Outcomes of Tympanoplasty After Cleft Palate Repair: A Systematic Review and Meta-analysis

Abstract

Objective.

To analyze graft success rates and hearing outcomes in patients with a history of cleft palate (CP) repair undergoing tympanoplasty.

Data Sources.

PubMed, Scopus, and CINAHL.

Review Methods.

Per PRISMA guidelines, the databases were searched from date of inception through December 14, 2021. Studies of patients with previous CP repair who underwent tympanoplasty were included. Meta-analysis of proportions, continuous measures, odds ratios (ORs), and meta-regression were used to analyze graft success and hearing outcomes after tympanoplasty.

Results.

A total of 323 patients with CP repair and 1169 controls were included. The proportion of graft success was 86.7% (95% CI, 76.1%-94.5%) in patients with CP repair and 88.8% (95% CI, 76.9–96.8) in controls. There was no difference in odds of graft success between patients with CP repair and controls (OR, 1.0 [95% CI, 0.5–1.8]; P = .870). Age was not a significant moderator of graft success in patients with CP repair (r = 0.1 [95% CI, −0.2 to 0.3]; P = .689) or controls (r = −0.0 [95% CI, −0.1 to 0.1]; P = .952). Comparing mean differences between pre- and postoperative air-bone gap was not statistically significant in patients with CP repair and controls (0.2 dB [95% CI, −3.1 to 3.4]; P = .930). Odds of functional success (postoperative air-bone gap < 20 dB) were not different between the groups (OR, 0.8 [95% CI, 0.5–1.4]; P = .450).

Conclusion.

This meta-analysis does not endorse anatomic or functional differences between patients with CP repair and controls after tympanoplasty. However, there is a paucity of evidence for younger children. Further studies are warranted to elucidate specific risk factors for tympanoplasty outcomes in young patients with previous CP repair.

Patients with cleft palate have been described as being at higher risk for eustachian tube dysfunction (ETD).^1,2 Multiple etiologies have been proposed: abnormal tensor veli palatini insertion, increased reflux into the eustachian tube, and lower volume of lateral lamina in the eustachian tube cartilage.^2–6 ETD increases the prevalence of middle ear disease, and upward of 98% of patients with cleft palate require tympanostomy tubes.^7–10 However, the effects of cleft palate repair on eustachian tube function are still debated. While 1 study did not endorse any change in tympanostomy tube requirements,¹¹ 2 other studies showed a recovery of eustachian tube function, albeit after 6 to 7 years of age.^12,13 This topic remains controversial, as some authors recommended against middle ear surgery in patients with cleft palate and orofacial cleft secondary to their ETD.^14–16

Tympanoplasty is a frequently performed procedure in the setting of tympanic membrane perforation. Posttympanostomy tube perforation ranges from 2.2% to 4.0% after spontaneous extrusion and 10% to 13% after surgical removal.^17–21 This procedure has a wide range of success rates, from 35% to 94% in children and 60% to 99% in adults.^22–24 Proper eustachian tube function is an important prognostic factor for a positive outcome after tympanoplasty; however, recent reports have been unable to find an association between ETD and graft failures.^25–27 Furthermore, recent studies have not shown an increased risk of poor tympanoplasty outcomes in patients after cleft palate repair.^28,29 It is currently unknown if patients with cleft palate and a prior palatoplasty have less favorable outcomes after tympanoplasty as compared with the general population.

There is a lack of consensus to guide clinicians in the timing and preoperative counseling for tympanoplasty among patients with previous palatoplasty. Therefore, we performed a systematic review and meta-analysis with the primary aim of analyzing graft success and hearing outcomes in patients with a history of palatoplasty for cleft palate as compared with controls. Secondarily, we wanted to determine if any demographic factors, such as age, influenced the outcomes of interest. We hypothesized that the meta-analysis would demonstrate that patients with a prior cleft palate repair would have lower success rates and reduced hearing improvements after tympanoplasty than controls. Furthermore, as eustachian tube function improves with age, we hypothesized that age at tympanoplasty may modulate graft success rates in this population.

