The Rate of Oronasal Fistula Following Primary Cleft Palate Surgery: A Meta-Analysis

Transcription

1 The Cleft Palate Craniofacial Journal 52(4) pp. e81 e87 July 2015 Ó Copyright 2015 American Cleft Palate Craniofacial Association ORIGINAL ARTICLE The Rate of Oronasal Fistula Following Primary Cleft Palate Surgery: A Meta-Analysis Michael R. Bykowski, M.D., M.S., Sanjay Naran, M.D., Daniel G. Winger, M.S., Joseph E. Losee, M.D. Background: Despite decades of craniofacial surgeons repairing cleft palates, there is no consensus for the rate of fistula formation following surgery. The authors present a metaanalysis of studies that reported on primary cleft palate to determine the rate of oronasal fistula and to identify risk factors for their development. Methods: A literature search for the period between 2000 and 2012 was performed. Articles were queried and strict inclusion and exclusion criteria were applied to focus on primary cleft palate repair. A meta-analysis of these data was conducted. Results: The meta-analysis included 11 studies, comprising 2505 children. The rate of oronasal fistula development was 4.9% (95% confidence interval, 3.8% to 6.1%). When analyzing a larger cohort, there was a significant relationship between Veau classification and the occurrence of a fistula (P,.001), with fistulae most prevalent in patients with a Veau IV cleft. The most common location for a fistula was at the soft palate-hard palate junction. One study used decellularized dermis in cleft repair with a fistula rate of 3.2%. Conclusions: Using 11 studies comprising 2505 children, we find the rate of reported fistula occurrence to be 4.9%. Furthermore, patients with a Veau IV cleft are significantly more likely to develop an oronasal fistula. When fistulae do occur, they do so most often at the soft palate-hard palate junction. A deeper understanding of fistula formation will help cleft palate surgeons improve their outcomes in the operating room and will allow them to effectively communicate expectations with patients families in the clinic. KEY WORDS: cleft palate, fistula, meta-analysis, oronasal fistula, palatoplasty Oronasal fistulae (ONF) are the bane of the cleft palate surgeon. Although there is exhaustive literature of individual and institutional experience and complications following primary cleft palate repair, the lack of a standardized definition of what constitutes a fistula results in a wide range of rates of occurrence, from 0.7% to.60% (Cohen et al., 1991; Rohrich et al., 1996; Emory et al., 1997; Muzaffar et al., 2001; Eberlinc and Kozelj, 2012; Mahoney et al., 2013). Comparative analysis has been made more difficult given the inconsistency of reporting surgical outcome details, inclusion or exclusion of submucous cleft palate repair, a wide range of patient populations, and myriad surgical techniques. Many factors have been suggested to increase risk for development of an ONF, including surgeon s experience (Jackson et al., 2004), Treacher Collins syndrome (Bresnick et al., 2003), extent of cleft, procedure selection (Bekerecioglu et al., 2005; Bindingnavele et al., 2008; Steinbacher et al., 2011), and age at palatal closure (Bresnick et al., 2003). In the face of these challenges, the goal of this study was to perform a meta-analysis to answer the questions: What is the rate of ONF formation following primary cleft palate repair, and what risk factors are associated with their development? Dr. Bykowski is Resident and Dr. Naran is Resident, University of Pittsburgh Medical Center, Department of Plastic Surgery; Mr. Winger is graduate student, University of Pittsburgh, Department of Statistics; Dr. Losee is Professor, Children s Hospital of Pittsburgh, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania. This paper was presented at the 59th Plastic Surgery Research Council Annual Meeting, March 7 9, 2014, in New York City at The Conrad Hotel; and the 71st Annual Meeting of the American Cleft Palate Craniofacial Association, March 24 29, 2014, at the Indianapolis Downtown Marriott Hotel, Indianapolis, Indiana. Submitted July 2014; Accepted June Address correspondence to: Dr. Joseph E. Losee, 45th and Penn, Pittsburgh, PA joseph.losee@chp.edu. DOI: / Literature Search METHODS The MEDLINE database was systematically reviewed for papers written in English and published between January 1, 2000, and October 1, 2012, using the following search items: cleft palate fistula and cleft palate surgery. These search items were arranged using the Boolean operator NOT. Additional searches were performed manually through reference lists of review articles and relevant studies. The Methods, Results, and 81

