|Year : 2017 | Volume
| Issue : 4 | Page : 220-225
Risk factors for a persistent gastrocutaneous fistula following gastrostomy device removal: A tertiary center experience
Abdulrahman Alshafei, Dawn Deacy, Brice Antao
Department of Paediatric Surgery, Our Lady's Children's Hospital, Crumlin, Dublin 12, Ireland
|Date of Web Publication||12-Sep-2017|
Department of Paediatric Surgery, Our Lady's Children's Hospital, Crumlin, Dublin 12
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Aim: The aim of this study is to identify the risk factors for a persistent gastrocutaneous fistula (GCF) after gastrostomy device (GD) removal in children.
Materials and Methods: A retrospective analysis of 59 patients that underwent GD insertion and removal over an 11-year period (2005–2015). Patients were divided into two Groups (A and B) according to persistence or closure of the gastrocutaneous tract. Data included patient demographics, comorbidities, age at insertion, gastrostomy site infections, size and type of device, duration of placement, and method of insertion and removal. Statistical analysis was done using Chi-square test and ANOVA test where P < 0.05 was considered statistically significant.
Results: A total of 34 patients (Group A) developed a GCF post-GD removal. The gastrostomy tract closed spontaneously in 25 patients (Group B). Underlying comorbidities did not influence spontaneous closure. Younger age at insertion (<2 years), longer duration of device placement, open gastrostomy insertion, upsizing the GD, changing a gastrostomy tube to a button, and site infections were significant risk factors for a persistent GCF.
Conclusions: Risk analysis of persistent GCF is important for patient counseling before removal or replacement of the GD. We have identified a number of potentially reversible risk factors for a persistent GCF and have made recommendations accordingly.
Keywords: Children, gastrocutaneous fistula, gastrostomy device, risk factors
|How to cite this article:|
Alshafei A, Deacy D, Antao B. Risk factors for a persistent gastrocutaneous fistula following gastrostomy device removal: A tertiary center experience. J Indian Assoc Pediatr Surg 2017;22:220-5
|How to cite this URL:|
Alshafei A, Deacy D, Antao B. Risk factors for a persistent gastrocutaneous fistula following gastrostomy device removal: A tertiary center experience. J Indian Assoc Pediatr Surg [serial online] 2017 [cited 2020 Jul 15];22:220-5. Available from: http://www.jiaps.com/text.asp?2017/22/4/220/214445
| Introduction|| |
Gastrostomy device (GD) insertion is a commonly performed procedure in pediatric patients. They are often used for various periods of time depending on the underlying condition and the child's need to establish enteral access for optimal nutrition. When the GD is no longer required to meet dietary requirements, it is removed and the tract between the stomach and the skin generally closes spontaneously in a matter of hours or days. However, a gastrocutaneous fistula (GCF) has been shown to persist in up to 45% of patients, and this is a distressing problem for both children and their families.,,
Few studies have attempted to determine the reason for a persisting GCF and have found the length of use to be a main contributing factor.,,,, The aim of our study is to evaluate a correlation between comorbidity and patient-related and gastrostomy-related risk factors for developing a GCF post-GD removal.
| Materials and Methods|| |
This was a retrospective review of 59 patients in the period between 2005 and 2015 that underwent GD insertion and removal. The patients were divided into two groups: Group A included 34 patients that developed a persistent GCF post-GD removal. Group B included 25 patients who did not develop a GCF after removal.
Data collected and analyzed included patient demographics, comorbidities, age at insertion, significant gastrostomy site infection (more than one culture proven infection requiring treatment with antibiotics), size and type of GD, duration of GD in situ, method of insertion and removal, and time from removal to closure of the fistula. Complete patient data and demographics are shown in [Table 1].
The GD was inserted using either an open Stamm technique  (n = 17, 29%) or a percutaneous endoscopic gastrostomy (PEG) (n = 42, 71%). Forty-four patients (75%) has a size 12fr Corflo ® tube (Corpak MedSystems, UK) inserted at initial gastrostomy placement. In 24 patients (41%), the tube was upsized at initial or subsequent replacements. In addition, thirty patients (51%) went on to have the gastrostomy tube changed to a MIC-KEY button ® (Kimberly-Clark, USA). The Corflo ® gastrostomy tubes were removed using either a traction method or endoscopically.