Methods

Search Criteria

This study adhered to the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-analyses).³⁰ Detailed search strategies were developed for PubMed (US National Library of Medicine, National Institutes of Health), Scopus (Elsevier), and CINAHL (EBSCO). A combination of subject headings (eg, MeSH in PubMed) and keywords around the concepts “tympanoplasty,” “cleft palate,” and “pediatrics” was used to develop a search strategy. The PubMed search strategy was formatted for the other 2 data-bases by maintaining similar keywords and replacing MeSH terms with appropriate subject headings when available. Data-bases were searched from the date of inception through December 14, 2021. Detailed search strategies, filters applied, and the number of results for each database are listed in Supplemental 1 (available online). Furthermore, the reference lists of relevant articles were manually searched to confirm the search strategy to ensure that no relevant articles were missed. The review management software Covidence (Veritas Health Innovation Ltd) was used for study selection.

Selection Criteria

The inclusion criteria consisted of studies that analyzed the outcomes of tympanoplasty in patients with a prior cleft palate repair. Studies were not required to have controls as comparisons. Studies comparing myringoplasty, tympanoplasty, and tympanomastoidectomy were included. Cholesteatoma was not an exclusion criterion. Double- or single-blinded randomized controlled trials, double- or single-blinded randomized comparison trials, nonrandomized controlled trials, and prospective or retrospective observational studies were considered for inclusion. The remaining exclusion criteria were non-English language, nonhuman studies, review articles, case reports of <4 patients, duplicates, and inaccessible articles.

Titles and abstracts were first independently assessed by 2 reviewers (N.S.P. and H.S.) to identify all articles that met the inclusion criteria. Conflicts were resolved by discussion with a third author (P.P.P.). Two reviewers (N.S.P. and H.S.) then independently assessed the full texts of articles to identify those that met all inclusion and exclusion criteria for the final analysis; any disagreements were resolved by a third author (P.P.P.). The Oxford Center for Evidence-Based Medicine criteria were used to critically assess the evidence level of articles.³¹ Risk of bias was assessed according to the Cochrane Handbook for Systematic Reviews of Interventions (version 6.0).³² Since all studies were retrospective reports, the ROBINS-I tool was used (Risk of Bias in Non-randomized Studies of Interventions).³³ The items assessed for risk of bias were as follows: bias due to confounding, bias in the selection of participants into the study, bias in classification of interventions, bias due to deviations from intended interventions, bias due to missing data, bias in the measurement of outcomes, and bias in the selection of the results. The risk of bias for each aspect was graded as low, unclear, or high. Two authors (N.S.P. and H.S.) independently performed a pilot assessment on 3 studies and compared results to check for consistency of assessment. Both authors then performed an independent risk assessment on the remaining studies, with all disagreements being resolved by a third party (P.P.P.).

Data Extraction

Two reviewers (N.S.P. and H.S.) independently extracted the data and compared them for accuracy. Extracted data included author, year of publication, and patient demographics (age and sex). The type of orofacial cleft, tympanoplasty procedure, presence of cholesteatoma, definition of procedure success, and length of follow-up were additionally extracted. The number of graft successes vs failures was recorded. Pre- and postoperative air-bone gaps (ABGs) were collected. Functional success, described as a postoperative ABG <20 dB, was also extracted.^34,35

Statistical Analysis

Categorical data were presented by frequencies with percentages and continuous data by mean ± SD or range (minimum, maximum). Meta-analysis of proportions was performed with MedCalc 20.011 (MedCalc Software). The pooled measures for this analysis consisted of graft success and no recurrence of cholesteatoma. These measures were weighted according to the number of patients in the outcome. The Freeman-Tukey transformation was used to calculate the weighted summary proportion from each study in the meta-analysis.³⁶ If these analyses suggested high heterogeneity (I² > 50% and P < .05), then the fixed effects model might be invalid, and the random effects model would be more appropriate.^37,38 A random effects model incorporates the random variation within and between studies.^37,38

