Introduction
Endoscopic retrograde cholangiopancreatography (ERCP) is the standard procedure for
treatment of pancreaticobiliary diseases that require use of fluoroscopy. Thus, personnel
involved in the ERCP room are at risk for radiation hazards [1 ]. Ionizing radiation from fluoroscopy potentially causes cell injury to various organs
(tissue reaction) [2 ] and increases risk of cancer or genetic defects (stochastic effect) [3 ]
[4 ]
[5 ]. To avoid this adverse effect, the “As Low As Reasonably Achievable (ALARA)” principle
is recommended for radiation safety [5 ]
[6 ].
Radiation doses to the patient and medical personnel depend on several factors such
as fluoroscopic time, thickness of the patient’s exposed body, distance between the
X-ray tube and personnel, and distance between image intensifier and patient [5 ]
[7 ]
[8 ]
[9 ]. With the automatic exposure control function, the fluoroscopy system will increase
the tube voltage in a thicker object, when compared with the thinner object, to maintain
the image quality [10 ]
[11 ]. The patient’s positions during ERCP can be either prone, supine or left lateral
decubitus (LLD) depending on the endoscopist’s preference [12 ]. Because the vertical body thicknesses of prone and supine are similar and thinner
than the LLD position, the adjusted beam by the fluoroscopic machine on these two
positions is speculated to be lower than in the LLD position [11 ]. Moreover, the increase in distance between the image intensifier and the X-ray
tube or patient could result in the radiation being more scattered in a thicker object.
We then hypothesized that the scattered radiation to ERCP personnel could be lower
in prone and supine positions than in the LLD position. In our experience, performing
ERCP in a supine position is more difficult and technically more challenging compared
to a prone position [12 ]
[13 ], while the LLD position had a comparable success rate to that of the prone position
[14 ]. We then aimed to compare the radiation exposure in ERCP personnel between patients
lying in a prone position and LLD position and chose the ocular lens, which is the
most susceptible organ, as our target of comparison [2 ].
Patients and methods
Patients
This was a parallel prospective randomized study performed at the Excellence Center
for Gastrointestinal Endoscopy of the King Chulalongkorn Memorial Hospital, Bangkok,
Thailand. Consecutive patients who were aged 18 or older and indicated for ERCP during
July to October 2016 were screened. Exclusion criteria were pregnancy, American Society
of Anesthesiology (ASA) physical status class III – IV, unstable vital signs or surgically
altered anatomy, need to change the patient’s position during the procedure, needing
a specific ERCP position (such as hilar biliary obstruction and pancreatic pathology),
and informed consent could not be obtained. The study protocol was approved by the
Chulalongkorn University Institutional Review Board (IRB number 624/58).
Fluoroscopy system and setting
The mobile C-arm, under-couch fluoroscopy system (BV Pulsera, Philips, Amsterdam,
The Netherlands), with the “last image hold” function producing pulsed fluoroscopy
at 12.5 pulses per second, was used in this study. The examination mode of the anatomically
programmed fluoroscopy was selected as the abdomen and the nominal II format was set
at 31 cm. Tube voltage and tube current-time were adjusted automatically to maintain
a constant radiation dose entering the over-couch image intensifier. The lead curtain
was mounted around the table during procedures. A well-trained assistant controlled
the fluoroscopy according to the endoscopist’s request but it was not adjusted for
image magnification.
Procedure and radiation measurement
Patient age, gender, body mass index, body thickness, and indications for ERCP were
recorded. Patients were randomized into two groups by computer-generated codes in
block-of-four. Randomization codes were inserted in the sequentially numbered envelopes.
An envelope was opened consecutively in the endoscopy room to assign the patient’s
position (prone or LLD). All ERCP personnel wore a wraparound lead apron and thyroid
collar. ERCP was performed according to the indication in a standard technique as
described elsewhere [15 ] under moderate sedation using intravenous meperidine and midazolam. The distance
of all ERCP personnel from the X-ray tube was approximately 30 to 40 cm for the primary
endoscopist and the nurse anesthetist and 60 cm for the secondary endoscopist ([Fig. 1 ]). The detector units, Personal Dose Meter (PDM) of the DoseAware system (Philips,
Amsterdam, the Netherlands), were placed outside the thyroid collar of each ERCP team
member on the side that was close to the fluoroscopy system; that is, on the left
side of the primary and secondary endoscopists, and on the right side of the nurse
anesthetist, and these represented the eye exposure of the involved personnel. The
PDM was calibrated in terms of the dose equivalent quantity H(p)(3) representing radiation
doses at the ocular lens [16 ]. In this study, the primary endoscopist began first and the attending endoscopist
replaced him whenever the ERCP procedure failed to progress. To maintain the correct
PDM positions, the detector units were swapped between the two endoscopists whenever
they swapped their positions. After the ERCP procedure, the total fluoroscopic time
(minute), fluoroscopy tube voltage (kV), fluoroscopy tube current (mA), patient entrance
skin dose rate (mGy/min), dose area product (Gy-cm2 ) and equivalent dose (mSv) were recorded. The ocular lens equivalent dose was presented
as the equivalent dose per procedure (mSv/procedure) and the equivalent dose per fluoroscopy
time (equivalent dose rate; mSv/hour) [17 ].
Fig. 1 Standing position of the ERCP personnel.
