Endoscopic retrograde cholangiopancreatography (ERCP) is a relatively new endoscopic procedure combined with fluoroscopy that is performed for multiple diagnostic and therapeutic indications. It carries a known risk of radiation exposure to patients and staff. We aimed to examine radiation administration techniques and to measure the radiation dose delivered by these techniques.
This was a retrospective analysis of 437 ERCP procedures performed at a tertiary care hospital between April 2015 and April 2017.
A total of 437 ERCP procedural charts were reviewed: fluoroscopy administration was endoscopist controlled (EC, n = 187, 42.79%) or technician controlled (TC, n = 250, 57.21%). The mean (and SD) fluoroscopy time (FT) was 2.107 ± 2.0 minutes. The mean (and SD) dose–area product (DAP) was 15,227.371 ± 16,784.738 Gy·cm2. The degree of ERCP difficulty was evaluated as recommended by the American Society for Gastrointestinal Endoscopy, and graded 1–4. Level I TC procedures had a mean FT and DAP of 1.600 minutes and 12,644.72 Gy·cm2, respectively. The FT and DAP values for level I EC procedures were 1.514 minutes and 12,966.71 Gy·cm2, respectively, as compared with level IV TC procedures (mean FT, 2.539 minutes; mean DAP, 19,469.94 Gy·cm2) and level IV EC procedures (mean FT, 4.890 minutes; mean DAP, 37,921.00 Gy·cm2).
DAP and FT are increased significantly in EC ERCP in American Society for Gastrointestinal Endoscopy 4 procedures. Comparison of the different degrees of difficulty indicated that there is a linear correlation between the degree of difficulty and both FT and DAP.
Endoscopic retrograde cholangiopancreatography (ERCP) is a combination of upper endoscopy and fluoroscopy for diagnostic and therapeutic purposes to examine the pancreatic and hepatobiliary tract. Endoscopic retrograde cholangiopancreatography has various therapeutic applications, and although these applications are relatively less invasive than the previously employed surgical procedures, they are nevertheless more invasive than the previous diagnostic uses of ERCP. Therefore, gastrointestinal societies have established guidelines for ERCP and have attempted to improve the quality of the procedure. The American Society for Gastrointestinal Endoscopy (ASGE) refers to quality indicators in the following three categories: (1) preprocedure quality indicators: indication, sedation, and informed consent; (2) intraprocedural quality indicators: cannulation, procedure success rate, radiation dose, and fluoroscopy time (FT); and (3) postprocedure quality indicators: reporting and adverse events.[1,2]
The use of fluoroscopy presents the inherent risk of radiation exposure to both staff and patients. The increased use of ERCP as a therapeutic modality has often led to increased ERCP duration, with a subsequent increase in the radiation dose administered.[2,3] Such exposure is not well documented, and little data on this topic have been collected to date. However, radiation exposure is of great concern for patients and medical staff because of its rapidly increasing frequency and potential carcinogenic effects.[2,4,5]
The radiation dose from fluoroscopy varies with different variables.[6–8] Fluoroscopy time is a quantitative measure that adds to duration as the patient undergoes fluoroscopy during ERCP. Fluoroscopy time has been proposed as a quality measure for ERCP; however, it is but one of several factors that determine radiation exposure. Another variable is dose–area product (DAP, expressed as Gy·cm2), which is the product of the dose absorbed and the area irradiated. It is an estimation of the entire dose of radiation the patient receives and is thought to correlate with long-term biological risk from radiation exposure. It is automatically calculated by the fluoroscopy machine based on the radiation dose and the surface area being imaged.[9,10]
A scale created by the ASGE is used, with a grading of 1–4, to determine the possible degree of difficulty of performing ERCP. Procedures ranked as grade 1 are more straightforward, and therefore have a lower level of complications and radiation dose. Higher grades indicate increased difficulty, complications, and radiation dose.[10–12,13] Apart from the indication scale, the physical habitus of the patient also plays a role. It may impose limitations on patient positioning during the procedure, which would affect the angle for imaging and scope.[14,15]
The primary objective of this study was to compare the radiation exposure of the patient during ERCP by using two fluoroscopy techniques: when fluoroscopy is administered by the endoscopist using a pedal (endoscopist controlled, EC) versus when the radiology technician assists in administering fluoroscopy (technician controlled, TC). We also evaluated the radiation dose for different indications and degrees of difficulty of ERCP.
