A Descriptive Study of Complications of Gastrostomy Tubes
Elizabeth Goldberg RN, MSN, CPNP⁎, Sharon Barton APRN-BC, PhD,
Melissa S. Xanthopoulos PhD, Nicolas Stettler MD, MSCE, Chris A. Liacouras MD
Division of Gastroenterology and Nutrition, The Children's Hospital of Philadelphia, Philadelphia, PA
Objectives: The purpose of this study was to determine the number and types of complications
experienced by children with gastrostomy tubes.
Methods: This is a prospective study of children with gastrostomy tube complications. Enrollment
occurred on the first 24 months of the study. Data were collected for 4 years, beginning at the enrollment
of the first participant. Demographic data and information on infections, granulation tissue formation,
and major complications were recorded.
Results: Infections occurred in 37% of patients, with most experiencing a single infection that occurred
within the first 15 days after tube placement. Granulation tissue developed in 68% of patients, with 17%
experiencing recurrent granulation tissue despite treatment. There was no difference in infection rates or
granulation tissue formation between subgroups based on gender, ethnicity, or parents' education level.
Major complications occurred in 4% of the patients.
Conclusion: Complications of infection and granulation tissue occur frequently and likely are a cause of
stress and increased burden of care for these children and families. Improved strategies for care are
© 2010 Elsevier Inc. All rights reserved.
GASTROSTOMY TUBE PLACEMENT is a long-term
solution for supporting nutrition in children. There are many
reasons for children to have difficulty consuming adequate
oral nutrition. With a reported prevalence of 25% in all
children and 80% in children with developmental disabil-
ities, impaired feeding is very common (Manikam &
Perman, 2000). Severe feeding problems are noted in 3%
to 10% of children and occur with greater prevalence in
children with physical disabilities, medical illness, or a
history of prematurity (Manikam & Perman, 2000).
Etiologies for feeding problems are many and include
medical, nutritional, behavioral, psychological, and environ-
Enteral tube feedings and parental nutrition are two
methods of nutrition support for severe cases of impaired
feeding. As many as 11,000 gastrostomy tubes are placed in
children younger than 18 years old in U.S. hospitals every
year (Healthcare Cost and Utilization Project, 2006). The use
of gastrostomy tube feedings or gastrojejunostomy tube
feedings has become an accepted method of long-term
nutritional support in children who are not able to consume
adequate calories by mouth yet have a functioning gut. Many
of the children who require gastrostomy tube feedings have
complex medical conditions such as neuromuscular dis-
orders, congenital heart disease, severe gastroesophageal
reflux, failure to thrive, or genetic/metabolic syndromes.
Their medical conditions often require a significant degree of
family involvement in daily care and frequent visits to
pediatric centers for medical care.
Psychological studies on families of children with
gastrostomy tube feedings reported high levels of burnout
and stress. One study found extremely high total stress scores
in the parents of enterally fed children compared with the
total stress scores of parents with healthy children, parents of
children with diabetes mellitus, and parents of children with
⁎Corresponding author: Elizabeth Goldberg, RN, MSN, CPNP.
E-mail address: email@example.com (E. Goldberg).
0882-5963/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
Journal of Pediatric Nursing (2010) 25, 72–80
growth deficiencies (Pedersen, Parsons, & Dewey, 2004).
One factor thought to contribute to the high stress scores of
parents with enterally fed children is the constant demand of
coordinating the day-to-day care of their child and the
ongoing effort and time involved in providing daily tube
feedings. Feeding is an important task of parenting.
