Facebook Twitter YouTube

Persistent Gastrocutaneous Fistula: Factors Affecting Need for Closure

Sunday, October 21, 2012: 10:32 AM
Versailles Ballroom (Hilton Riverside)
Andrew P. Bozeman1, Melvin S. Dassinger1, Richard J. Jackson1, Robert T. Maxson1, Karen R. Kelley1, Donna L. Mathews1, Jingyun Li2, Christopher J. Swearingen2 and Samuel D. Smith1, (1)Department of Pediatric Surgery, University of Arkansas for Medical Sciences, Arkansas Children's Hospital, Little Rock, AR, (2)Department of Biostatistics, University of Arkansas for Medical Sciences, Arkansas Children's Hospital, Little Rock, AR


The occurrence of gastrocutaneous fistula (GCF) is a well-known complication.  We explore multiple factors to ascertain their impact on the rate of persistent GCF formation.


We retrospectively reviewed patient records for all gastrostomies (GT) constructed at our children’s hospital from 2007- 2011.  Method of placement (laparoscopic vs. open), concomitant fundoplication, elapsed time of therapy, and demographics, were evaluated.  Both Mann-Whitney-Wilcoxon and Chi-square tests were used for data analysis.  A logistic regression model estimated the odds of requiring surgical closure secondary to GCF for each of the two surgical placement methods adjusting for the elapsed time of GT, patient’s age at the time GT placement, gender and need for Nissen fundoplication.


Among the 950 patients who had GTs during the study period, 148 patients had GTs removed; of those, 47 (32%) patients required a surgical closure secondary to persistent GCF.  Laparoscopic and open procedures comprised 79/148 (53%) and 69/148 (47%), respectively.  Of the 79 patients with laparoscopic GTs, 17 (22%) developed a persistent GCF, compared to 30 (43%) in the open group (OR= 2.52, P= 0.035). 

In total, 71 patients had a Nissen fundoplication in combination with GT (34/79 patients in the laparoscopic group; 37/69 patients in the open group).  Of the patients with Nissen fundoplication, 31/71 (44%) developed GCF, compared to 16/77 (21%) of patients without Nissen fundoplication (OR= 4.94, P= 0.002).

There was a significant relationship between the duration of elapsed time between placement and removal of GT and the incidence of persistent GCF.  When compared to the 25th percentile (301 days had estimated GCF incidence of 13%), those patient in the 50th (537 days had 45% incidence) experienced a greater incidence of GCF formation (OR = 3.41, P < 0.001).  Those patients in 75th percentile (848 days) had the highest incidence (50%) (OR = 4.52, P = 0.011). 

Patients aged 7.2 weeks (25th percentile, 37% incidence) were more likely to develop GCF when compared to patients aged 34.4 weeks (75th percentile, 24% incidence) (OR = 2.39, P = 0.017).

When considered together, duration of GT therapy and concomitant Nissen fundoplication have an additive effect of the risk of GCF development.  (Figure 1)


Duration of GT (Days)

Nissen Fundoplication

Incidence of GCF
















Figure 1



In our study, open GTs have twice the rate of GCF formation as compared to the laparoscopic approach.  An even greater reduction in GCF is realized without concomitant Nissen fundoplication.  In addition, younger patient age and greater duration of therapy are associated with higher risk of GCF development.