One of the more invasive endoscopic procedures is the percutaneous endoscopic placement of a gastrostomy tube. This procedure was originally described in pediatric patients (1) but was quickly adapted to all age groups. The procedure is safe and without the longer recovery associated with surgical placement. The sooner the feedings are started, the sooner the education of the patient's caregivers can be initiated. In some adult centers, it is reported that the procedure can be performed in the morning and the patient discharged the same day (2). In pediatric patients, a short stay for teaching gastrostomy use and care may be beneficial for the child's eventual care at home. The standard followed by the Pediatric Gastroenterology Division at Riley Hospital was to initiate feedings at 6 hours following percutaneous endoscopic gastrostomy (PEG) placement. This waiting period is based on a clinical report by Werlin et al (3). This study was uncontrolled, but the authors reported no unusual complications that led to feeding discontinuation in 24 pediatric patients taking Pedialyte (Abbott Nutrition, Columbus, OH) 6 hours after PEG placement (3). The aim of our study was to determine whether earlier initiation of feeds is safe and comparable to the current standard practice.
PATIENTS AND METHODS
All of the patients referred to the Pediatric Gastroenterology Division for PEG placement were considered for inclusion in the study regardless of age, race, or sex. The standard protocol of the procedure was unaltered. All of the children received primary gastrostomy button placement by the standard “pull” technique (1). Patients with known anatomic abnormalities or gastrointestinal dysmotility were excluded.
Pediatric patients requiring a PEG placement were randomly assigned to feedings at 3 and 6 hours after placement. Randomization was performed using a random number table before the procedure. All of the patients received a dose of intravenous antibiotic before the placement and 2 doses afterward. It was not possible to blind the nursing staff to the feeding time assigned to each patient. The first feeding was with Pedialyte (Abbott Nutrition) at a volume of 60 mL by bolus.
Any postoperative vomiting was recorded, and the abdominal girth was measured immediately before the first feeding and 1 hour after the feeding was started. The residual volume in the stomach before the second feeding 3 hours later was also recorded. The second feeding was the standard formula prescribed for the patient. Patients who could take oral solids safely would be allowed to eat after the typical 6-hour bolus was tolerated. Measurable regurgitation or a fluid volume in the stomach greater than the previous bolus before the next feeding was considered evidence of feeding intolerance. The length of stay was recorded for patients who underwent the procedure as scheduled outpatients. Inpatients who required a PEG were enrolled in the study but excluded from the length-of-stay data.
Before the initiation of the study, a power calculation indicated that 20 patients in each group would give an 80% power to detect a difference of 10%, assuming a deviation of 10% with α set to 0.05. The data were compared with a standard t test or the Mann-Whitney rank sum test, if not normally distributed, using SigmaStat software (SSPS Inc, Chicago, IL). The results are presented as median values and then mean ± standard error of the mean.
The present study was approved by the Indiana University/Purdue University institutional review board. Consent was obtained from the patient's caregivers before enrollment in the study.
There were no statistical differences in the patient populations randomly assigned to the 3- or 6-hour feeding times. The age range was the same for both groups (median 3 hours for 1.48 years vs 1.49 years for 6 hours, mean 3.5 ± 0.9 years for 3 hours vs 2.8 ± 0.6 years for 6 hours; P = 0.76) as was the gender division (3 hours for 15/20 males, 6 hours for 14/20 males, P = 1.0). The most common diagnosis for patients undergoing PEG placement was failure to thrive or feeding problems; there was a variety of other diagnoses that are listed in Table 1. No infectious or other placement complications occurred in either group.
The change in girth data demonstrated no statistical differences between the 3- and 6-hour groups (median 0 cm for 3 hours vs 0.5 cm for 6 hours, the mean was 0.08 ± 0.15 cm for 3 hours vs 0.27 ± 0.19 cm for 6 hours; P = 0.29). The number of emesis events recorded was not statistically significant between the 2 groups (median for both groups was 0, the mean was 0.53 ± 0.17 for 3 hours and 0.68 ± 0.3 for 6 hours; P = 0.82). Gastric residuals between the 2 groups were not statistically different. The majority of the patients had no residuals and the ones that did only had several milliliters, therefore, the median gastric residual for both groups was 0 (mean 7.0 ± 5.1 mL for 3 hours vs 0.06 ± 0.06 mL for 6 hours; P = 0.47). There were no recorded events that would indicate an increased incidence of infection, including fever or site erythema, in either study group.
A post hoc analysis of the data for PEGs performed in children younger than 1 year of age revealed no statistical differences in the outcome data; however, the number of patients was small (3 hours n = 5, 6 hours n = 4) and possibly inadequate to demonstrate an actual statistical difference.
The majority of the study patients were scheduled outpatients; however, some were inpatients hospitalized for other reasons. These inpatients were excluded from the length-of-stay data. The length of stay was not significantly different between the 2 groups (n = 14 for the 3-hour group with a median of 31 hours of observation and n = 19 for the 6-hour group with a median of 30 hours of observation; P = 0.92).