Therefore, the failure of oral feeding and the subsequent
need for tube feeding may raise concerns about being a
successful parent. Feelings of guilt over the need for a
feeding tube and the development of tube-related problems
that further complicate the child's care may also contribute to
parental stress and self-doubt. In a study by Guerriere,
McKeever, Llewellyn-Thomas, and Berall (2003), mothers
were interviewed at the time of gastrostomy tube placement
in their children. The majority reported a lack of information
regarding gastrostomy tube use and complications associated
with enteral feedings. Parental lack of knowledge as to the
causes or treatments for various feeding tube problems may
result in parental fear. They often fear that the tube may
malfunction and deny their child access to fluids, nutrition,
or medications. The pain associated with a poorly fitting
tube, skin excoriation, granulation tissue formation, or
infection adds to parental stress and worry. In addition,
parental perception that many health care providers lack
experience in caring for enteral devices can leave a parent
with a feeling of little support.
Few studies have investigated complications associated
with gastrostomy tube use in children. A survey by Crosby
and Duerksen (2005) included 55 adults and children who
had been fed with a gastrostomy tube for a mean duration of
25.9 months. Common complications included granulation
tissue formation (67%), broken or leaking tube (56%),
leakage from around the tube site (60%), and stomal
infection requiring antibiotics (45%). Another study (Fried-
man, Ahmed, Connolly, Chait, & Mahant, 2004) evaluated
complications associated with radiologically placed gastro-
stomy and gastrojejunostomy tubes in children. In a chart
review of 208 children aged 7 days to 18 years, researchers
reported a major complication rate of 5% and a minor
complication rate of 73%. Major complications included
subcutaneous abscess, peritonitis, septicemia, gastrointesti-
nal bleeding, and death. Minor complications were tube
dislodgement, tube leakage, and gastrostomy site skin
infections. It has been estimated that gastrostomy site skin
infections occur in one fourth to more than one third of all
patients (Crosby & Duerksen, 2005; Gossner, Keymling,
Hahn, & Eli, 1999).
There have been few studies on infection or granulation
tissue formation in children with gastrostomy tubes. These
are complications that can be anticipated and hopefully
prevented. Information on incidence and natural history of
these complications can help guide clinical practice to
reduce complications and provide safe and effective care to
children and their families. A greater understanding of
gastrostomy tube complications may result in improved
anticipatory guidance for patients and families and a
lessening of stress associated with the use of these devices.
The purpose of this study was to determine the number and
types of complications experienced by children with
gastrostomy tube placement.
Materials and Methods
The division of gastroenterology at a tertiary children's
hospital was the site for the study. The institutional review
board approved the study protocol to assess gastrostomy tube
complications by prospective data collection and retrospec-
tive review of medical records. All patients who presented
for tube placement or tube-related care between July 1, 2001,
and July 30, 2005, were asked to participate in this study.
Enrollment occurred throughout the first 24 months of the
study. Data were collected for 4 years from the enrollment of
the first participant. Patients/Families were approached for
consent by the principal investigator (PI), an experienced
gastroenterology pediatric nurse practitioner. The families
reported their demographic information and date and method
of tube insertion, diagnosis, reason for tube placement,
parental education level, and family income (Table A in
Using a data collection form (Table B in Appendix A)
developed by the PI, information on complications involving
infections, granulation tissue formation, and major compli-
cations requiring hospitalization or surgical intervention was
recorded at every visit or follow-up telephone call. Infection
involving the G-tube site was defined as having three of the
following symptoms: spreading erythema of the peristomal
skin, increased tenderness or pain, induration, fever,
development of a furuncle, or purulent discharge. Recurrent
infections were defined as having two or more infections,
with the repeat infection occurring at least 2 weeks after the
completion of therapy for a prior infection. Granulation
tissue was defined as a proliferation of capillaries that
manifest clinically as exuberant, red, raw, beefy, painful, or
bleeding tissue extruding from the inside of the stoma.
Treatment outcomes for these complications were recorded.
Every study participant was contacted either by telephone
or in a clinic every 4months. Patient assessments, reported or
observed complications, treatments, treatment outcomes, and
patient/family education were performed by the PI. If the
family or patient expressed concerns regarding the condition
of the tube or stoma site, the patient was evaluated by the
investigator. If the family lived over an hour drive from the
hospital, the patient was seen by the primary care physician
in the community or if after office hours at the emergency
department. Information concerning clinical assessments and
treatments of gastrostomy tube complications not seen by the
PI was collected by either a follow-up telephone call to the
family and/or a retrospective review of the medical record.