The outcome measure was a persistent GCF post-GD removal. Risk factors for development of a GCF were categorized as outlined in [Table 2]. Statistical analysis was done using Chi-square test and ANOVA test where P < 0.05 was considered statistically significant. Approval for this study was obtained from our Institutional Ethics Committee.
|Table 2: Potential risk factors for development of a gastrocutaneous fistula postgastrostomy device removal|
Click here to view
| Results|| |
In Group A, 34 patients (14 male [41%], 20 females [59%]) developed a persistent GCF after GD removal. The mean age at gastrostomy insertion was 24.03 ± 39.59 months and mean duration of GD in situ was 57.65 ± 39.42 months. In Group B, 24 patients (14 males [56%], 11 females [44%]) underwent GD insertion at a mean age of 52.24 ± 67.89 months and the mean duration of GD in situ was 31.68 ± 30.19 months. The indications for insertion of GDs are summarized in [Table 3].
Patient-related risk analyses are summarized in [Table 4]. There were no significant comorbidity-related risk factors identified [Table 5]. Gastroesophageal reflux disease (GERD) was the most common comorbidity in both groups (Group A: 22 cases [65%], Group B: 15 cases [60%]), of which 14 cases (41%) in Group A and 7 cases (29%) in Group B were on regular anti-reflux medication (P = 0.256) at the time of GD removal. Gastrostomy-related risk analyses are summarized in [Table 6].
All patients in Group A underwent an open surgical closure of the persistent GCF at a mean duration of 10.89 ± 13.87 months after removal of the GD.
| Discussion|| |
Although complications related to the insertion of a GD are well documented, complications following removal are not well described in the literature. The reported incidence of a persistent GCF ranges from 16% to 45%.,,,, Our data show an incidence rate of 58% of developing a persistent GCF after GD removal. We believe that our higher incidence may be related to the fact that we have a significant number of young patients aged <2 years that had a GD inserted and in situ for a prolonged duration of time (>2 years). Our high incidence may have also been slightly overestimated due to lack of follow-up in all cases. Although all gastrostomy tubes and most MIC-KEY buttons ® are removed in a hospital setting, a small proportion of cases would have their MIC-KEY buttons ® removed in the community with subsequent spontaneous closure of their gastrostomy tracts.
Most previous studies did not show a correlation between age at which the GD was inserted and a persistent GCF.,,, Evaluating our patient-related factors, gender of the child was not a risk factor for the development of a persistent GCF. However, there was a correlation between age and persistent GCF, similar to the findings from a study by Wyrick et al., where there was a higher incidence of persistent GCF when the GD was inserted at a younger age. In our cohort, children <2 years of age had a higher incidence of GCF. The agreed theory is that with time, the tract eventually completely epithelializes creating a GCF that fails to close spontaneously. It has been shown that epithelialization is a slow process and occurs at a rate of approximately 7 mm/day, over several weeks to months depending on the size of the wound and local conditions. An increased incidence of GCF in younger children could be related to less subcutaneous tissue resulting in a shorter tract and early epithelialization. Furthermore, we had a higher incidence of local wound infections, open gastrostomies, device changes, and prolonged use in this younger cohort of patients, all of which are potential risk factors for a persistent GCF. In our study, patient comorbidities such as GERD, neurological impairment, cardiac anomalies, and immunosuppression did not seem to contribute to a persisting GCF. This is in accordance with the few other reports in the literature that have attempted to determine predictive factors for developing a GCF.,,,,
Several studies have shown that the length of time the GD is in place is the major determining factor in developing a persistent GCF.,,,,, GERDon and Janik and Janik et al. went on to show that GDs removed before 9 months had a significantly lower rate of tract persistence., We found that the duration of time the GD was left in situ was also a significant risk factor for development of a persistent GCF. The majority of patients (80%) who developed a persistent GCF had the GD in situ for more than 2 years with the shortest duration being 1.5 years. There is a tendency to leave the GD in place longer than they are required. Clinicians must therefore be aware that unless clinically indicated the devices must be removed as soon as possible to avoid this complication. This needs to be a decision made with the multidisciplinary team involved in the child's care.