Study Setting and Population
The study was conducted at a tertiary level hospital. We conducted a chart review of ERCP procedures collected from the hospital's electronic systems as well as the radiology and gastroenterology divisions' internal systems. After each endoscopic procedure, physicians are required to write a full report, available from the division and the hospital system, stating the indications, findings, type of procedure performed, and postprocedure recommendations. The variables included were the patient's age, sex, reason for the procedure, duration of the procedure, DAP, FT, the degree of difficulty of each procedure, and the endoscopist. The FT and DAP were calculated by the fluoroscopy machine. The degree of difficulty was obtained from the physician report.
All ERCP from April 2015 to April 2017 performed and completed on patients aged less than or equal to 18 years and performed by a gastrointestinal endoscopist were included; noncompleted procedures were excluded. The sample size was 437.
The study design was a chart review–based retrospective cohort study, with data retrieved via the hospital's electronic system. The Internal Review Board approved the study.
Equipment and Location
Endoscopic retrograde cholangiopancreatography procedures are conducted at King Abduaziz Medical City in the fluoroscopy room at the medical imaging department or Endoscopy suite 5 in the endoscopy unit using Olympus duodenoscopes Evis Exera II (TJF-Q180V; Olympus Corporation, Center Valley, Pennsylvania) and General Electric (GE, Wauwatosa, Wisconsin) MPI fluoroscopy systems.
The demographics and clinical statistics are reported as frequencies and percentages for categorical data; continuous data are reported as the mean and SD. Appropriate transformations were done if the data had a skewed distribution. To compare the means of two groups, the two-sample t-test or nonparametric Wilcoxon test was used for continuous variables; analysis of variance (ANOVA) or the Kruskal–Wallis method was used for more than two groups, as appropriate. To determine the association between categorical variables, the Pearson χ2 or Fisher exact test was used. A p-value less than 0.05 was considered significant; the statistical analysis was performed using SAS 9.4 (SAS Institute, Inc., Cary, NC).
Patient Demographics and Clinical Characteristics
A total of 437 ERCP procedural charts were reviewed. The mean patient age was 53.67 ± 20.2 years. There were 199 male patients (45.54%) and 238 female patients (54.46%). There were 187 EC (42.79%) and 250 TC (57.21%) procedures. The variable reason for exam was classified into five categories: (1) obstructive jaundice, common bile duct stone (n = 332, 75.97%); (2) stent exchange, recurrent pancreatitis, or other (n = 18, 4.12%); (3) benign strictures (n = 21, 4.81%); (4) bile leakage (n = 13, 2.97%); and (5) malignant stricture (n = 53, 12.13%). The gastroenterology division has seven endoscopists who perform ERCP, one of whom administered fluoroscopy unassisted in 187 procedures (42.79%). All other endoscopists performed the remaining 250 procedures (57.21%) with technician assistance to administer fluoroscopy. The FT was measured in minutes; the mean FT was 2.107 ± 2.0 minutes; the minimum and maximum FT were 0.10 and 15.5 minutes, respectively. The mean DAP was 15,227.371 ± 16,784.738 Gy·cm2, while the minimum and maximum DAP were 11.64 Gy·cm2 and 139,069.000 Gy·cm2, respectively. For the degree of ERCP difficulty, we followed the ASGE grading system, with grading of 1–4. The level I TC procedures had a mean FT of 1.600 minutes and mean DAP of 12,644.72 Gy·cm2. The FT and DAP values for level I EC procedures were1.514 minutes and 12,966.71 Gy·cm2, respectively, as compared with level IV TC procedures (mean FT = 2.539 minutes, mean DAP = 19,469.94 Gy·cm2) and level IV EC procedures (mean FT = 4.890 minutes, mean DAP = 37,921.00 Gy·cm2).
Comparison of the EC Group Versus FT and DAP
An independent two-sample t-test was performed for the EC group versus FT, displayed as a histogram with normal plots (Fig. 1 and 2). The pooled and Satterthwaite (equal and unequal group) methods for variances showed that statistical significance (p < 0.0001) was small in comparison with the assumed alpha significance level of 0.05. Hence, the EC versus TC groups had a statistically significant difference for FT, which supports the conclusion that FT was significantly increased in the EC group. For DAP, an independent two-sample t-test was performed, and pooled and Satterthwaite variance testing for equal and unequal groups were performed. Based on the evidence, the null hypothesis was rejected, since the p-value was less than 0.05 (t = −3.34, p = 0.0009), suggesting a significant difference in DAP and supporting the conclusion that DAP was significantly increased in the EC group.
Degree of Difficulty Versus FT and DAP
The degree of difficulty was graded on a 1–4 scale, where 1 and 4 indicate the lowest and highest level of difficulty, respectively.