Electronic medical records were also utilized when available.
Medical record data were used to verify or expand upon
73 A descriptive study of complications of gastrostomy tubes in children
patient/parent reports. Retrospective chart reviews involved
30% of the sample. Data extraction was not verified by
another researcher. Evidence for infection and granulation
tissue was determined using standardized definitions and
prescribed therapy (e.g., the use of an antibiotic in the
absence of any other source of infection). The data were
coded and placed into an electronic database. Patient
information was protected by assigning every participant a
personal identity number. The database was stored on a
secure computer that was password protected. Upon
completion of data collection, the data were deidentified
prior to analyses. The statistical analyses for categorical data
used frequencies and percentages to describe the sample.
Inferential statistics including chi-square were used to
determine if the type and rate of complications were affected
by gender, ethnicity, age, or parents' education level.
During the 2-year enrollment period, a convenience
sample of 94 patients were enrolled into the study. Ongoing
assessments were collected on 70 patients until the end of the
study. Of the remainder, 10 of the patients died during the
no longer required a G tube. Demographics of the sample are
2 months to 26 years, with 49 females and 45 males.
Thirty of the study participants were enrolled at the time of
tube placement in the interventional radiology unit. The
for 1 to 143 months prior to inclusion in the study; 55 tubes
by surgery. None of the tubes were placed endoscopically.
Most patients were Caucasian (73%), African American
(20%), Hispanic (4%), and Asian (3%). The reasons for
enteral tube placement included gastrointestinal problems
(41%; such as failure to thrive, inflammatory bowel disease,
severe food allergies, etc.), neurological or neuromuscular
disorders (28%), pulmonary disease (7%; mainly cystic
fibrosis), and cardiac disorders (4%), with the remainder
having metabolic disorders, immunological disease, or
cancer. In our sample, 37% of the families had a reported
annual income of greater than $80,000, with 33% reporting a
yearly income of less than $40,000.
Gastrostomy-related infections occurred in 35 (37%) of
the 94 patients. Twenty-one (60%) of the 35 patients
experienced a single infection. In 12 (34%) of these 35
patients, the first infection developed within 15 days of initial
tube placement, despite the use of prophylactic systemic
antibiotics in 10 of the patients. Recurrent infections were
noted in 14 (40%) of the 35 patients. Two of the 14 patients
with recurrent infections required surgical closure of a
gastric–cutaneous fistula that had become chronically
infected despite repeated use of local and/or systemic
antibiotics. Prior to surgery, both patients had been treated
with targeted antibiotics determined by wound cultures.
Most of the single infections were treated with an oral
antibiotic such as cephalexin, topical mupirocin, or a
combination of oral and topical therapy. In the last year of
the study, the use of clindamycin as the first-choice antibiotic
increased with the rising prevalence of methicillin-resistant
Staphylococcus aureus. All 14 children with recurrent
Demographic Characteristics of the Sample
Age at time of G-tube placement (years)
Annual household income ($)
74 E. Goldberg et al.
infections were treated with oral antibiotics, and 11 (79%)
required intravenous antibiotics. There were no statistically
significant differences in infection rates between subgroups
based on gender, ethnicity, or parental education level. In
children younger than 1 year old, 27% had at least one
infection. Fifty-two percent of the children aged 1 to 5 years
experienced at least one infection, and 20% of children older
than 5 years old had at least one infection (Figure 1). The
incidence of infection in children aged 1 to 5 years was
statistically significant (p b .05).
