Abstract
Abstract
Purpose:
Traditional methods for securing a laparoscopic gastrostomy (LG) involve the placement of two monofilament transabdominal (TA) sutures to be removed after a short interval of 5 days. A modified technique employing an absorbable suture tunneled subcutaneously has been adopted by many surgeons. The aim of this study was to compare wound complications between these techniques.
Methods:
A retrospective review of patients who underwent LG placement between 2010 and 2016 was conducted, dividing patients into two cohorts by securing stitch type, TA and subcutaneous (SC), and evaluating for complications.
Results:
A total of 740 children underwent laparoscopic gastrostomy tube (GT) placement, of whom 554 (75%) patients had a TA stitch and the remaining 186 (25%) had a SC stitch. Demographic data were comparable in both groups. The most common wound complication was granulation tissue (22%), dislodgement (19%), external drainage (16%), cellulitis (10%), erosion (3%), and abscess formation (2%). Seven patients required operative revision for dislodgement; TA patients comprised the majority of these patients. Operative times were significantly longer in the SC group (22 minutes versus 28 minutes, P < .05). Rates of granulation, erosion, external and internal leakage, and dislodgement were equivalent between cohorts. There were higher rates of cellulitis (7.3% versus 19%, P < .05) and abscess (0.8% versus 7.6%, P < .05) noted in the SC group. Time to external leakage was significantly earlier in the SC group (P < .05); however, all other complications occurred at comparable times following initial operation. Persistent gastrocutaneous fistula requiring surgical closure occurred at equal rates with no difference in times to closure from GT discontinuation in both groups.
Conclusion:
While both techniques are feasible, there was a significant increase in infectious complications and operative times observed in the SC stitch patients, suggesting this may not be the optimal securing method.
Background
F
The most commonly described and implemented laparoscopic technique employs the use of a temporary U-stitch fixation of a primarily placed gastrostomy button.2,3 However, rising concerns for poor gastric apposition to the anterior abdominal wall and subsequent intraperitoneal leak with this method have resulted in the development of alternative techniques for gastrostomy button fixation. A recent modification utilizes an absorbable suture tunneled in the subcuticular space to secure the gastrostomy button in a more permanent fashion to the abdominal wall. Currently, smaller series exist evaluating this technique citing lower complication rates.4,5,7
Our institution has adopted both techniques for gastrostomy fixation in the past several years. Anecdotally, we have noted the possibility of increased wound complications. As such, the aim of this study was to compare complications between transabdominally placed U-stitches to the newer subcutaneously tunneled permanent suture.
Materials and Methods
After Institutional Review Board approval (15060241), a retrospective chart review of patients who underwent a laparoscopic gastrostomy (LG) at a single center from May 2010 to May 2016 was done. Patients were divided into two cohorts based on securing suture type: transabdominal (TA) and subcutaneous (SC).
As previously described, the TA cohort had two monofilament sutures placed as a U-stitch, traveling through the abdominal wall, through the stomach, and back out through the abdominal wall.2,3 The needle is bent and pulled cephalad before the needle driver is turned to exit the abdomen. This intraoperative measure diminishes the superior to inferior space between entry and exit points of the TA stitch, resulting in a more secure stitch without pinching of the skin. Sutures were tied overtop the flange of the gastrostomy button and removed 5 days after initial placement. Similar to the previous technique, the SC stitch was passed through the abdominal wall, through the stomach, out through the fascia and then, tunneled in the SC space with the knot tied in a buried fashion beneath the skin. 7 These sutures remain in place until they slowly become absorbed over the next few months with wound healing. Specific intraoperative methods for LG sizing were not uniform among all surgeons.
Demographic data were collected. The primary objective was to compare wound complications between securing techniques. Complications were identified by appraising all hospital documentation, including inpatient admissions, emergency room visitations, outpatient clinical evaluations, and phone calls recorded along with operative reports following initial gastrostomy and before initial gastrostomy change in the surgical clinic.
