Since the introduction of the laparoscopic technique for gastrostomy tube placement in the 1990s, there have been comparisons made between percutaneous and laparoscopic techniques. Unfortunately, these studies have suffered from heterogeneity in the principles of the PEG and LapGT methods as well as in the materials used for each. Most recent comparison studies describe PEGs placed using the pull or push technique, leaving gastrostomy tubes with semi-rigid internal retention bumpers and no fixation of the stomach to the abdominal wall. These aspects differ greatly from the laparoscopic technique and may account for the increased complications noted with PEGs in these studies [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]. At our institution, both PEGs and LapGTs employ similar mechanics, using gastrostomy tubes with internal retention balloons, inserted by way of Seldinger technique and stabilized with temporary fixation of the stomach to the abdominal wall. These parallel placement techniques allow more fundamental comparison of the endoscopic versus laparoscopic methods.
The push and pull PEG techniques require pulling wires and tubes back and forth across the abdominal wall, stomach and esophagus, an action that can be traumatic to tissue, especially in the neonatal population or patients with head and neck pathologies. In our PEG technique, the MIC gastrostomy tube is placed directly into the stomach via a percutaneous sheath, which prevents unnecessary trauma.
The semi-rigid retention bumpers used in the pull and push technique are also associated with an increased risk of buried bumper syndrome [14]. This is a serious complication in which the internal retention device of the gastrostomy tube migrates through the gastrostomy tract and outside of the stomach. This complication often requires surgical intervention in the OR and may help account for the increase in major complications noted with the PEG technique in earlier studies.
The transcutaneous gastric fixating sutures in our PEG and LapGT techniques give time for the stomach to adhere to abdominal wall and for the tract to form and may prevent dislodgements and other major complications [15, 16].
As in our study, Koh et al. found the initial PEG procedure to be a quicker operation than LapGTs [10]. Yet, with the pull technique used in their study, patients in the PEG group returned to the OR for tube changes leading to no difference in utilization of the operating room between the two groups. By using MIC gastrostomy tubes that have an inflatable internal retention balloon in both our PEGs and LapGTs we find common ground between the two techniques. Exchanges of this type of gastrostomy tube are done exclusively at the bedside or in the office, which greatly reduces the time and cost associated with gastrostomy tube exchange and thus preserves the savings in operative time that we see in the PEG technique at our institution.
When comparing different surgical techniques for the same operation, one must consider which adds more value. The value of a surgical procedure can be influenced in two ways, quality (i.e., complications) and cost. Our study saw zero major complications after 30 days in both the PEG (0/56) and LapGT (0/37) groups, a stark difference not readily seen in previous studies and meta-analyses where LapGTs were often found to be safer. We postulate that this may be due, in part, to the differences in PEG technique applied in this study as compared to the technique employed in previous analyses.
Overall operative charges were significantly less for the PEG group compared to the LapGT group, with an average savings of $4,500 per patient (p < 0.001). This reported monetary benefit is significantly underestimated, as the private-practice anesthesia billing was not included in the savings calculation. Both procedure time and total time spent in the OR were significantly less in the PEG group, a point seen in some previous studies [4, 5, 6, 7, 8, 9, 11]. Since OR and anesthesia services are billed in half-hour increments, mean OR times become important, in terms of cost-effectiveness, when they cross 30 min thresholds. By keeping average OR time for PEGs under an hour, considerable cost savings are realized when compared to LapGTs that have an average OR time of 85 min. Shorter OR times add value to the patient by saving in anesthesia and operative charges and by minimizing intubation and anesthesia time in this fragile cohort. They additionally benefit the hospital and anesthesia practice by improving OR efficiency.
If safety criteria were not met during upper endoscopy for a planned PEG, then a LapGT was performed. We do not see this as a failure of PEG, but rather proof that the safety criteria helped prevent misplaced gastrostomy tubes during the study period. In our report, two planned PEGs were changed to the LapGT, technique, representing a conversion rate of 3.5%. Although low, the possibility of PEG failure makes us recommend the presence of laparoscopy equipment readily available, but unopened, in the OR suite to minimize OR time if change to the LapGT technique is required.
Although the quantity of time spent on the endoscopic and laparoscopic portions of the two converted cases was not recorded, we can gather a sense of how time was used from the operative notes. For the patient in whom the area of transillumination was above the costal margin, procedure time and OR time were 41 min and 80 min, similar to the average LapGT times. In the patient with a distended colon who required an additional trocar and maneuvers to identify the stomach, procedure and OR times were longer (147 min and 186 min, respectively). We hypothesize that outlier operative times for this child were secondary to the difficult anatomy and would have occurred with or without the addition of the endoscopy at the start of the case.
This study is limited by the potential effects of confounding and bias inherent to retrospective study designs. The smaller sample size characteristic of a single center study may miss small differences between two groups. Additionally, the decision to proceed with PEG or LapGT was based on surgeon preference. This preference is no doubt influenced by the surgeon’s skill and experience with their technique of choice. Complications were assessed and reported in the 30-day postoperative period, and therefore lack long term follow up data. We also recognize that surgeon comfort with upper endoscopy is not uniform across institutions. It should be noted that pediatric surgeons at our institution perform their own upper endoscopies, and as such did not require intraoperative assistance from gastroenterology colleagues. Therefore, we did not analyze how the addition of a gastroenterologist to the PEG procedure would change operative cost or outcomes.