2 82 Cleft Palate Craniofacial Journal, July 2015, Vol. 52 No. 4 TABLE 1 Collection Inclusion and Exclusion Criteria Applied to Data Inclusion Criteria Primary cleft palate repair Average/median age at surgery,4 y Postoperative follow-up.3 mo Number of fistulae reported in article (Anatomic definition of oronasal fistula) Exclusion Criteria Preclinical animal studies Case reports Patients with PFCS type V VII fistulae* Submucous cleft palate repairs Case series of wide cleft palate repairs * PFCS ¼ Pittsburgh Fistula Classification System. Statistical Analysis FIGURE 1 Schematic of the Pittsburgh Fistula Classification System. * Reprinted with permission from The Cleft Palate Craniofacial Journal. Smith DM, Vecchione L, Jiang S, Ford M, Deleyiannis FW, Haralam MA, Naran S, Worrall CI, Dudas JR, Afifi AM, et al. The Pittsburgh Fistula Classification System: a standardized scheme for the description of palatal fistulas. Cleft Palate Craniofac J. 2007;44: doi: / Conclusions sections of each article were reviewed in detail to evaluate for relevance. Study Eligibility Criteria Inclusion criteria included (1) primary cleft repair; (2) average or median age at time of surgery of,4 years; (3) postoperative follow-up period of.3 months; (4) a description of an ONF as a failure of healing or a breakdown of the primary surgical repair of palate (Muzaffar et al., 2001); and (5) the number of fistulae must have been reported. Exclusion criteria included (1) preclinical animal studies; (2) case reports; (3) patients with a type V through type VII fistula, as defined by the Pittsburgh Fistula Classification System (Fig. 1); (4) repair of submucous cleft palates; and (5) case series of wide cleft palate repairs. These criteria are outlined in Table 1. Outcome Measures The primary outcome targeted for analysis was ONF formation following primary cleft palate repair. For subgroup analysis, a group of studies were analyzed together if the definition of the ONF was described in the report as a failure of healing or a breakdown of the primary surgical repair of palate (Muzaffar et al., 2001). For each selected article, data on number of subjects, rate of ONF formation, study period, retrospective versus prospective study, inclusion of syndromic children, location of fistula, Veau classification of cleft, type of surgery, whether decellularized dermis was used, and follow-up times were collected. A random effects meta-analysis of proportions and exact confidence intervals was performed in Stata 11 (StataCorp. LP, College Station, TX). For Veau classifications, an extension of the Cochran-Mantel- Haenszel Test for a series of tables was used in SAS 9.3 (SAS Institute, Inc., Cary, NC). The Cochran Q statistic was performed to determine whether the selected articles were homogeneous enough to be analyzed together in the meta-analysis. The heterogeneity analysis is a chi-square test used to evaluate the statistical significance of the heterogeneity between studies, and a P value of less than 0.1 was deemed to show heterogeneity. Heterogeneity was further quantified by calculating I-squared, which is the percentage of variability in the results due to heterogeneity rather than sampling error or chance. To further focus on a homogeneous subset of studies, those with a fistula rate more than four standard deviations away from the original meta-analysis pooled estimate were excluded. A forest plot was used to graphically represent effect sizes. RESULTS Between January 1, 2000, and October 1, 2012, there were 70 research articles (of which 86.6% were retrospective studies) that met the inclusion/exclusion criteria (Table 1), with the exception that the authors of these articles did not specifically describe their definition of a fistula. A separate analysis is discussed below to include only those that define an ONF as a failure of healing or a breakdown of the primary surgical repair of palate. The 70 studies presented 12,410 patients with a fistula rate of 6.9% (852 fistulae). The locations of the ONF, based upon the Pittsburgh Fistula Classification System (PFCS; Fig. 1), were as follows: uvular (type I), 0.0%; soft palate (type II), 15.7%; soft palate-hard palate junction (type III), 50.0%; and hard palate (type IV), 32.6%, with the remaining reported as a combination of locations not otherwise specified. There was a signifi-