Sixty-four (68%) of the 94 study patients experienced
gastrostomy-tube-induced granulation tissue. In the sub-
group of patients followed since the time of initial tube
placement, 25 (72%) of 35 patients experienced granulation
tissue, with most developing granulation tissue within 3
months of tube placement (Figure 2). Recurrence of
granulation tissue occurred in 11 (17%) of the 64 patients
with granulation tissue and 12% of the overall sample. In
these 11 patients, therapy which included the use of silver
nitrate, steroid cream (triamcinolone 0.1%), stomadhesive
powder, or topical antibiotics either did not result in
complete resolution or the granulation tissue returned
immediately upon completion of the therapy. In addition,
all 11 patients continued to have evidence of granulation
tissue throughout the 4 years of the study. In 1 patient,
granulation tissue was a problem for over 10 years until she
underwent revision of the gastrostomy site in surgery. In 53
out of the 64 patients, resolution of granulation tissue
occurred within 2 years after G-tube placement. Most of
these patients had intermittent granulation tissue formation
several times during the first 2 years after tube placement.
Granulation tissue was seen as early as 7 days after initial
tube placement. There was no difference in the incidence of
granulation tissue formation between subgroups based on
gender, ethnicity, or parents' education level. Fifty-one
percent of children younger than 1 year old had granulation
tissue. Eight-one percent of children between 1 and 5 years
old had granulation, and 73% of children older than 5 years
old had granulation tissue (Figure 1). The differences
between groups were not statistically significant.
Four of the patients in this study required surgery for a
gastrostomy-tube-related complications. Two of the 4
patients required surgery for closure of a chronically infected
fistula; 1 patient had chronic leakage from the stoma, and 1
reoccurring granulation tissue. One patient experienced an
extended hospital stay after radiological placement of the G
tube from the “buried bumper,” a condition that occurs when
too much tension upon the tube causes erosion of the gastric
mucosa from the internal disc of the tube. Eleven of the 14
patients with recurrent infections remained in the hospital for
IV antibiotics secondary to infection. None of the 10 patients
who died during the study died of tube-related causes.
Other Minor Complications
One patient had a latex allergy to a red rubber catheter that
was used to access her stoma. The redness, tenderness, and
swelling of her site initially thought of as infection cleared
upon removal of the latex catheter. Three patients experi-
enced recurrent breakage of the internal balloon, and two of
these patients were fitted with gastrostomy tubes that had a
mushroom-style internal bumper. One patient developed a
false, dead-ended track that did not connect into the stomach.
To successfully place a G tube into his stomach, fluoroscopic
guidance to avoid the false track was necessary.
Advances in medical care and improvements in nutrition
will likely continue to increase survival and longevity in
children with complex health issues. Overall, it is reasonable
Granulation and infection rates.
(n = 25).
First granulation tissue after initial tube placement
75 A descriptive study of complications of gastrostomy tubes in children
to expect that the number of children who rely upon enteral
feedings will continue to increase. Although complications
associated with the use of enteral feedings often occur, life-
threatening complications are few. Complications, such as
infection and granulation tissue formation, frequently occur
and not only add to the burden of care but also increase the
economic costs associated with enteral feeding. Although
most complications (granulation tissue, leaking or dislodged
tubes, equipment malfunctions, or minor skin infections) are
considered of little clinical importance to health care
providers, these complications frequently cause concern
and stress for patients and families.
Infections involving the G-tube site are reported to occur
in 25% to 45% of all patients (Crosby & Duerksen, 2005;
Gossner et al., 1999).
In this study, 37% developed an infection, with most
experiencing a single infection that occurred within 15 days
of tube placement. Both previous research and this study
demonstrate an elevated incidence of infectious complica-
tions related to gastrostomy tubes. One possible explanation
for this elevated infection rate may be that tertiary pediatric
hospitals care for children with extreme illness whose
complex health conditions make them more vulnerable to
infections. Other factors to consider would be the nutritional
state of the patient prior to tube placement and the
increasing bacterial resistance to antibiotics. The choice of
prophylactic antibiotics should be based upon the evolving
characteristics of bacterial resistance and the antibiotic-
resistant patterns unique to the geographic area. Very few of
the infected sites were cultured for sensitivities. Often,
confusion regarding a diagnosis of infection occurs.
