Vascular and Interventional Radiology
Technical Innovation
Radiologic Retrograde Gastrostomy Tube Insertions Without the Use of Barium: Implications for Radiation Dose in Children
Audio Available | 
Share
OBJECTIVE. Historically, percutaneous radiologic retrograde gastrostomy tube insertions in children were performed using barium to opacify the colon. The purpose of this study was to examine whether performing radiologic gastrostomy tube insertions without barium affects patient radiation dose.
CONCLUSION. The method of radiologic gastrostomy tube insertion without barium significantly reduced patient dose, fluoroscopy time, dose-area product (DAP), and weight-and time-adjusted DAP.
Keywords: barium, gastrostomy, pediatric, radiation
Gastrostomy tubes are commonly placed in children who cannot meet their nutritional needs by mouth and have been shown to improve quality of life for these patients. Radiologically placed gastrostomy tubes are associated with good outcomes [1, 2]. In children, antegrade and retrograde radiologic gastrostomy tube insertion techniques have been described [3]. Colonic puncture is a recognized risk of both techniques and occurs in 1–2% of cases [3–6]. Colon opacification with barium has traditionally been used in children to avoid puncture of the colon. This is achieved by either administering barium orally the night before, anticipating that it will reach the colon by the procedure time, or infusing dilute barium rectally via Foley catheter at the start of the procedure. More than a decade ago, our institution switched from the former to the latter administration technique; thus, rectal infusion of dilute barium became our standard practice. However, barium in the imaging field increases the radiation dose required through automatic exposure control (AEC) feedback [7]. Hence, we now place gastrostomy tubes using radiologic techniques without barium administration. The purpose of this study was to determine whether radiologic gastrostomy tube insertion without the use of barium affects the radiation dose imparted to the child.
| Radiologic Gastrostomy Tube Insertion Techniques |   |
|---|
Preliminary mapping of liver and spleen is performed with ultrasound. Stomach inflation is performed through a nasogastric tube (NGT) before puncture, usually after administration of IV glucagon. A single puncture of the stomach is made with a needle pre-loaded with one retention anchor suture by using fluoroscopic guidance, and the retention anchor suture is deployed with a 0.035-inch wire. After tract dilation, the gastrostomy tube is advanced using the Seldinger technique (Fig. 1).
 View larger version (486K) | Fig. 1 —Fluoroscopic image shows gastrostomy tube advancement technique. Air-filled stomach with nasogastric tube in situ (black arrow) is visualized, and puncture needle tip (white arrow) is seen poised cephalad to transverse colon, filled with dilute barium (B). |
After ultrasound mapping of the abdomen, a Foley catheter is placed into the rectum, and dilute barium is infused to delineate the transverse colon. The site for gastric puncture is chosen using fluoroscopic guidance in different projections, avoiding adjacent bowel, particularly the opacified colon. The stomach is then punctured under real-time fluoroscopic guidance as described already.
After mapping of the liver and spleen, real-time ultrasound examination is used to visualize the empty stomach and adjacent bowel loops below the left lobe of the liver (Fig. 2A). Stomach distention through the NGT is visualized with ultrasound and fluoroscopy as it distends and displaces the adjacent bowel caudad (Fig. 2B). Real-time ultrasound examination is used to choose the site for puncture, which is confirmed fluoroscopically. The needle is advanced to the stomach wall, away from the adjacent bowel, under ultrasound guidance. The stomach is then punctured with a combination of real-time fluoroscopic and ultrasound guidance as described already (Figs. 2C and 2D).
 View larger version (232K) | Fig. 2A —Real-time ultrasound and fluoroscopy for radiologic gastrostomy tube placement without barium. A, Sagittal ultrasound image (linear 12-MHz probe) of 6-month-old girl shows empty stomach (S) below left lobe of liver (L) with colon (C) placed beside it. |
 View larger version (292K) | Fig. 2B —Real-time ultrasound and fluoroscopy for radiologic gastrostomy tube placement without barium. B, Sagittal ultrasound image (curvilinear 8-MHz probe) of different patient from A shows stomach (S) distended with air below liver (L) with colon (C) displaced caudad. |
 View larger version (201K) | Fig. 2C —Real-time ultrasound and fluoroscopy for radiologic gastrostomy tube placement without barium. C, Matched lateral fluoroscopic image of same patient as B shows stomach (S) distended with air and colon (C) displaced caudad. Needle (black arrow) inserted into stomach and nasogastric tube (white arrow) are visible. |
 View larger version (211K) | Fig. 2D —Real-time ultrasound and fluoroscopy for radiologic gastrostomy tube placement without barium. D, Sagittal ultrasound image (curvilinear 8-MHz probe) of additional patient shows stomach (S) with wire inserted (arrow) and colon (C) positioned more caudally. |
| Materials and Methods | |
|---|
Institutional approval was granted for this quality-improvement project. By use of the hospital PACS, radiologic gastrostomy tube insertions between 2011 and 2013 were reviewed. Inclusion criteria were patients between 5 and 20 kg who underwent radiologic gastrostomy tube insertion as a single procedure, in either of two modern interventional radiology (IR) suites with radiation dose reports exported to a PACS. Exclusion criteria were patients who underwent a radiologic gastrostomy tube insertion in an older IR suite without available dose reports, those who underwent gastrojejunostomy tube insertions, those who underwent enterostomy maintenance procedures, children less than 5 kg or more than 20 kg, or those undergoing concurrent fluoroscopic procedures under the same sedation. Patient records, PACS, IR database, and morbidity and mortality records were reviewed for complications pertaining to the procedure that were reported within 90 days after the procedure.