3 Bykowski et al., RATE OF ORONASAL FISTULA FOLLOWING CLEFT PALATE SURGERY: A META-ANALYSIS 83 TABLE 2 Reported Incidence of Combined Oronasal Fistula Rates of 12,410 Patients Separated by Surgical Technique Surgical Technique Fistula Rate (%) Furlow 6.6 Two-flap 5.1 Wardill-Kilner 12.5 von Langenbeck 11.5 Sommerland 14.3 Other 6.7 cant relationship between Veau classification and the occurrence of an ONF (P,.001), with ONF most prevalent in patients with a Veau IV cleft. The following are the ONF rates for Veau I, II, III, and IV cleft palates: 2.6%, 8.0%, 6.9%, and 9.1%, respectively. The rate of fistula occurrence did not correlate with the surgical technique used for palate repair (Table 2). One study used decellularized dermis in cleft repair with a fistula rate of 3.2% (Helling et al., 2006). Specifically, the decellularized dermis was placed between the nasal and oral closure layers at the soft palate-hard palate junction. Subgroup analysis was used to further analyze the collected pool of data (Fig. 2). After evaluating the initial 70 studies, we excluded 53 studies because the definition of ONF was not outlined in the article. Of the remaining 17 studies, six were then excluded because they had a detrimental effect on the homogeneity of studies (as described previously in the Statistical Analysis section); they were eliminated because they were four standard deviations from the original meta-analysis pooled estimate FIGURE 2 Flow diagram demonstrating inclusion and exclusion of studies as well as the multiple levels of analysis. from the 17 studies. After excluding the six outliers, the data set comprised 11 studies (100% of the studies were retrospective reviews) and 2505 children. The final subgroup analysis contains only 20.2% of patients who were included in the initial analysis of 70 research reports. These studies were found to be statistically comparable to each other, meeting the homogeneity assumption with an acceptable I-squared value of 25.3% and a nonsignificant FIGURE 3 Forest plot for weighted oronasal fistula rates of the final 11 studies that were included in the meta-analysis.

4 84 Cleft Palate Craniofacial Journal, July 2015, Vol. 52 No. 4 heterogeneity chi-square P value (P ¼.203). The primary outcome is the weighted rate of ONF formation, which we found to be 4.9% (95% confidence interval [CI] 3.8 to 6.1%) in our meta-analysis (Fig. 3). DISCUSSION For decades, cleft surgeons have developed many techniques spanning the reconstructive ladder to prevent and address ONF formation following cleft palate repair. Despite the effort spent addressing this problem, its magnitude is unknown, with reported rates ranging from 0to.60% (Cohen et al., 1991; Rohrich et al., 1996; Emory et al., 1997; Muzaffar et al., 2001; Eberlinc and Kozelj, 2012; Mahoney et al., 2013). This wide range is of little prognostic value and results, in part, from the lack of a standardized definition of ONF. The goal of this metaanalysis was to comprehensively evaluate the literature using strict and specific inclusion/exclusion criteria to determine the true incidence of ONF formation and to identify risk factors that contribute to their development. To our knowledge, this report is the first long-term metaanalysis studying ONF rate and risk factors following primary palatoplasty. Limitations of the Literature Comparative analysis is difficult due to the inconsistency of reporting surgical outcome details, inclusion/ exclusion of submucous cleft palate repair, a wide range of patient populations, and many surgical techniques with modifications. Some groups report a fistula only if it is clinically significant (Pigott et al., 2002; Inman et al., 2005), which is inherently subjective. Our final inclusion criteria excluded these studies so that only studies that defined an ONF as a failure of healing or a breakdown of the primary surgical repair of palate were included. In a prior effort to standardize ONF, our research group has previously published a classification system based on anatomic location the Pittsburgh Fistula Classification System (Fig. 1) (Smith et al., 2007). In this meta-analysis, the inclusion/exclusion criteria were used to evaluate children