Peristomal redness (which may be a result of friction from
the tube rubbing on the skin) with a yellow mucoid
discharge (which is often a normal occurrence) in a fussy
child may be mistaken for an acute infection, and antibiotics
were prescribed by well-meaning practitioners “just in
case.” Cellulitis in a newly placed gastrostomy tube may be
either bacterial or chemical. Chemical cellulitis may result if
gastric contents leak into the surrounding subcutaneous
tissue and result in an inflammatory response that is often
difficult to differentiate from bacterial cellulitis. Cellulitis
always involves a spreading diffuse erythema that is
associated with varying degrees of swelling and pain
under normal-appearing skin. In this study, over half of
the children between the ages of 1 to 5 years experienced an
infected G-tube site. Confirmation of these results in a larger
prospective study is indicated, as is the need to tract
infections rates in patients from the time of tube placement.
The parent population in this study was well educated and
with high income. Future research including lower income
families would be interesting and may help to delineate the
impact of education and socioeconomic status upon
infection rates and treatment outcomes. It would also be
interesting to investigate in-home practices concerning the
daily care of gastrostomy tube sites, such as cleansing
practices of the site or equipment.
A poor nutritional state may also contribute to post-
procedural infections. One way to manage this possibility
would be to promote adequate nutrition in children prior to
placement of G tubes via temporary nasogastric feedings.
Although gastrostomy tube placement provides ready access
for feeds, the improvement in the child's nutritional state is
dependent upon consistent provision of feedings and
continued monitoring of caloric and nutrient intake.
Inadequate calories and micronutrient deficiencies in tube-
fed children given noninfant formulas have been documen-
ted (Skelton, Havens & Werlin, 2006). Nursing plays an
important role in educating parents about proper growth and
nutrition. A future study including assessments of the
nutritional status of children with recurrent infections or
infections requiring IV antibiotics would be useful.
The exact etiology for the development of granulation
tissue is unclear. Granulation tissue development in 68% of
the patients was in keeping with the study report of 67% by
Crosby and Duerksen (2005) and provides additional
evidence that granulation tissue development involving the
G tube is a very common problem. Granulation tissue was
treated in 70% of our patients with silver nitrate. The
application of silver nitrate is often associated with pain, and
if improperly applied, secondary discoloration or burning of
the surrounding skin may occur. Children are often fearful of
this procedure. Because silver nitrate usually must be applied
more than once, its use is associated with a great deal of
stress for patients, families, and medical caretakers. Alter-
natively, the use of steroid creams in the treatment of
granulation tissue has been utilized; however, their effec-
tiveness has never been demonstrated in clinical trials. In
addition, we found that 17% of our patients experienced
recurrent granulation tissue that never resolved during the 4
years of the study despite the use of both therapies (silver
nitrate and a midpotency steroid cream). For these reasons,
the treatment of granulation tissue should only occur if the
tissue is painful or bleeding and never for solely cosmetic
reasons. Future studies to evaluate or develop new therapies
especially for patients with recalcitrant granulation tissue
formation are needed.
In many hospitals, gastrostomy tubes are placed endo-
scopically or surgically. At this study site, approximately 95
to 100 G tubes are placed by interventional radiology every
year. Most tubes (86%) in this study were placed
radiologically. The nonsurgical placement of a gastrostomy
tube under radiological guidance was introduced in 1981.
The combined use of ultrasound and fluoroscopy allows for
the safe identification of a window into the upper two thirds
of the stomach through which a G tube can be introduced.
This window into the stomach must avoid the liver or
overlying loops of large intestine. Using an antegrade
technique, the G tube is advanced by way of the esophagus
into the stomach and then exits the stomach through a small
incision (or stoma) on the abdomen. There have been studies
that have compared complications in G tubes placed
radiologically, endoscopically, and surgically but only in
76E. Goldberg et al.
adults. This study did not have a sufficient number of
surgically placed tubes to allow for an adequate comparison.