Meta-analysis of continuous measures compared the following: ABG for preoperative cleft palate repair vs preoperative control, postoperative cleft palate repair vs postoperative control, and pooled cleft palate repair difference vs pooled control difference. Odds ratios (ORs) compared graft success between the cleft palate repair and control groups. Both analyses were performed with Cochrane Review Manager version 5.4 (Nordic Cochrane Centre, Cochrane Collaboration). Fixed and random effects models depended on the I² or heterogeneity test results.^37,38

Comprehensive Meta-analysis (version 3; Biostat) was used to perform a meta-analysis regression to determine the associations between graft success and age of tympanoplasty. Meta-analysis regression relates statistical heterogeneity between study effect sizes and modulator variables via regression-based techniques.^39,40 This analysis calculated r (95% CI), R², and residual I². R is the selected covariates’ correlation coefficient; R² describes the between-study variance explained by the variables; and I² is a proportion used to assess if the between-study variation is due to heterogeneity or sampling variation. Method of moments and z distribution were used for the analysis, and the regression was graphed by logit of graft success rate and age.

Additionally, Egger tests with funnel plots were performed to assess the risk of publication bias.^41,42 Potential publication bias was evaluated by an Egger regression test, which statistically examines the asymmetry of the funnel plot.⁴¹ Visual inspection of the funnel plot was also used, as publication bias results in an asymmetrical funnel plot. A P value <.05 was considered to indicate a statistically significant difference for all statistical tests.

Results

Search Results and Study Characteristics

The literature search yielded 431 unique articles, with title and abstract screening excluding 400 articles. A full-text review of the remaining studies excluded 19 articles, leaving 12 remaining articles for inclusion in the final data extraction and analysis. The PRISMA flowchart outlining the entire search process is shown in Figure 1. Critical appraisal of studies indicated an acceptably low risk of bias for most studies (Figure 2). Bias due to missing data and bias in the selection of the results were the most common biases. A funnel plot with Egger test suggested little publication bias as all studies were within the funnel with no asymmetry (−6.9 [95% CI, −17.0 to 3.3]; P = .151; Supplemental 2, available online). Articles selected for inclusion were level 3b and 4 studies based on the Oxford level of evidence and were published between 1992 and 2020 (Table 1).

Figure 1.

PRISMA flowchart of study selection.

Figure 2.

Risk of bias in selected studies.

Table 1.

Description of Included Studies With Patients With Cleft Palate Repair.

Author (year)	OLE	Design	Total, No.	Age, y, mean, (range)	Male, %	Cleft lip and palate, No.	Cleft palate, No.	T, No.	T&M, No.	Cholesteatoma, %	Description of procedure success	Follow-up, mo, mean ± SD
Gardner (2002)29	3b	CC	20	24.0 (2.0–55.0)	NR	16	4	17	9	19.2	No graft failure	25.0 ± NR
Harterink (2014)43	3b	CC	32	10.5 (4.9–15.3)	25.0	9	23	32	0	0.0	No perforation, revision surgery	51.2 ± 36.6
Imbery (2017)44	4	CS	24	NR (NR)	NR	NR	NR	24	0	0.0	NR	NR
Juergens (2019)45	4	CS	16	11.5 (3.9–22.2)	50.0	10	6	16	4	5.0	Intact tympanic membrane	40.1 ± 38.4
Knapik (2012)46	3b	CC	15	12.3 (NR)	60.0	NR	NR	15	0	0.0	Intact tympanic membrane	47.5 ± 34.8
Kopcsányi (2014)50	3b	CC	21	10.7 (4.7–16.0)	40.7	NR	NR	NE	NE	0.0	NR	NR
Kopcsányi (2015)51	3b	CC	19	9.5 (5.5–13.7)	60.0	NR	NR	NE	NE	100.0	No recurrent cholesteatoma	NR
Lourenxçone (2018)47	4	CS	8	25.8 (14.0–34.0)	12.5	8	0	0	13	100.0	No recurrent cholesteatoma	NR
Lourenxçone (2020)48	4	CS	97	27.0 (10.0–57.0)	49.5	69	28	0	118	100.0	No recurrent cholesteatoma	160.8 ± 70.6
Metrailer (2014)52	4	CS	37	NR (4.0–19.0)	NR	20	17	45	0	NR	No revision surgery	NR
Vartiainin (1992)28	4	CS	16	16.5 (2.0–41.0)	NR	NR	NR	NR	NR	36.8	No revision surgery	78.0 ± NR
Vincenti (2014)49	4	CS	18	10.3 (6.0–16.0)	61.1	8	10	0	16	100.0	No recurrent cholesteatoma	96.0 ± NR