who underwent primary cleft palate repair with adequate follow-up, with the goal of determining fistula rate in the typical case (i.e., a routine cleft palate repair). As such, studies with children who were on average older than 4 years old or case series that specifically investigated children with wide clefts were excluded. Bardach defined a wide cleft palate as one in which the distance between the medial edges of the hard palate is.1.5 cm (Bardach, 1999). Multiple prior studies draw a link between palatal cleft width and development of fistula (Helling et al., 2006; Parwaz et al., 2009; de Agostino Biella Passos et al., 2013); although, other studies exist that are not consistent with this finding (Mahoney et al., 2013). However, there is no consensus in the literature on what defines a wide cleft. As such, we excluded any study that stated specifically that they focused on wide clefts. Our meta-analysis determined that the reported rate of ONF formation following primary cleft palate repair is 4.9% (95% CI, 3.8% to 6.1%). To allow our analysis to be maximally useful to a large population of children who undergo cleft palate repair, we used subgroup analysis to evaluate the effect of potential risk factors on ONF formation. In this regard, Veau classification appears to modify risk for development of ONF. Discrepancies exist that complicate surgical outcome analysis following primary cleft palate repair. Because there is wide variability in the reporting of such repair (Eberlinc and Kozelj, 2012), a comprehensive analysis must include subgroup analysis. Inclusion of fistulae that are intentionally not repaired at the time of palatoplasty specifically, lingual-alveolar and/or labial-alveolar fistulae (PFCS types VI and VII) can be misleading. Rather, more clinically interesting are true ONF that represent surgical failure. Our goal was to conduct a more clinically relevant analysis by excluding PFCS types V, VI, and VII fistulae (junction of primary and secondary palate, lingual-alveolar, and labial-alveolar, respectively). The PFCS type I fistulae (uvular) are often not repaired because they do not seem to cause dyslalia, hypernasal resonance, or nasal air emission. Some studies (Schultz, 1986; Muzaffar et al., 2001; Losee et al., 2008) included only patients with clinically significant ONF; whereas, others did not specify the clinical significance (Cohen et al., 1991; Brusati and Mannucci, 1994; Wilhelmi et al., 2001; Sommerlad, 2003; Inman et al., 2005). Phua and de Chalain (2008) and Emory et al. (1997) included only palate fistulae posterior of the incisive foramen. Due to a retrospective nature and incomplete medical records, some studies report their fistula rate while appropriately noting that they were unable to determine whether some fistulae were deliberately left open (Phua and de Chalain, 2008). In fact, most studies (87% to 100%, depending of level of exclusion) in this meta-analysis were retrospective and, thus, vulnerable to the inherent biases associated with a retrospective review. First Level of Analysis Our first level of analysis included 12,410 patients; however, many of these reports lack the details (including how they defined a fistula) necessary for critical evaluation and comparison. Nonetheless, even though some details are missing, evaluation of this large cohort is useful.

5 Bykowski et al., RATE OF ORONASAL FISTULA FOLLOWING CLEFT PALATE SURGERY: A META-ANALYSIS 85 Consistent with several other papers (Amaratunga, 1988; Moore et al., 1988; Rohrich et al., 1996; Landheer et al., 2010), half of the ONF that were reported occurred at the soft palate-hard palate junction (PFCS type III) notably, a dynamic area that is often closed under tension. Accordingly, surgeons have developed techniques in attempts to ensure soft tissue closure and prevent breakdown, including hamulus fracture, wide relaxing incisions, and, even more aggressive, buccal flaps or microvascular free flaps (Zemann et al., 2011; Christiano et al., 2012). Furthermore, our group has successfully instituted placement of acellular dermal matrix (ADM) prophylactically during primary cleft repair with fistula rate of 0.76% (Losee et al., 2008). Although sufficient prospective studies do not exist, ADM may decrease ONF formation when placed at the soft palate-hard palate junction (PFCS type III) (Clark et al., 2003; Cole et al., 2006; Helling et al., 2006; Kirschner