Wound complications of interest were cellulitis, abscess formation, pressure-related erosion or necrosis, internal and external leakage, granulation tissue, and dislodgement rates. When reviewing documentation, cellulitis was defined as erythematous skin findings requiring a course of antibiotics to improve. An abscess was a collection of purulent material surrounding the LG that required drainage and was not able to be managed with antibiotics alone. Pressure-related erosion was noted as ulcerative findings resulting in local breakdown requiring adjunct local wound care. External leakage was leakage occurring around the tube, including gastric contents and/or feeds. Internal leakage was defined as leakage around the tube within the peritoneal space as defined by fluoroscopic gastrostomy tube. Dislodgement included any accidental and unintentional removal of the gastrostomy button.
Descriptive statistics were reported as mean ± standard deviation unless otherwise stated. Categorical data as reported were compared using a Fisher's exact test and continuous variables with a Student t-test or Mann–Whitney U, where appropriate.
Results
Demographics
Seven hundred forty patients underwent an LG during the 5-year study period. Three hundred seventy-four (51%) patients had an inpatient surgery and one-fifth of patients had a concomitant fundoplication. Four hundred twenty-one (57%) patients were male with an overall median age of 9 months at operation (interquartile range [IQR]: 3 month–2.5 years). At surgery, the median weight was 6.8 kg (IQR: 4.3–11.6 kg).
Five hundred fifty-four (75%) patients had a TA stitch and 186 (25%) had an SC stitch. In evaluating operative reports, 635 dictations had suture details described. For the TA cohort, 361 patients had a PDS and 103 had a Prolene. In the SC group, 23 patients had a Monocryl, and 148 had a Vicryl. The most common gastrostomy tube used was a 12 French 0.8 cm low-profile MIC-KEY button accounting for 57% of patients.
There were no differences in patient gender, age at operation, and weight at operation between both groups.
Operative data
The overall median operative time was 29 minutes (IQR: 18–57). Median operative time was significantly longer in the SC group (22 minutes versus 28 minutes, P = .002). Intraoperatively, there were 4 patients who had a gastric injury requiring primary repair; all of these patients had a TA stitch. These patients, after primary repair, had two monofilament TA sutures used to secure their new gastrostomy button. One patient returned to the operating room within a 24-h period for persistent bleeding around the gastrostomy tube. One patient had an intraoperative cardiopulmonary arrest upon induction of anesthesia with return of spontaneous circulation (ROSC) with chest compressions and another patient suffered postoperative cardiopulmonary arrest event with ROSC after compressions.
Outcomes
Nine patients died from pathologies unrelated to their LG and before their 2-month follow-up. Nine patients either moved out of state or were lost to follow-up. These 18 patients were excluded from further analysis.
Of the remaining 722 patients, 538 patients remained in the TA cohort and 184 patients in the SC group. Nearly half of all patients had a wound complication. Forty-six percent of TA stitch patients developed a wound complication and 53% of SC patients. Regardless of the technique, the most common overall wound complication was granulation tissue (22.5%). This rate was followed by dislodgement (19%), external drainage (16%), cellulitis (10%), erosion (3%), and abscess formation (2%). There were no deaths related to gastrostomy-related complications. The mean follow-up for initial gastrostomy change was 2 ± 0.7 months for all patients.
Table 1 summarizes the comparison of wound complications between both techniques. There were increased infectious complications with 2.5 times more cellulitis in the SC cohort. Additionally, abscesses requiring drainage were observed at a higher rate in SC patients. TA patients seemed to have earlier findings of cellulitis and abscess with a trend toward significance in the rate of abscess formation. All other complications, including granulation, external and internal leakage, dislodgement rates, and erosion were comparable between these patients.
Values shown in bold denote statisically significant values as defined in the methods.
—, n too small for comparative analysis.
IQR, interquartile range; SC, subcutaneous; TA, transabdominal.
Dislodgement occurred in 1 out of every 5 patients. Despite an equal number of inpatient and outpatient LG, a higher rate of dislodgements occurred outside of the hospital, accounting for 78% of dislodgements. Family members were able to replace the gastrostomy tube at home 28% of the time. The majority (60%) of patients had successful replacement of gastrostomy tube at bedside in the urgent care or emergency department setting. Operative replacement was requiring in only 7 patients; 5 of these patients were noted to have contrast extravasation consistent with internal leakage after attempted bedside gastrostomy replacement. There was a higher preponderance of TA patients (6 patients) who required gastrostomy revision in this setting. Only 1 patient in the SC group required operative gastrostomy replacement.
