the heterogeneity across settings, participants, and sample size [15].

Statistical heterogeneity was calculated using the Higgins I² statistic, which describes the percentage of variability in the effect estimate due to heterogeneity rather than sampling error. Inconsistency was examined using I² and the following grades were applied: < 25% (very low), 25 to < 50% (low), 50 to < 75% (moderate), and ≥ 75% (large) [16, 17]. The possibility of small study effects was assessed qualitatively by a visual estimate of the Funnel plot and quantitatively by calculation of the Egger and Begg’s tests [18, 19].

Besides, we performed subgroup analysis to test interactions according to the study design (retrospective, RCT). A Meta-regression was also performed to check the possible influence of confounders (e.g., age and gender) on the results.

The database search yielded 244 results. Figure 1 shows the PRISMA algorithm. Six studies were included in the review [20‐25]. Characteristics of the studies are summarized in Table 2.

Two studies were from Iran and one each from the United States, France, Saudi Arabia, and India. Of the six studies, two were randomized control trials and four were retrospective cohort studies. A total of 1298 patients were evaluated, with a mean age of 39.9 ± 6.27 years, and a female to male ratio of 1.5:1.

The methodological quality of included studies was checked independently by two authors (AS and PM) using the Downs and Black checklist [26]. It has been ranked in the top six quality assessment scales suitable for use in systematic reviews [27]. The checklist consists of 27 items to assess the quality of reporting, external validity, internal validity (bias and confounding) and power. As in other systematic reviews [28, 29], the tool was modified slightly for use in this review simplifying the scoring for question 27 dealing (statistical power) to assign either 1 point or 0 rather than a score from 0 to 5. A score of 1 was assigned to studies that included a power calculation, while a score of 0 was given to studies without any power calculation.

Downs and Black score ranges were grouped into the following 4 quality levels: excellent [26‐28], good [20‐25], fair [15‐19], and poor [14].

Any disagreements were discussed until a consensus was reached or resolved by a third author (MR). Two studies were evaluated good, two were evaluated fair and one was evaluated poor (Table 3).

In the included studies, there were 327 patients in the gelfoam group, 314 in the tragal perichondrium group, 203 in the no graft group, 184 in the lobule fat group, 138 in the vein graft group, 100 in the hyaluronic acid group, and 32 in the adipose graft tissue around the piston group. A difference in the ABG calculation method was observed: Angeli’s [20], Pradhan's [24], and Schmerber’s [25] studies used the means of the thresholds for bone and air conduction at 0.5, 1, 2, and 4 kHz, while Faramarzi [22], Bawazeer [21], and Faramarzi [23] adopted 0.5, 1, 2, and 3 kHz. The latter was calculated as an average of 2 and 4 kHz frequencies. Because of the heterogeneity of the studies included, we decided to focus our investigation on fat versus other materials and gelfoam versus other materials.

In our study, we identified as primary outcome the ABG after the intervention. Because of the heterogeneity of the treatments adopted, we checked two main methods: the use of fat compared to all other treatments, and the use of the gelfoam compared to all other treatments.

The first analysis found that four studies [20, 22‐24] were eligible. The total population included 216 subjects who underwent a surgical procedure using fat, compared to 212 subjects who underwent a surgical procedure not using fat. The meta-analysis did not identify differences between the groups (p = 0.74), with a low heterogeneity of the results (I² = 28.36%; Hedge’s g = 0.04, 95% CI − 0.19 0.27; Fig. 2A). The Funnel plot and the Egger and Begg’s test did not show small study effects publication bias (Fig. 2B). Then, to check the possible interaction of the study design with these results, we performed a sub-analysis. Both in retrospective and RTC studies, no statistically significant differences were found in ABG closure between the two groups, whereas in the retrospective studies a very low heterogeneity was found (I² = 0.0%, p = 0.38), while the RTCs had a moderate heterogeneity (I² = 66.04%, p = 0.81). Noteworthily both the subgroups included only two studies (Fig. 3A, B).

Then, investigating the possible effect on post-operative ABG of the use or not of gelfoam, we found only three studies eligible [21, 22, 24]. The total population included 327 subjects who underwent a surgical procedure using gelfoam, compared to 310 subjects who underwent a surgical procedure not using gelfoam.

The meta-analysis did not identify differences between the groups (p = 0.97), with a low heterogeneity of the results (I² = 28.91%; Hedge’s g = 0.00, 95% CI − 0.20 to 0.21; Fig. 4A). The Funnel plot and the Egger and Begg’s test did not show small study effects publication bias (Fig. 4B). Then, to check the possible interaction of the study design with these results, we performed a sub-analysis. In the retrospective studies, no statistically significant differences (p = 0.38) were found in ABG between the two groups, whereas a very low heterogeneity was found (I² = 0.0%). The RTCs included only one study, making analysis impossible (Fig. 5A, B).

Since the absence of differences in ABG levels between patients treated with or without fat, we ran a meta-regression analysis to seek potential moderators, such as age and percentage of the female population. Both these parameters did not affect the findings (age: beta = 0.0104; 95% CI − 0.0378 to 0.0586, p = 0.673; percentage of female: beta =  − 0.0086; 95%CI − 0.0309 to 0.0136, p = 0.447).

Similarly, a meta-regression was performed, to check the possible influence of the age and the percentage of the female population as moderators on the ABG levels in patients treated with or without gelfoam. Again, both these factors did not affect the findings (age: beta = 0.0177; 95% CI − 0.0757 to 0.0404, p = 0.550; percentage of female: beta =  − 0.0027; 95% CI − 0.0395 to 0.0340, p = 0.884).

This systematic review was conducted using transparent methods and a priori defined criteria in accordance with the guidelines of the Cochrane Collaboration and PRISMA. The review protocol was registered in the International Prospective Register of systematic reviews (PROSPERO). To our knowledge, this represents the first metanalysis exploring the effect of the fat or the gelfoam with respect to the other treatments on the ABG closure.

Limitations included a restricted search in language and a few cases due to the unavailability of quality studies. Only two papers were randomized controlled trials while the others were retrospective comparative studies. Another limitation is the relatively short duration of follow-up. The authors of the included studies were not contacted for further information and thus the results are solely based on the published data. Surely the main limitation is the small number of studies introduced that, however, represent the only available in the literature up to date.