et al., 2006; Steele and Seagle, 2006). Although many studies and case series of primary cleft palate repair have been reported, the primary aim of many of these studies was not to assess fistula rates. For this reason, it is not surprising that critical evaluation of the studies had to be further analyzed and many studies excluded. After analyzing the 12,410 patients, we then excluded all of the studies that did not specifically define an ONF as a failure of healing or a breakdown of the primary surgical repair of palate. Coincidentally, despite applying these exclusion criteria, the fistula rate was similar at 6.4%. Subsequently, the remaining 17 papers were analyzed. Six of these papers were considered outliers. Finally, the fistula rate of the 11 papers (2,505 patients) was noted to be 4.9% (95% CI, 3.8% to 6.1%), which again is similar to the rate of the 12,410 patients. Predictive Risk Factors Our analysis also demonstrates that the rate of ONF varies with the type of cleft. Corroborating other studies, we found that the ONF rate was highest for Veau type IV clefts (Cohen et al., 1991; Muzaffar et al., 2001; Andersson et al., 2008; Parwaz et al., 2009; Eberlinc and Kozelj, 2012). The rate of Veau IV clefts was 9.1% in our analysis, which is similar to those of other groups (Phua and de Chalain, 2008; Eberlinc and Kozelj, 2012). Intuition tells us that the more severe the cleft, the more technically challenging the reconstruction and the higher the chance of ONF formation. Several groups report that there is an association between the extent of clefting and the fistula rate (Amaratunga, 1988; Cohen et al., 1991; Rohrich et al., 1996; Muzaffar et al., 2001; Eberlinc and Kozelj, 2012; de Agostino Biella Passos et al., 2013). We find it interesting that there is a small study reporting that not only is the width of the cleft important, but the ratio of cleft width to the posterior arch width is a predictor of ONF formation (Parwaz et al., 2009). Surgical Technique Many surgical techniques and modifications exist to reconstruct the cleft palate, but there are some basic techniques that are widely used. When evaluating 12,410 patients, we found differences in the aggregate raw data comparing rate of ONF occurrence across surgical techniques (Table 2). Although it is interesting to evaluate these aggregate raw data, applying metaanalysis statistical method is inappropriate. There are many confounding factors (e.g., operating surgeon, study population, modifications to procedure) that can potentially impact ONF rate. Our data demonstrate fistula rates vary only slightly among the techniques widely used for primary repair of cleft palate. Nonetheless, previous reports have described a high ONF rate following von Langenbeck palatoplasty compared with Veau palatoplasty (Amaratunga, 1988). Schultz (1986), however, found no correlation between the type of palatoplasty and the rate of ONF occurrence. We classified only primary surgical techniques (Table 2). Many surgeons have their own modification. For example, Bindingnavele et al. (2008) studied results when the Furlow double-opposing Z-plasty was performed with or without greater palatine flap mobilization and islandization. Although these are different procedures, two different surgeons may refer to each as a Furlow palatoplasty. However, Bindingnavele et al. (2008) found that there was a statistically significant difference in fistula formation between the two modifications of the same procedure. Of course, this complicates analysis of ONF formation among surgeons. Study Limitations A meta-analysis is a powerful tool to combine studies typically randomized controlled trials to generate an integrated cohort of patients under similar experimental or observational conditions. However, randomized controlled trials are largely impractical in plastic surgery outcomes research. Instead, the plastic surgery literature comprises individual and institutional case series. Despite the utility of a meta-analysis, there are drawbacks, such as the inherent publication biases associated with each report that composes the metaanalysis. Of course, there is variability among the studies included in a meta-analysis, such as which patient factors are reported from each study. For example, some studies in this meta-analysis included children with syndromes associated with clefts, others did not, and others did not comment.