In addition, we were able to establish a correlation between a persistent GCF and a number of gastrostomy-related factors, namely, insertion method, upsizing of the GD, and changing a gastrostomy tube to a button device. In contrast to other studies,, we found that upsizing the GD at the time of replacement was a significant factor contributing to a persistent GCF. Forty-nine patients (83%) had a size 12 Fr Corflo ® tube inserted at initial gastrostomy placement which was upsized to a larger diameter tube in 24 patients (41%). In our institution, a size 12 Fr Corflo ® is the smallest diameter tube used in the majority of pediatric patients. The decision to upsize a tube was generally made by the consultant surgeon at the time of replacement based on the age and weight of the patient. Indications for replacing a gastrostomy tube may include inadvertent dislodgment, blockage, or damage from long-term use. Our data also showed that changing a Corflo ® tube to a MIC-KEY button ® also lead to a persisting GCF. This association, however, could also be a factor of time since children with MIC-KEY buttons ® tend to have devices in situ for a prolonged duration and have frequent changes of the same in the hospital or community. It is difficult to know, if cases of MIC-KEY buttons ® inserted as a primary GD would have a similar significant association with GCF, as none of our cases had a primary MIC-KEY buttons ® inserted. Some patients and families request MIC-KEY buttons ® as they are low profile and more convenient, especially in older children where tubes can interfere with sport activities. They are inserted at the parent's wishes once the tract has fully healed and matured which, in our center, is usually 3–6 months after Corflo ® PEG insertion. These findings should alert physicians that gastrostomy tubes should not be routinely upsized or changed to a MIC-KEY button ® unless absolutely necessary. Parents or carers need to be aware of an increased risk of developing a GCF before deciding on such changes to the GD.
Gastrostomy site infection was another clinical parameter in our study that was significantly associated with a persistent GCF. In the cohort of patients that developed a persistent GCF, 91% of cases experienced at least two gastrostomy site infections requiring treatment with antibiotics. This association is probably explained by the resultant healing and scar tissue formation that creates a tract that is unlikely to close on its own.
While GD insertion has traditionally been performed through the open Stamm technique, this method is now being widely replaced by percutaneous and laparoscopic approaches. The majority of our patients (71%) underwent gastrostomy tube placement using the percutaneous endoscopic technique (PEG) while only 29% were done using the open method all of which during concomitant fundoplication. GERDon and Langer were the first to suggest that tubes inserted percutaneously were less likely to develop a persisting fistula than those inserted with the Stamm technique although their data did not reach statistical significance. Several other studies also showed that the insertion technique, whether it was percutaneous or open, had no effect on gastrostomy closure., However, it is important to note that the number of PEG insertions in those studies was possibly too small to arrive at any conclusion. We report a higher number of PEG insertions, and our data have shown a lower incidence of GCF with this technique compared to the open technique. Although studied in comparison to the laparoscopic approach, Wyrick et al. also concluded that open operations and concurrent fundoplication were statistically significant factors causing persistent GCF. We were not able to compare the laparoscopic approach as our cohort only included the open and PEG technique. However, when we utilize the laparoscopic approach, the gastrostomy tube is inserted using the percutaneous endoscopic pull technique, as described in the literature., As such, we would concur with Wyrick et al. that laparoscopic gastrostomy insertion would have a lower incidence of GCF compared to the open technique. The removal method, whether with traction or endoscopic retrieval, did not seem to significantly affect spontaneous gastrostomy closure.
GCFs are generally managed expectantly for a period of time to allow for spontaneous closure. Although pharmacological agents such as Somatostatin analogs have been successful in the treatment of enterocutaneous fistulas, its role in the management of GCF has not been well document., Although the majority of our patients (65%) were on proton-pump inhibitors for gastroesophageal reflux, Somatostatin analogs were not used in any of our patients. Our policy is watch-and-wait for 2–3 months and any patient with a persistent fistula beyond that time is planned for open surgical repair. Our preferred approach is to excise the fistula tract down to the level of the stomach, and the gastric defect is repaired in two layers with absorbable sutures. This is followed by a layered abdominal wall and subcutaneous tissue closure. We have no documented leaks or recurrences after surgical closure. There have also been recent reports with variable results of minimally invasive methods including endoscopic cautery, clip closure, and fibrin glue injection.,
| Conclusions|| |
Risk analysis of persistent GCF is crucial for patient counseling as well as identifying potentially reversible factors that could reduce the incidence of this complication. Our study identifies the following contributing factors for a persistent GCF post-GD removal:
- Younger patients less than 2 years of age
- Open Stamm gastrostomy technique
- Upsizing of GD
- Change of a gastrostomy tube to a button
- Significant gastrostomy site infections
- Prolonged duration of GD more than 2 years.