The PEG technique for gastrostomy placement was developed as an alternative to surgical placement with the advantage of a faster recovery. The rationale for waiting to feed is a long-standing vestige of the belief that anesthesia suppresses motility. This dogma has been challenged in the surgical literature in recent years. A study in adults undergoing elective gastrointestinal surgery found that patients tolerated a regular diet 8 hours postoperatively (4). A study of newborns requiring gastrointestinal surgery who were given trophic feedings within 12 hours of surgery found earlier stool output and shorter hospital stays (5). These newborns were started without waiting for flatus or stool passage.
Common complications of PEG placement should not be affected by timing of the feedings because the 3-hour difference would not allow resolution of any potential complications. There have been several studies of early feeding after PEG placement in adults. Two studies with >20 patients in each group found no difference in complication rates in patients fed 3 hours after placement compared with those fed the following day (2,6). Two studies in adults infused water-soluble contrast into PEGs 3 hours after placement and found no leakage on an x-ray of the abdomen (7,8). One of these studies was of primary gastrostomy button placement (8). There has been 1 more rigorous study in adults that randomized patients to receive formula feedings at 4 hours versus 24 hours after PEG placement (9). That study had more than 50 patients in each group and measured postfeeding residuals. No differences were noted in the incidence of complications or in the volume of gastric residuals. In an older review article, Kirby et al (10) reviewed their first 55 PEGs in adults and stated that they began a trial infusion of water at 2 hours after the procedure without increased complications. A recent Spanish article with an English abstract reports a randomized trial of immediate feeding in more than 30 adult patients with no increase in complications (11).
The study with earliest feeding in children after PEG placement (6 hours) is by Werlin et al (3). This was an uncontrolled series of 24 pediatric patients, and they found no increase in the complication rate with the earlier feedings (3). Our procedures are performed under general anesthesia and afterward patients recover in a general postanesthesia care area. This area is not designed to allow anyone to perform feedings or initiate them. Arrival on the short-stay unit, approximately 3 hours later, is the first practical time that feedings can be initiated.
Because there were no changes in complication rate with the 3-hour feeding, it did not increase the overall cost of the procedure. The length-of-stay data did not demonstrate any change, which was expected because this procedure is normally performed with the expectation that the postoperative care will be brief. When Werlin et al (3) fed their patients at 6 hours they were able to discharge the patients in 24 hours, which was when feedings had been initiated in the past. The present study's findings have the potential to result in shortening the length of stay even further. The practice has been to perform the PEG and bring the patient into the hospital for a 23-hour observation stay. During this short stay the patient completes antibiotics and caregivers are taught the care and use of the gastrostomy. The present study raises the possibility that with appropriate logistics, patients could complete their recovery, and the training takes place without an overnight stay. This would result in reducing the costs of the procedure.
In summary, we found that pediatric patients randomly assigned to use a PEG for feedings at 3 hours after placement did so with excellent tolerance. There was no increase in complications with a shorter recovery time before feedings were started. This change in practice should lead to earlier initiation of feedings and foster increased caregiver comfort with feeds before the discharge home.
1. Gauderer MW, Ponsky JL, Izant RJ. Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 1980; 15:872–875.
2. Brown DN, Miedema BW, King PD, et al
. Safety of early feeding after percutaneous endoscopic gastrostomy. J Clin Gastroenterol 1995; 21:330–331.
3. Werlin SL, Glicklich M, Cohen RD. Early feeding after percutaneous endoscopic gastrostomy is safe in children. Gastrointest Endosc 1994; 40:692–693.
4. Lucha PA Jr, Butler R, Plichta J, et al
. The economic impact of early enteral feeding in gastrointestinal surgery: a prospective survey of 51 consecutive patients. Am Surg 2005; 71:187–190.
5. Ekingen G, Ceran C, Guvenc BH, et al
. Early enteral feeding in newborn surgical patients. Nutrition 2005; 21:142–146.
6. Choudhry U, Barde CJ, Markert R, et al
. Percutaneous endoscopic gastrostomy: a randomized prospective comparison of early and delayed feeding. Gastrointest Endosc 1996; 44:164–167.
7. Nolan TF, Callon R, Choudry U, et al
. Same day use of percutaneous endoscopic gastrostomy tubes: radiographic evidence of safety [abstract]. Am J Gastroenterol 1994; 89:1743.
8. Yarze JC, Herlihy KJ, Fritz HP, et al
. Prospective trial evaluating early initiation of feeding in patients with newly placed one-step button gastrostomy devices. Dig Dis Sci 2001; 46:854–858.
9. McCarter TL, Condon SC, Aguilar RC, et al
. Randomized prospective trial of early versus delayed feeding after percutaneous endoscopic gastrostomy placement. Am J Gastroenterol 1998; 93:419–421.
10. Kirby DF, Craig RM, Tsang T-K, et al
. Percutaneous endoscopic gastrostomies: a prospective evaluation and review of the literature. J Parenter Enteral Nutr 1986; 10:155–159.
11. Carmona-Sanchez R, Navarro-Cano G. Percutaneous endoscopic gastrostomy. Is it safe to start eating immediately? Rev Gastroenterol Mex 2002; 67:6–10.