Abbreviations: CC, case-control; CS, case series; NE, not extractable; NR, not reported; OLE, Oxford level of evidence; T, tympanoplasty; T&M, tympanoplasty and mastoidectomy.

Patient Characteristics

A total of 1492 patients were included from all 12 studies (Tables 1 and 2). The group with cleft palate repair had 323 patients, and the control group had 1169 patients. The weighted mean age was 18.5 years (range, 2.0–57.0) in the group with cleft palate repair and 26.4 years (range, 2.4–37.8 years) in the control group. In studies reporting sex, males composed 47.8% of patients with cleft palate repair and 59.4% of patients without cleft palate repair. Patients with both cleft lip and cleft palate repair made up 61.4% of the group with cleft palate repair. Most studies analyzed outcomes after myringoplasty or tympanoplasty,^29,43–46 but some studies included tympanomastoidectomy.^47–49 Four studies excluded patients with cholesteatoma^43,44,46,50, 4 studies included only patients with cholesteatoma^47–49,51, and 3 studies included a range of patients with cholesteatoma (5.0%-36.8%).^28,29,45 Different descriptions of procedure success were extracted, with the most common being no revision surgery and intact tympanic membrane. The 4 studies of patients with cholesteatoma considered recurrent cholesteatoma as their procedure success.^47–49,51 Mean follow-up ranged from 25.0 to 160.8 months for patients with cleft palate repair and 31.3 to 75.0 months for controls.

Table 2.

Description of Included Studies With Controls.

			Control
Author (year)	OLE	Design	No.	Age, y, mean (range)	Male, %	T, No.	T&M, No.	Cholesteatoma, %	Description of procedure success	Follow-up, mo, mean ± SD
Gardner (2002)29	3b	CC	52	NR (NR)	NR	NR	NR	NR	No graft failure	NR
Harterink (2014)43	3b	CC	32	10.5 (5.5–16.7)	65.6	32	0	0.0	No perforation, revision surgery	75.0 ± 42.6
Knapik (2012)46	3b	CC	144	11.0 (NA)	52.1	144	0	0.0	Intact tympanic membrane	NR
Kopcsányi (2014)50	3b	CC	119	10.8 (3.3–21.3)	44.4	NE	NE	0.0	NR	31.3 ± 24.3
Kopcsányi (2015)51	3b	CC	151	10.7 (2.4–19.5)	61.1	NE	NE	100.0	No recurrent cholesteatoma	NR
Vartiainin (1992)28	4	CS	671	37.8 (NR)	NR	NR	NR	33.4	No revision surgery	NR

Abbreviations: CC, case-control; CS, case series; NA, not applicable; NE, not extractable; NR, not reported; OLE, Oxford level of evidence; T, tympanoplasty; T&M, tympanoplasty and mastoidectomy.

Graft Success

Eight studies had extractable outcome data on graft success for patients with a cleft palate repair,^{28,29,43–46,50,52} and 5 studies had similar data for controls.^{28,29,43,46,50} Graft success was defined by each study’s criteria at its particular follow-up timepoint. The graft success rate was 86.7% (95% CI, 76.1%-94.5%) for the group with cleft palate repair and 88.8% (95% CI, 76.9–96.8%) for the control group (Figures 3 and 4). Odds of graft success did not favor the group with cleft palate repair or controls (OR, 1.0 [95% CI, 0.5–1.8]; P = .870). The forest plot is represented in Figure 5. For patients with a prior cleft palate repair, age was not a significant moderator of graft success (r = 0.1 [95% CI, −0.2 to 0.3]; P = .689). Age also did not significantly influence the variance of graft success in controls (r = −0.0 [95% CI, −0.1 to 0.1]; P = .952).