6 86 Cleft Palate Craniofacial Journal, July 2015, Vol. 52 No. 4 Not every visible ONF causes dysfunction and thus some do not require surgical repair. To the patient, the most important measure is whether the fistula is symptomatic. Unfortunately, few studies report this information, and again, there is no standardized way to objectively report palatal dysfunction. It is for this reason that our meta-analysis focused on studies that defined an ONF as an anatomic entity rather than a functional one. Despite being an objective assessment, defining an ONF strictly by anatomical measures can be misleading because not all fistulae are clinically significant. The nature of a meta-analysis precludes conclusions of causality and only allows us to draw associations. Although the goal is to combine similar studies to increase the sample size, there are inherent biases and confounding factors in a meta-analysis. For example, there is the possibility that some patients who developed an ONF did not follow up with the same surgeon. Because there was no randomization process in the individual studies, there is potential for multiple unidentified confounders. Common shortcomings of case series (e.g., potential loss of follow-up, publication bias, and outcome reporting bias) can be functionally amplified when these studies are analyzed together. A large, multicenter, randomized trial is necessary but challenging in cleft palate surgery. CONCLUSIONS Many studies have been conducted assessing ONF formation following cleft palate repair. Most of these studies were small, underpowered, and the clinical experience of one author. Therefore, we conducted the first long-term, comprehensive systematic review and meta-analysis to evaluate the rate and risk factors for ONF formation. Consistency of classifying ONF is imperative to appropriately assess surgical outcomes, which allows meaningful comparisons, and ultimately leads to improving treatment strategies and providing families with a more thorough informed surgical consent. We offer a cohesive group of data to which practitioners can refer when preoperatively counseling caregivers of children undergoing palatoplasty. REFERENCES Amaratunga NA. (1988). Occurrence of oronasal fistulas in operated cleft palate patients. J Oral Maxillofac Surg. 1988;46: Andersson EM, Sandvik L, Semb G, Abyholm F. Palatal fistulas after primary repair of clefts of the secondary palate. Scand J Plast Reconstr Surg Hand Surg. 2008;42: Bardach J, ed. Atlas of Craniofacial and Cleft Surgery. Vol. 2. Philadelphia: Lippincott Raven; Bekerecioglu M, Isik D, Bulut O. Comparison of the rate of palatal fistulation after two-flap and four-flap palatoplasty. Scand J Plast Reconstr Surg Hand Surg. 2005;39: Bindingnavele VK, Bresnick SD, Urata MM, Huang G, Leland HA, Wong D, Hammoudeh J, Reinisch J. Superior results using the islandized hemipalatal flap in palatoplasty: experience with 500 cases. Plast Reconstr Surg. 2008;122: Bresnick S, Walker J, Clarke-Sheehan N, Reinisch J. Increased fistula risk following palatoplasty in Treacher Collins syndrome. Cleft Palate Craniofac J. 2003;40: Brusati R, Mannucci N. Repair of the cleft palate without lateral release incisions: results concerning 124 cases. J Craniomaxillofac Surg. 1994;22: Christiano JG, Dorafshar AH, Rodriguez ED, Redett RJ. Repair of recurrent cleft palate with free vastus lateralis muscle flap. Cleft Palate Craniofac J. 2012;49: Clark JM, Saffold SH, Israel JM. Decellularized dermal grafting in cleft palate repair. Arch Facial Plast Surg. 2003;5:40 44, discussion 45. Cohen SR, Kalinowski J, LaRossa D, Randall P. Cleft palate fistulas: a multivariate statistical analysis of prevalence, etiology, and surgical management. Plast Reconstr Surg. 1991;87: Cole P, Horn TW, Thaller S. The use of decellularized dermal grafting (AlloDerm) in persistent oro-nasal fistulas after tertiary cleft palate repair. J Craniofac Surg. 2006;17: de Agostino Biella Passos V, de Carvalho Carrara CF, da Silva Dalben G, Costa B, Gomide MR. Prevalence, cause, and