Degree of difficulty versus FT
ANOVA was used to compare the degree of difficulty versus FT and DAP. We applied a similar approach to general linear modeling (GLM) using SAS software PROC GLM procedures. The GLM model provided evidence that the overall model is statistically significant, and suggested that FT was significantly different in the different degree-of-difficulty groups (Fig. 3a, 3b).
Degree of difficulty versus DAP
A analysis similar to that described above based on the GLM model was done for the DAP on a log scale; the overall model was statistically significant (Fig. 4).
The use of fluoroscopy is inherent to ERCP. Since the quality measures proposed by the joint ASGE/ACG (American College of Gastroenterology) task force, there has been a proposal to include FT as an additional quality indicator to monitor patient radiation exposure during ERCP. In addition, most previous studies that examined patient radiation exposure during ERCP relied on FT. For example, a group that observed the factors associated with increased patient exposure used FT as a measure of radiation exposure. In another study, which found that radiation exposure during ERCP was lower when more experienced providers performed the procedure, used FT as the outcome measured. These studies assumed that radiation exposure correlated strongly with FT. Prior studies that examined the relationship between DAP and FT had a small number of patients, 20, 73, and 54 patients.[19,20] While there is a good correlation between DAP and FT, there is wide variability in DAP that is not accounted for by FT. Thus, the correlation is not accurate, resulting in both underestimations and overestimations of radiation dose if FT is relied upon alone. One large study showed that DAP is a more accurate indicator of patient radiation exposure during ERCP than FT, but did not emphasize the indication or degree of procedure difficulty. Therefore, the present study not only aimed at measuring the DAP and FT during ERCP, but also at correlating radiation exposure with different indications and degrees of difficulty, with comparison between different operation techniques.
Here, we observed significantly increased DAP and FT for EC radiation administration in ASGE grade 4 procedures. One possible explanation for this finding is the number of tasks that endoscopists are responsible for, such as focusing on the patient's condition, procedure techniques, and attempting to capture good images using their skill set. Although it might be more convenient for the radiation to be EC, this should be avoided because it increases the radiation exposure to both the patient and the staff. Comparison of the different degrees of difficulty indicated that there is a linear correlation between the degree of difficulty and both FT and DAP. There was little to no variance in procedures with low levels of difficulty, but the variance increased with the difficulty of the procedure (Table 1).
Procedures that required stent insertion had higher FT. The reason could be that fluoroscopy is required to confirm proper stent positioning. Benign strictures had the highest mean FT at 2.800 minutes, and bile leak had the lowest mean FT at 1.546 minutes. The highest difference in EC versus TC radiation administration was seen in ERCP involving benign strictures, where the FT was almost more than double. At the time of our study, we could not find any literature on fluoroscopy administration techniques during ERCP. However, fluoroscopy administration in other procedures such as urology and cardiology fluoroscopy has been well studied. The authors of a study of urological procedures requiring fluoroscopy found no difference between surgeon controlled versus TC fluoroscopy. There was an exception for one of the more complicated urology procedures in comparison to the other studied procedures, where the surgeon-controlled fluoroscopy appeared to show a significant reduction in radiation. Similar results were found in the published literature for cardiology procedures, where no significant difference was found in routine coronary angiography between cardiologist-controlled versus TC administration. In contrast, the present study shows that TC administration resulted in significantly less radiation exposure. The strength of our study is that we not only aimed to measure the DAP and FT during ERCP, but also took into consideration radiation exposure with regard to variable indications and degree of difficulty, with comparison between different operation techniques with a larger sample size.
The present study has some limitations. First, we used data from only one center. However, the data were retrospectively collected, which meant that the endoscopists were not aware that the data would be used for the study; this reduced the risk of bias and increased the accuracy of our data. In addition, the direct effect of the number of instruments used and specific therapeutic procedures performed during ERCP, such as basket sweeps, biopsies, and sphincterotomies, on the DAP and FT was not studied separately. Another limitation is that only one endoscopist administered radiation unassisted, which could be a confounder. However, the same endoscopist performed 187 of 437 procedures, which created a proper sample and distribution for comparing the two techniques studied.
The study shows that radiation is lower with TC fluoroscopy compared with EC in grade 4 ASGE procedures. Therefore, we recommend encouraging more technician involvement in ERCP. We also encourage further research on procedural radiation administration techniques to acquire sufficient data, and steer guideline formulation for safer radiation administration techniques during ERCP for the patients and medical professionals involved.
Source of Support: None. Conflict of Interest: None.