Figure 3.

Forest plot of proportion of graft success in patients with cleft palate repair.

Figure 4.

Forest plot of proportion of graft success in control patients.

Figure 5.

Forest plot of comparison of graft success in patients: cleft palate repair vs control.

A subanalysis of children (age <18 years) was performed, with Harterink et al and Kopcsányi et al reporting data on graft success for exclusively pediatric patients with a cleft palate repair and with Harterink et al reporting on exclusively pediatric controls.^43,50 Children with cleft palate repair had a success rate of 89.6% (95% CI, 75.8%-98.0%; Figure 6), and Harterink et al cited a success rate of 84.4% for their controls.⁴³

Figure 6.

Forest plot of proportion of graft success in exclusively children with cleft palate repair.

Hearing Outcomes

Five studies reported preoperative ABG,^{28,43,46,50,51} and 6 studies reported postoperative ABG for patients with cleft palate repair and controls.^{28,29,43,46,50,51} There was not a statistically significant mean difference between these groups for preoperative ABG (1.3 dB [95% CI, −3.3 to 5.9]; P = .570) and postoperative ABG (0.4 dB [95% CI, −2.3 to 3.2]; P = .770; Figure 7). When overall change between pre- and postoperative ABG was compared, there was no difference in patients with cleft palate repair and controls (0.2 dB [95% CI, −3.1 to 3.4]; P = .930; Figure 8). Regarding functional success, patients with cleft palate repair did not have lower odds of success than controls (OR, 0.8 [95% CI, 0.5–1.4]; P = .450). The forest plot is shown in Figure 9.

Figure 7.

Forest plot of comparison of postoperative air-bone gap in patients: cleft palate repair vs control.

Figure 8.

Forest plot of comparison of mean differences in air-bone gap in patients: cleft palate repair vs control.

Figure 9.

Forest plot of comparison of functional success in patients: cleft palate repair vs control.

Cholesteatoma Recurrence

Only 4 studies analyzed patients with cleft palate repair and cholesteatoma,^47–49,51 and just 1 study reported on controls.⁵¹ For patients with a prior cleft palate repair, the proportion of cholesteatoma recurrence was 30.5% (95% CI, 8.7%-58.4%; Figure 10). In the 1 study analyzing controls, 2.9% of patients had a recurrence of cholesteatoma.

Figure 10.

Forest plot of proportion of cholesteatoma recurrence in control patients.

Discussion

Patients with a cleft palate have a higher prevalence of ETD and middle ear disease.^1,2,7–9 The effects of cleft palate repair on improving these pathologies are debated.^11–13 Often, tympanoplasty success rates are higher with healthy eustachian tube function; for this reason, it has been postulated that patients with cleft palate may differ in their outcomes after tympanoplasty.^14–16 However, recent evidence has questioned the impact of ETD on tympanoplasty outcomes.^25–29 Therefore, this systematic review and meta-analysis aimed to evaluate graft and hearing outcomes after tympanoplasty in patients with a history of cleft palate. We did not find any evidence of improved graft success or hearing outcomes in controls after tympanoplasty as compared with patients with a former cleft palate repair. Moreover, studies in the current literature did not identify age as a significant factor of success in this group, although the patient population consisted of older children and adults rather than younger children.