location of palatal fistula in operated complete unilateral cleft lip and palate: retrospective study. Cleft Palate Craniofac J. 2014;51: Eberlinc A, Kozelj V. Incidence of residual oronasal fistulas: a 20-year experience. Cleft Palate Craniofac J. 2012;49: Emory RE Jr, Clay RP, Bite U, Jackson IT. Fistula formation and repair after palatal closure: an institutional perspective. Plast Reconstr Surg. 1997;99: Helling ER, Dev VR, Garza J, Barone C, Nelluri P, Wang PT. Low fistula rate in palatal clefts closed with the Furlow technique using decellularized dermis. Plast Reconstr Surg. 2006;117: Inman DS, Thomas P, Hodgkinson PD, Reid CA. Oro-nasal fistula development and velopharyngeal insufficiency following primary cleft palate surgery an audit of 148 children born between 1985 and Br J Plast Surg. 2005;58: Jackson IT, Moreira-Gonzalez AA, Rogers A, Beal BJ. The buccal flap a useful technique in cleft palate repair? Cleft Palate Craniofac J. 2004;41: Kirschner RE, Cabiling DS, Slemp AE, Siddiqi F, LaRossa DD, Losee JE. Repair of oronasal fistulae with acellular dermal matrices. Plast Reconstr Surg. 2006;118: Landheer JA, Breugem CC, van der Molen AB. Fistula incidence and predictors of fistula occurrence after cleft palate repair: two-stage closure versus one-stage closure. Cleft Palate Craniofac J. 2010;47: Losee JE, Smith DM, Afifi AM, Jiang S, Ford M, Vecchione L, Cooper GM, Naran S, Mooney MP, Serletti JM. A successful algorithm for limiting postoperative fistulae following palatal procedures in the patient with orofacial clefting. Plast Reconstr Surg. 2008;122: Mahoney MH, Swan MC, Fisher DM. Prospective analysis of presurgical risk factors for outcomes in primary palatoplasty. Plast Reconstr Surg. 2013;132: Moore MD, Lawrence WT, Ptak JJ, Trier WC. Complications of primary palatoplasty: a twenty-one-year review. Cleft Palate J. 1988;25: Muzaffar AR, Byrd HS, Rohrich RJ, Johns DF, LeBlanc D, Beran SJ, Anderson C, Papaioannou AA. Incidence of cleft palate fistula: an institutional experience with two-stage palatal repair. Plast Reconstr Surg. 2001;108:

7 Bykowski et al., RATE OF ORONASAL FISTULA FOLLOWING CLEFT PALATE SURGERY: A META-ANALYSIS 87 Parwaz MA, Sharma RK, Parashar A, Nanda V, Biswas G, Makkar S. Width of cleft palate and postoperative palatal fistula do they correlate? J Plast Reconstr Aesthet Surg. 2009;62: Phua YS, de Chalain T. Incidence of oronasal fistulae and velopharyngeal insufficiency after cleft palate repair: an audit of 211 children born between 1990 and Cleft Palate Craniofac J. 2008;45: Pigott RW, Albery EH, Hathorn IS, Atack NE, Williams A, Harland K, Orlando A, Falder S, Coghlan B. A comparison of three methods of repairing the hard palate. Cleft Palate Craniofac J. 2002;39: Rohrich RJ, Rowsell AR, Johns DF, Drury MA, Grieg G, Watson DJ, Godfrey AM, Poole MD. Timing of hard palatal closure: a critical long-term analysis. Plast Reconstr Surg. 1996;98: Schultz RC. Management and timing of cleft palate fistula repair. Plast Reconstr Surg. 1986;78: Smith DM, Vecchione L, Jiang S, Ford M, Deleyiannis FW, Haralam MA, Naran S, Worrall CI, Dudas JR, Afifi AM, et al. The Pittsburgh Fistula Classification System: a standardized scheme for the description of palatal fistulas. Cleft Palate Craniofac J. 2007;44: Sommerlad BC. A technique for cleft palate repair. Plast Reconstr Surg. 2003;112: Steele MH, Seagle MB. Palatal fistula repair using acellular dermal matrix: the University of Florida experience. Ann Plast Surg. 2006;56:50 53, discussion 53. Steinbacher DM, McGrath JL, Low DW. Is nasal mucoperiosteal closure necessary in cleft palate repair? Plast Reconstr Surg. 2011;127: Wilhelmi BJ, Appelt EA, Hill L, Blackwell SJ. Palatal fistulas: rare with the two-flap palatoplasty repair. Plast Reconstr Surg. 2001;107: Zemann W, Kruse AL, Luebbers HT, Jacobsen C, Metzler P, Obwegeser JA. Microvascular tissue transfer in cleft palate patients: advocacy of the prelaminated radial free forearm flap. J Craniofac Surg. 2011;22:

8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.