Within the selected time period, 308 patients who underwent radiologic gastrostomy tube insertion were identified, 112 of whom met the inclusion criteria (52 with barium administration, 60 without barium administration). Analysis included clinical data (patient age, weight, procedural complications) and radiation data (total dose [in mGy], dose-area product [DAP] [in μGy ⋅ m2], total fluoroscopy time [in minutes], time-adjusted DAP [in μGy ⋅ m2/min], and weight-adjusted DAP [in μGy ⋅ m2/kg]). Procedures were performed by five staff interventional radiologists (n = 7, 15, 29, 29, and 32 for each radiologist). No procedures were performed by residents; any performed by fellows were under direct staff supervision.
Statistical analysis of the data was performed using SPSS Statistics software (version 21.0 for Windows, IBM). Tests were performed to assess normal distribution of patient weight. ANOVA was performed to examine any difference in exposure between both groups (with barium and without barium). Binary logistic regression was used to identify any difference in complications between each group.
| Results | |
|---|
Both groups were well matched for age and weight. The mean (± SD) weight among the without-barium group was 10.347 (± 5.913) as compared with 11.402 (± 6.189) kg in the with-barium group. The mean age of patients in the without-barium group was 2.95 (± 3.546) as compared with 3.084 (± 3.307) years in the with-barium group. No statistical difference was found between both groups.
ANOVA showed significant reduction in several radiation metrics between the with-barium and the without-barium groups, with mean fluoroscopy time ranging from 3.167 to 7.207 minutes (p < 0.05), dose from 2.750 to 7.043 mGy (p < 0.05), and DAP from 27.488 to 85.060 μGy ⋅ m2 (p < 0.05), respectively (Table 1). Results remained significant after adjusting for weight differences (p < 0.05). The time-adjusted DAP also significantly decreased from 9.926 to 3.917 μGy ⋅ m2/min (p < 0.05). Results of one-way ANOVA comparing interoperator differences in DAP were not significant between the groups (p > 0.05). Complication rates were within the published range (Table 2).
| Discussion | |
|---|
This study showed a reduction in fluoroscopy time, DAP, time-adjusted DAP (i.e., per minute; reflecting the absence of barium in the image and AEC response), and weight-adjusted DAP (i.e., per kilogram) in the radiologic gastrostomy tube insertions without barium. The advantages of not using barium are threefold. First, ultrasound is less invasive than rectal administration of barium, especially in an immune-compromised patient. The alternative of orally administered barium the night before the procedure is inconsistent in its opacification of the transverse colon. Second, the time-adjusted DAP is lower, indicating the dose per unit time is reduced. This is due to absence of contrast medium in the imaging field and AEC response, and it is important regardless of whether barium is given rectally or orally. Third, fluoroscopy time is significantly reduced because intermittent fluoroscopy of the colon during filling is not required, and image planning for the puncture is largely by use of ultrasound rather than entirely fluoroscopically. Implicit benefits for occupational dose in terms of reduced scatter were not specifically assessed and are outside the scope of this study.
The advantages of successful radiologic gastrostomy tube insertion without barium administration must be balanced against any increase in complication rate. The present study was unable to show significant difference in complications between both groups (by use of logistic regression), owing to insufficient power at this sample size (< 50% power at 95% CI). However, the complication rates were within limits previously reported in the literature for both groups [4–6, 8]. All nongastric punctures (colon, small bowel) occurred in patients with suboptimal gastric distention (in two cases, glucagon was not used) and challenging anatomy. The overall rate of bowel puncture in our study (1.8%) was within reported rates in the literature [6]; all were immediately recognized, and no patient experienced subsequent sequelae. Nevertheless, this emphasizes the importance of good gastric distention and careful ultrasound technique.
Radiation protection and dose reduction are especially important in pediatric patients, given their increased radiosensitivity, proximity of radiosensitive organs, higher overall cell turnover, and long life expectancy. Cumulatively, this population is often subject to many imaging studies during their lives. Therefore, under the as low as reasonably achievable (ALARA) principle, it is reasonable to consider radiologic gastrostomy tube insertion without barium to be superior to the technique with barium administration from a dose reduction perspective.
| Acknowledgments | |
|---|
We thank Avnesh Sinh Thakor for valuable contributions to the preparation of the manuscript, the Image-Guided Therapy team for their help with this study, and especially Doina Filipescu for provision of the data.
| References | |
|---|
