Keywords
gastrostomy - G-tube - percutaneous endoscopic gastrostomy - endoscopic - fluoroscopic
Introduction
Gastrostomy tubes have been used for > 100 years in patients requiring enteral nutrition
for an extended period of time.[1] Traditionally, gastrostomy tubes were placed surgically under general anesthesia
in the operating room.[2] Since then, however, new techniques that are simpler to perform and require only
local anesthesia such as percutaneous endoscopic insertion and fluoroscopically guided
insertion have emerged and mostly displaced traditional surgery from clinical practice.[3]
[4]
[5]
[6]
[7] Early studies of both minimally invasive techniques have shown high success rates
and low peri- and postoperative complication rates.[8]
[9]
[10]
[11] However, much of the current literature on fluoroscopic gastrostomy, in particular,
has been limited to being outside of the United States or performed in a single institution
or region.[7]
[8]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19] This study aims to evaluate the complication rates associated fluoroscopically placed
gastrostomy tubes nationally in patients covered by Medicare parts A and B in the
United States.
Methods
Medicare Standard Analytic Files containing 100% of inpatient and outpatient facility
records billed to Medicare from 2007 to 2012 were retrospectively analyzed. Total
30,327 patients were identified as having undergone fluoroscopic gastrostomy placement
defined by Current Procedural Terminology (CPT) codes CPT-49440. Patients were stratified
by age, sex, and comorbidities. Presence of comorbidities: diabetes mellitus (DM),
hypertension (HTN), hyperlipidemia (HLD), atrial fibrillation (a-fib), obesity, history
of smoking/smoker, and/or history of cancer; perioperative complications: pneumoperitoneum,
ileus, esophageal and gastric perforation, and/or damage to other intra-abdominal
organs; and postoperative complications: mechanical complication of gastrostomy, other
gastrostomy complications, surgical site infection, necrotizing fasciitis, bleeding,
ulceration, gastric outlet obstruction, colocutaneous fistula, and/or abdominal wall
pain were defined by International Classification of Diseases (ICD) 9 codes (detailed
in Supplementary Tables S1–S3, online only).
Results
Total 30,327 patients were identified as undergoing fluoroscopic gastrostomy. Regional
breakdown of patients represented include 6,042 from the mid-west; 5,014 from the
northeast; 12,146 from the south; 7,124 from the west; and 1 unknown with > 2,031
hospital centers represented. Age breakdown is noted ([Table 1]) with most patients age[3] 65 years (70.48%). Males made up 53.66% of the population (n = 16,272). Prevalence of comorbidities noted in patient group showed 38.81% with
DM (n = 11,771), 77.16% with HTN (n = 23,401), 58.68% with HLD (n = 17,795), 23.44% with a-fib (n = 7,111), 2.27% with obesity (n = 689), 37.26% with history of smoking (n = 11,299), and 28.32% with history of cancer (n = 8,589).
Table 1
Descriptive characteristics of patients undergoing fluoroscopic gastrostomy placement
|
N, (%)
|
|
Total procedures
|
30,327
|
|
|
Age (y)
|
|
|
Unknown
|
274 (0.90)
|
|
≤ 64
|
6,224 (20.52)
|
|
65–69
|
5,142 (16.96)
|
|
70–74
|
4,782 (15.77)
|
|
75–79
|
4,345 (14.33)
|
|
80–84
|
4,278 (14.11)
|
|
≥ 85
|
5,282 (17.42)
|
|
|
Sex
|
|
|
Male
|
16,272 (53.66)
|
|
Female
|
13,780 (45.44)
|
|
Unknown
|
275 (0.91)
|
|
|
Comorbidities
|
|
|
Diabetes
|
11,771 (38.81)
|
|
Hypertension
|
23,401 (77.16)
|
|
Hyperlipidemia
|
17,795 (58.68)
|
|
Atrial fibrillation
|
7,111 (23.44)
|
|
Obesity
|
689 (2.27)
|
|
History of smoking
|
11,299 (37.26)
|
|
History of cancer
|
8,589 (28.32)
|
Incidences of perioperative complications noted showed 0.20% with pneumoperitoneum
(n = 61), 0.43% with ileus (n = 130), 0.05% with esophageal/gastric perforation (n = 16), and 0.09% with intra-abdominal injury (n = 30). Incidences of 30-day postoperative complications showed 4.73% with mechanical
complications (n = 1,435), 2.73% with other gastrostomy complications (n = 828), 1.46% with surgical site infection (n = 443), 0% with necrotizing fasciitis (n = 1), 4.46% with bleeding (n = 1,353), 0.06% with ulceration (n = 17), 0.30% with outlet obstruction (n = 92), 0.22% with colocutaneous fistula (n = 67), and 9.25% with abdominal wall pain (n = 2,808) ([Table 1]).