A comparison of complications associated with the three
methods of gastrostomy tube insertion has not been studied
There are few, if any, other prospective studies that have
evaluated G-tube-related complications in either adults or
children. This prospective study of children (only five
participants were older than 18 years old) took place over 4
years, with 25 months as the mean time followed in the
study. Although the number of patients was limited, it
included children of all ages. A strength of the study was that
it provided more than just a snapshot assessment of
complications; instead, it allowed us a broader look into
the lives of children with gastrostomy tubes over time.
Although the number of children followed from the time of
tube insertion was small, a larger prospective study of these
patients could outline in more detail the incidence of
postprocedural infections, the impact of prophylactic antibi-
otic use, and perhaps the etiology of granulation tissue
formation. Another possible limitation would be diagnostic
confusion. Not every participant was seen by the PI when
patients/families had G-tube-related complaints. A limited
assessment of a G-tube site may fail to differentiate
granulation tissue from scar tissue or gastric prolapse and
may not diagnose an infection from the chronic irritation of a
poorly fitted tube. Most participants were followed closely
by the PI but not every child for every complaint. As the
patient population was a convenience sample, the results of
the study may not be generalizable to all children with
gastrostomy tubes. The lack of widely recognized and
validated definitions of gastrostomy tube complications
limits our ability to make comparisons between studies,
medical centers, and standard assessments of therapies. The
development of validated definitions for gastrostomy-tube-
related complications is needed. Because the study was
descriptive and enrolled all families who consented, it was
not possible to choose a balanced number of participants in
each age group.
The use of gastrostomy tube feedings in children is a
widely practiced method of providing long-term nutrition
support. However, enteral feeding is not without complica-
tions, and it is likely that these complications contribute to
stress in patients and their care providers. This prospective
study of mostly children provides some insights into two
complications of gastrostomy tube use: frequent infections
and granulation tissue formation. To improve the care of
children with gastrostomy, standardized and validated defini-
tions of complications will be necessary, and a systematic
prospective follow-up of these patients from the time of tube
placement, perhaps a multicenter registry, is urgently needed.
Improved knowledge and care of children with gastrostomy
will likely result in improved patients' and parents' quality of
this increasing population of patients.
77 A descriptive study of complications of gastrostomy tubes in children
78 E. Goldberg et al.
79 A descriptive study of complications of gastrostomy tubes in children
Crosby, J., & Duerksen, D. (2005). A retrospective survey of tube-related
complications in patients receiving long-term home enteral nutrition.
Digestive Diseases and Sciences, 50, 1712−1717.
Friedman, J. N., Ahmed, S., Connolly, B., Chait, P., & Mahant, S. (2004).
Complications associated with image-guided gastrostomy and gastro-
jejunostomy tubes in children. Pediatrics, 114, 458−461.
Gossner, L., Keymling, J., Hahn, E. G., & Eli, C. (1999). Antibiotic
prophylaxis in percutaneous endoscopic gastrostomy (PEG): A
prospective randomized clinical trial. Endoscopy, 31, 119−124.
Guerriere, D. N., McKeever, P., Llewellyn-Thomas, H., &Berall, G. (2003).
Mother's decisions about gastrostomy tube insertion in children; factors
contributing to uncertainty. Developmental Medicine and Child
Neurology, 45, 470−476.
Healthcare Cost and Utilization Project. (2006). Healthcare cost and
utilization project (HCUP). Retrieved October 2006 from. http://
Manikam, R., & Perman, J. A. (2000). Pediatric feeding disorders. Journal
of Clinical Gastroenterology, 30, 34−46.
Pedersen, S. D., Parsons, H. G., & Dewey, D. (2004). Stress levels
experienced by the parents of enterally fed children. Child: Care, Health
and Development, 30, 507−513.
Skelton, J.A., Havens, P. L., & Werlin, S. L. (2006). Nutrient deficiencies in
tube-fed children. Clinical Pediatrics, 45, 37−41.
80E. Goldberg et al.