Graft success was similar between patients with a prior cleft palate repair and controls (86.7% vs 88.8%), and odds of graft success were comparable between these groups. This rate is on the higher end of the range of pediatric tympanoplasty success noted in the literature (35% to 94%),^22–24 but is closely in line with a recent systemic review of successful type I tympanoplasties (includes underlay, inlay, and onlay techniques as well as fat graft myringoplasty) in children at 83.4%.⁵³ One of the concerns of poor tympanoplasty outcomes for these patients is whether having the cleft palate repaired improves ETD. Interestingly, the association between eustachian tube function and tympanoplasty success has been questioned. One retrospective study was unable to find ETD as a predictive factor for failure.²⁷ An extensive review was unable to find eustachian tube testing preoperatively to be a significant prognostic value for tympanoplasty success.⁵⁴ The most recent systematic review on pediatric type I tympanoplasties also did not endorse eustachian tube function as a factor in graft success.⁵³ Furthermore, it is unclear whether palatoplasties improve eustachian tube function. One study did endorse a slight improvement in eustachian tube function after straight-line closure or Z-plasty by using tympanostomy tubes and antibiotics as the outcome measure; however, 27.8% and 20% of patients still required tympanostomy tubes postoperatively.⁵⁵ Two other studies compared eustachian tube function pre- and postpalatoplasty by using the forced response test.^56,57 One group found an increase in appropriate eustachian tube dilation after swallowing from 33% to 83% in infants, and the other group found an appropriate eustachian tube dilation in 60% of children post-palatoplasty.^56,57 A review on this topic noted that while the tensor veli palatini plays a vital role in eustachian tube opening in healthy patients, its effects on patients with cleft palate are uncertain.⁵⁸ Moreover, it remains unclear whether the type of surgical technique used to re-create the velar sling can impact eustachian tube function.⁵⁸ This meta-analysis suggests that patients with cleft palate may have a similar risk profile of failure as patients without a cleft palate, although there is still ambiguity surrounding when ETD resolves in patients with cleft palate. Thus, close long-term follow-up is required to monitor and reduce significant complications, such as cholesteatoma.⁵⁹

While most studies in this systematic review focused on tympanoplasty results in children, the weighted means of the 2 groups were 18.5 years for those with cleft palate repair and 26.4 years for the controls. Some results endorse a higher success rate of tympanoplasty in adults than children, but most studies show nonsignificant differences.^60–64 Furthermore, we did not find age to be a modulator of graft success between patients with cleft palate repair and controls. While an early systematic review did endorse a higher success rate for older healthy children,⁶⁵ an updated systematic review did corroborate these findings.⁵³ A recent review also argued that other patient variables, rather than just age, may be more predictive of tympanoplasty outcomes; these variables included site and size of the perforation, technique, otorrhea, eustachian tube function, and status of the contralateral ear.⁶⁶ An important distinction is that our meta-regression could incorporate only mean ages into the regression analysis. The lowest mean age was 10.3 ± 2.5 years, which is 3 to 4 years after the suggested return of eustachian tube function after cleft palate repair according to 2 studies.^12,13 Moreover, a recent study found that over half of eustachian tube maturation occurred by 8 years of age and reached adult morphology by early adolescence. Therefore, our meta-regression analyzing the modulating effect of age was ineffective at examining younger age groups. While we did not find different graft take results between children with cleft palate and controls in studies exclusively focusing on children <18 years old (89.6% vs 84.4%), this analysis was limited to just 2 studies. Further studies are needed to address the optimal timing of tympanoplasty for young children with tympanic membrane perforations.

Pre- and postoperative hearing outcomes did not differ between patients with prior palatoplasty and controls. Pediatric tympanoplasty has generally been shown to result in excellent hearing outcomes.^27,67–69 It is well known that patients with cleft palate usually have conductive hearing loss secondary to otitis media with effusion that can frequently be relieved with tympanostomy tubes.^70–72 However, there is some evidence that hearing improvements from tympanostomy tubes are weaker in the higher frequencies and in patients >4 years old.^73,74 Patients with cleft palate generally have satisfactory hearing outcomes with early tympanostomy tube placement.⁷⁵ Furthermore, early cleft palate repair is an important factor in hearing and speech outcomes in patients with cleft palate.^76,77 There is competing evidence on long-term hearing outcomes after cleft palate repair, with 1 study finding 19.2% of patients with residual hearing loss in the higher frequencies and another finding resolution of hearing loss by the age of 6 years.^78,79 Therefore, our findings of similar hearing outcomes between the groups do not seem to contradict the current evidence. Overall, this study supports similar functional outcomes after tympanoplasty between patients with cleft palate repair and controls.