Discussion
Our analysis of 30,327 patients undergoing fluoroscopic gastrostomy showed low rates
of perioperative complications. An early retrospective study by Hicks et al on 158
patients undergoing fluoroscopic gastrostomy reported rate of ileus large gastric
residual at 4% (n = 6).[6] A later retrospective study by Neeff et al of 18 fluoroscopic gastrostomy patients
reported a similar rate of ileus at 5.6% (n = 1).[13] Our study noted much lower rates of ileus with 130 (0.43%) patients only. A randomized
controlled study by Cosentini et al of 44 patients undergoing fluoroscopic gastrostomy
noted rate of pneumoperitoneum at 18% (n = 8).[7] Rates of perioperative esophageal/gastric perforation or other intra-abdominal injury
have not been specifically reported in previous literature. Our study noted low rates
of both esophageal/gastric perforation (n = 16, 0.05%) and intra-abdominal injury (n = 30, 0.09%).
Our analysis of postoperative complications following fluoroscopic gastrostomy was
similarly low and mostly in concordance with reported rates in the literature. An
early randomized control study of 66 patients by Hoffer et al reported five patients
with mechanical failure of either tube dislodgement, fracture, leakage, or block (7.6%),
five with wound infection (7.6%), none with bleeding (0%), and one with ulceration
(1.5%).[8] A later retrospective study of 193 patients by Silas et al similarly reported three
patients with mechanical failure of dislodgement or leakage (1.7%), four with wound
infection (2.3%), and three with pain (4%).[20] Rates of postoperative complications of “other gastrostomy complications” such as
tumor seeding or herniation, necrotizing fasciitis, gastric outlet obstruction, or
colocutaneous fistula have not been reported in the literature history. Our findings
of postoperative complications following fluoroscopic gastrostomies were comparable
with 1,435 patients experiencing mechanical complications (4.73%), 828 experiencing
“other gastrostomy complications” (2.73), 443 experiencing infection (1.46%), 1 experiencing
necrotizing fasciitis (0%), 17 experiencing ulceration (0.06%), 92 experiencing outlet
obstruction (0.030%), 67 experiencing colocutaneous fistula (0.2%), and 2,808 patients
experiencing abdominal pain (9.25%).
Overall, the peri- and postoperative complication rates of fluoroscopic gastrostomies
in this study are similar to those reported in the literature. Summary of the previous
studies note rates of ileus at 4%[6]; rates of mechanical failure including removal, leakage, and dysfunction ranging
at 1 to 6.2%[6]
[8]
[10]
[20]
[21]; rates of infection ranging at 3 to 7.6%[8]
[20]
[21]; rates of bleeding ranging at 0 to 3%[6]
[8]; rates of ulceration at 1.5%[8]; and rates of pain at 1.6%.[20] Of note, the aforementioned studies were limited to those performed in a hospital
or surgical center located in the United States only. Past studies performed outside
of the United States noted interestingly higher rates of complications for mechanical
failure ranging at 2.9 to 38%,[7]
[13]
[16]
[22] rates of infection at 3 to 22%,[7]
[13]
[14]
[15] and rates of pain at 35%,[14] though differing standards of practice and patient criteria may account for these
differences. Our study showed relatively low rates of complications, with rates of
abdominal pain (9.25%), bleeding (4.46%), and mechanical failure (4.73%) being the
most common. Our reported rates of infection were considerably lower than those reported
in the literature (1.46% in our study vs. range of 3–7.6% in the literature).
Limitations
Though the use of administrative data allows for access to large numbers of medical
data files across national hospitals with long-term tracking within the coding system,
analysis of such data does not allow for controls for individual procedural methods,
surgeon expertise, standardization of quality of care, or insight into criteria for
selection of patients. Administrative data are typically meant for administrative
and financial purposes rather than research, which may subject such data to errors
in accuracy and comprehensiveness due to reliance on interpretation of physician records
by a medical reviewer.
Table 2
Rates of peri- and postoperative complications following fluoroscopic gastrostomy
|
N, (%)
|
|
Perioperative complications
|
|
|
Pneumoperitoneum
|
61 (0.20)
|
|
Ileus
|
130 (0.43)
|
|
Esophageal/gastric perforation
|
16 (0.05)
|
|
Intra-abdominal injury
|
30 (0.09)
|
|
|
Postoperative complications
|
|
|
Mechanical complication
|
1435 (4.73)
|
|
Other gastrostomy complication
|
828 (2.73)
|
|
Surgical site infection
|
443 (1.46)
|
|
Necrotizing fasciitis
|
1 (0)
|
|
Bleeding
|
1353 (4.46)
|
|
Ulceration
|
17 (0.06)
|
|
Gastric outlet obstruction
|
92 (0.30)
|
|
Colocutaneous fistula
|
67 (0.22)
|
|
Abdominal wall pain
|
2808 (9.25)
|
Conclusion
Fluoroscopic guidance for gastrostomy placement is a safe procedure with low rates
of peri- and postoperative complications.
Institutional Approval
This study was approved by the institutional review board (IRB)—study number PDUVA001820.