After 2 months of initial follow-up, 450 patients had continued use of their gastrostomy button at the end date of our chart review. Forty-two patients died during the chart review period for nongastrostomy-related etiologies. Seventy-eight patients either moved out of the state or were lost to follow-up. The remaining 152 patients had their gastrostomy button removed at a median of 14 months after initial placement (IQR: 9–23 months). Of these, 43 (28%) required surgical closure at a median of 3 months (IQR:1–6 months). Persistent gastrocutaneous fistula occurred at an equal rate between groups: 32 (6%) patients in the TA cohort and 11 (6%) patients in the SC group (P = 1).
Discussion
Classically, a TA U-stitch has been a method for securing an LG tube.2,3 These temporary sutures were removed at various times due to concerns for erosion. Concern over possible dislodgement led some surgeons to tunnel the stitch and bury the knot to provide security while the gastrostomy is healing.4,7 This study has the largest dataset to date comparing these two techniques to highlight difference in wound complication rates.
Gastrostomy dislodgement can be catastrophic. Removal of temporary sutures used in the TA method raises the concern of inadequate healing of the stomach to the anterior abdominal wall. One series reported a rate of 2.6% of TA U-stitch patients requiring operative revision due to gastrostomy dislodgement. 3 Our data revealed that the incidence of dislodgement was quite common with most patients tolerating replacement at bedside without reoperation. Patients with internal leakage made up the majority of patients requiring operative revision. While there was a slightly higher rate of dislodgement in TA patients, this was not significantly different from SC patients. Additionally, TA patients comprised most patients who ultimately required operative replacement of their gastrostomy. Regardless, the likelihood of a true operative emergency resulting from dislodgement was low in both cohorts.
More significant was the finding of increased infectious complications of both cellulitis and abscess rates in patients with an SC stitch. We found that nearly 1 out of every 5 patients with a permanent stitch developed cellulitis requiring treatment with antibiotics, despite a <1% of drainable abscess rate. These rates, particularly for cellulitis, were higher than what has been reported in the literature.3,5,7 This finding was likely due to inclusion of all hospital documentation with a particularly high amount of data coming from nonsurgical services with more liberal antibiotic use. Additionally, the SC stitch patients had significantly earlier times to external drainage and slightly higher rates of external drainage around the tube that might account for the physiology of wound breakdown and possible infection. Across both cohorts, we found that the majority of patients who developed erosion had preceding external drainage. Interestingly, there were earlier times to development of infectious findings in the TA cohort despite these patients having an overall lower rate of actual incidence of infection. It is difficult to know what criteria were truly used to qualify these complications in documentation.
Even with definitions to help delineate wound complications, data were still collected retrospectively and as such, inherently limited. The surgical service plays a heavy clinical role in initial consultation, with gastrostomy teaching and in the event of emergent gastrostomy replacements. For complications that are nonoperative in nature, such as minor leakage, irritation, and surgical site infections, there is more limited surgical documentation. As such, clinical decisions and definitions may be variable and are broadly defined by a heterogeneous population of providers. Evaluating these wound complications prospectively with multispecialty involvement to devise better diagnostic and treatment guidelines would result in more consistent and reliable data.
Along with being a commonly performed procedure, complications from pediatric gastrostomy have a significant impact on healthcare burden.8–10 These costs are often not related to major medical emergencies but rather from minor mechanical or infectious complications. 8 Our study found increased infectious complications and operative times in those patients who underwent an SC stitch. Additionally, we did not identify significantly higher rates of dislodgement and erosion in TA patients. Understanding modifiable factors that may help minimize these complications are vital steps to improve outcomes and overall healthcare costs.
Footnotes
Authors' Contributions
A.S.P. acquired data, performed data analysis, drafted initial article, critically reviewed and revised the final article, and approved it in its final form. S.R., A.M., F.F., H.B. assisted with acquiring data. R.S. performed data analysis and critically reviewed and revised the final article, and approved it in its final form. R.M.R., K.L.W., and R.J.H. critically reviewed and revised the final article and approved it in its final form. S.D.S. conceptualized the study, supervised data acquisition and analysis, critically reviewed and revised the article, and approved it in its final form.
Disclosure Statement
No competing financial interests exist.