A small subgroup of studies reported tympanomastoidectomy outcomes regarding cholesteatoma recurrence. Cholesteatoma recurrence after tympanomastoidectomy was highly variable in the population with cleft palate repair but was higher than the 1 study reporting cholesteatoma recurrence in control patients. The proportion of cholesteatoma recurrence in the population with cleft palate repair was 30.5%, but the confidence interval was wide (8.7% to 58.4%). In addition, just 1 study analyzed controls, making conclusions difficult to draw from this analysis. Pediatric cholesteatoma typically has been shown to have a recurrence rate of 33%,⁸⁰ but this can be as low as 2.6% to 7.6% if a canal wall down mastoidectomy is performed.⁸¹ Unfortunately, due to the methodology of studies, we could not separate outcomes based on canal wall up vs canal wall down mastoidectomies. However, there is significant evidence that patients with cleft palate are at a much higher risk of acquired cholesteatoma than patients without cleft palate.^82–84 We are unable to comment on this with our data, as many of the studies explicitly included or excluded cholesteatoma in patients with cleft palate and controls, therefore skewing our findings. Furthermore, while there is limited evidence concerning the effects of palatoplasty on cholesteatoma occurrence, 1 study found that supplementing palatoplasty with an inferior or superior pharyngeal flap did not decrease the risk of cholesteatoma.⁸⁵ Another study noted that adults with previously repaired submucous cleft palates were more likely to have cholesteatoma.⁸⁶ Overall, while patients with cleft palate have an elevated risk of cholesteatoma, this study could not comment on their recurrence rates after tympanomastoidectomy as compared with control patients.

A major limitation of this analysis is the lack of standardized reporting leading to heterogeneity among the studies. First, inconsistent reporting of variables may have affected how graft and hearing outcomes were interpreted. Variables such as perforation location and size, opposite ear status, preoperative imaging findings, ossicular chain status, type of graft, type of cleft palate, palatoplasty technique, and the number of previous tympanostomy tubes were unable to be collected due to being unextractable or not reported. Because of this, meaningful subgroup analyses were unable to be performed to analyze other risk factors that may influence poor tympanoplasty outcomes. In addition, successful surgical outcomes were defined differently in each study; therefore, one study’s failure (eg, repeat tympanostomy tube placement, conductive hearing loss) may still be defined as a successful outcome in another study if the primary outcome was an intact tympanic membrane. What is more, these analyses may be subject to selection bias given the variety in techniques used at the surgeon’s discretion, such as cartilage graft and tympanostomy tube placement at the time of tympanoplasty. For example, some studies considered the absence of revision surgery to be a success, whereas others required no recurrent perforation. Notably, Metrailer et al and Gardner and Dornhoffer regarded postoperative tympanostomy tubes a success, whereas Juergens et al defined this outcome as a failure of the tympanoplasty.^29,45,52 These differences question graft success comparisons between studies and how the results would have changed if a uniform definition had been adopted. Follow-up was also variable among the studies, ranging from 25.0 to 160.8 months. As the mean time from placement of tympanostomy tube to healed perforation has been shown in the population with cleft palate to be 43 months, lack of long-term follow-up in some studies could have misrepresented the graft success rate.⁴⁴ Moreover, if the age at tympanoplasty for children with a cleft palate repair was delayed due to concerns from the literature, this might have contributed to our limited analyses on exclusively pediatric patients. Furthermore, the fact that all studies were retrospective limits how this analysis translates to the general population. Therefore, a prospective observational trial using standardized definitions of graft success, particularly among younger children with cleft palate, would better elucidate tympanoplasty outcomes in this population.

Conclusion

When compared with controls, patients with a prior cleft palate did not have differing functional or anatomic outcomes after tympanoplasty. Age was not a modulator of graft success in both groups; however, this review identified a paucity of studies among younger children undergoing tympanoplasty. These findings suggest that older patients with cleft palate having undergone prior palatoplasty have a similar risk profile for tympanoplasty as patients without a cleft palate. Nevertheless, there was significant heterogeneity among studies. Future studies should aim to identify risk profiles for tympanoplasty outcomes, particularly in terms of an optimal age for tympanoplasty in children with a history of cleft palate.