Based on the above findings, the authors make the following recommendations to reduce the risk of developing a persistent GCF after removal of the GD:
- Use of the percutaneous endoscopic technique, whenever feasible
- Use of the smallest size tube possible that meets patient requirements
- Avoid upsizing the GD or changing the type of device unless absolutely necessary
- Antibiotic cover during the procedure and adequate and prompt treatment of gastrostomy site infections
- Remove GD as soon as feasible and avoid tendency to keep it in place as a “safe measure” unless clinically indicated.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
El-Rifai N, Michaud L, Mention K, Guimber D, Caldari D, Turck D, et al.
Persistence of gastrocutaneous fistula after removal of gastrostomy tubes in children: Prevalence and associated factors. Endoscopy 2004;36:700-4.
GERDon JM, Langer JC. Gastrocutaneous fistula in children after removal of gastrostomy tube: Incidence and predictive factors. J Pediatr Surg 1999;34:1345-6.
Janik TA, Hendrickson RJ, Janik JS, Landholm AE. Analysis of factors affecting the spontaneous closure of a gastrocutaneous fistula. J Pediatr Surg 2004;39:1197-9.
Bratu I, Bharmal A. Incidence and predictors of gastrocutaneous fistula in the pediatric patient. ISRN Gastroenterol 2011;2011:686803.
Wyrick DL, Bozeman AP, Smith SD, Jackson RJ, Maxson RT, Kelley KR, et al.
Persistent gastrocutaneous fistula: Factors affecting the need for closure. J Pediatr Surg 2013;48:2506-10.
Shellito PC, Malt RA. Tube gastrostomy. Techniques and complications. Ann Surg 1985;201:180-5.
Gauderer MW. Percutaneous endoscopic gastrostomy-20 years later: A historical perspective. J Pediatr Surg 2001;36:217-9.
Kobak GE, McClenathan DT, Schurman SJ. Complications of removing percutaneous endoscopic gastrostomy tubes in children. J Pediatr Gastroenterol Nutr 2000;30:404-7.
Gauderer MW, Stellato TA. Gastrostomies: Evolution, techniques, indications, and complications. Curr Probl Surg 1986;23:657-719.
Almeleh R. Spontaneous accelerated epithelialization in deep dermal burns using an oxygen-delivering hydrogel: A report of two cases. Wounds 2013;25:E18-25.
Stringel G, Geller ER, Lowenheim MS. Laparoscopic-assisted percutaneous endoscopic gastrostomy. J Pediatr Surg 1995;30:1209-10.
Yu SC, Petty JK, Bensard DD, Partrick DA, Bruny JL, Hendrickson RJ. Laparoscopic-assisted percutaneous endoscopic gastrostomy in children and adolescents. JSLS 2005;9:302-4.
Papavramidis TS, Mantzoukis K, Michalopoulos N. Confronting gastrocutaneous fistulas. Ann Gastroenterol 2011;24:16-9.
Dorta G. Role of octreotide and somatostatin in the treatment of intestinal fistulae. Digestion 1999;60 Suppl 2:53-6.
Teitelbaum JE, Gorcey SA, Fox VL. Combined endoscopic cautery and clip closure of chronic gastrocutaneous fistulas. Gastrointest Endosc 2005;62:432-5.
González-Ojeda A, Avalos-González J, Muciño-Hernández MI, López-Ortega A, Fuentes-Orozco C, Sánchez-Hochoa M, et al.
Fibrin glue as adjuvant treatment for gastrocutaneous fistula after gastrostomy tube removal. Endoscopy 2004;36:337-41.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]