Sample size and statistical analysis
A previous study [18 ] measured radiation exposures in 4 different areas of the endoscopist, including
left eye, thyroid, left forearm and left leg, while performing ERCP on patients with
prone and LLD positions. The sample size was calculated based on the data for ocular
lens equivalent dose of the endoscopist in patients with prone and LLD positions (0.059
and 0.084 mSv, respectively) from a previous study [18 ]. To demonstrate a 20 % difference in the ocular lens equivalent dose at a power
of 90 % and type I error of 5 %, the calculated number of needed patients in each
group was 27. Continuous variables were displayed as the mean ± standard deviation
(SD), or median (interquartile range, [IQR]), and the difference between the two groups
was analyzed with a Student’s t-test, or Mann-Whitney U test where appropriate. Categorical
variables were displayed as the percentage or proportion and the differences between
the two groups were analyzed with a Chi-squared or Fisher’s exact test where appropriate.
Statistical analysis was performed with IBM SPSS statistics 19. A two-sided P value < 0.05 was considered to be significant.
Results
During the study period, there were 71 consecutive patients who underwent ERCP, and
16 patients were excluded because of a hilar lesion (n = 14) or unstable vital signs
(n = 2). Fifty-five patients were randomized to prone (n = 27) and LLD (n = 28) position
groups. One patient in the LLD group was excluded because the position was changed
to prone during the procedure because of difficult cannulation and a double guidewire
technique to achieve deep biliary cannulation was required. The final analysis was
made from these 54 patients (27 patients in each group; [Fig. 2 ]). Demographic parameters including age, gender, body mass index (BMI) in the prone
position and LLD position groups were not different ([Table 1 ]). Indications for ERCP in the prone and LLD groups were choledocholithiasis (63 %
vs. 67 %), malignant biliary stricture (30 % vs. 22 %) and benign biliary stricture
(7 % vs. 11 %; P = 0.780), respectively. The switch-over rate from primary to secondary endoscopist
were 74 % and 67 % in prone and LLD groups, respectively (P = 0.766). The mean vertical thickness in the prone and LLD groups was 27.2 vs. 20.2 cm
(P < 0.001), respectively. Median fluoroscopy time, median fluoroscopy tube voltage,
median fluoroscopy tube current, median dose area product, and median patient entrance
skin dose rate in the prone and LLD groups were 4.14 vs. 4.06 min (P = 0.993), 70 vs. 72 kV (P = 0.549), 2.30 vs. 2.29 mA (P = 0.659), 23.2 vs. 22.3 Gy-cm2 (P = 0.742), and 5.5 vs. 5.7 mGy/min (P = 0.197), respectively ([Table 1 ]).
Fig. 2 Flowchart of the study.
Table 1
Baseline characteristics and fluoroscopic parameters of patients in prone and left
lateral decubitus (LLD) positions.
Parameters
Prone
LLD
P value
Age[1 ] (years)
67 (15)
65 (27)
0.723
Male:Female (n)
15:12
10:17
0.17
BMI[2 ] (kg/m2 )
22.2 ± 4.07
22.9 ± 3.45
0.476
Indication for ERCP (%)
63
67
0.78
30
22
7
11
Switch over from primary to secondary endoscopist n (%)
20 (74)
18 (67)
0.766
Vertical thickness (cm)[2 ]
20.2 ± 4.18
27.2 ± 3.71
< 0.001
Fluoroscopy time[1 ] (minutes)
4.14 (4)
4.06 (4)
0.993
Fluoroscopy tube voltage[1 ] (kV)
70 (12)
72 (7)
0.549
Fluoroscopy tube current[1 ] (mA)
2.30 (0)
2.29 (0)
0.659
Dose area product[1 ] (Gy-cm2 )
23.2 (19)
22.3 (24)
0.742
Patient entrance skin dose rate[1 ] (mGy/min)
5.5 (1)
5.7 (4)
0.197
BMI, body mass index; ERCP, endoscopic retrograde cholangiopancreatography
1 Data presented as the median (interquartile range; IQR)
2 Data presented as the mean ± standard deviation (SD)
Median ocular lens equivalent doses in the primary endoscopist were significantly
lower in the prone versus the LLD positions (0.0192 vs. 0.0307 mSv, P = 0.035) and the nurse anesthetist median ocular lens equivalent doses were also
significantly different (0.0173 vs. 0.0442 mSv, P = 0.002), but the secondary endoscopist did not show a difference ([Table 2 ]). Median ocular lens equivalent dose rates were significantly lower in the prone
position than in the LLD position in all 3 personnel (0.28 vs. 0.43 mSv/hr; P = 0.001) in the primary endoscopist, 0.18 vs. 0.25 mSv/hr (P = 0.015) in the secondary endoscopist and 0.23 vs. 0.54 mSv/hr (P < 0.001) in the nurse anesthetist ([Table 2 ]).
Table 2
Ocular lens equivalent dose of ERCP personnel in prone and LLD positions.
ERCP personnel
Median (IQR) ocular lens equivalent dose (mSv)
P value
Median (IQR) ocular lens equivalent dose rate (mSv/hour)
P value
Prone
LLD
Prone
LLD
Primary endoscopist
0.0192 (0.0207)
0.0307 (0.0245)
0.035
0.28 (0.21)
0.43 (0.20)
0.001
Secondary endoscopist
0.0096 (0.0135)
0.0154 (0.0192)
0.113
0.18 (0.09)
0.25 (0.12)
0.015
Nurse anesthetist
0.0173 (0.0250)
0.0422 (0.0346)
0.002
0.23 (0.18)
0.54 (0.41)
< 0.001
IQR, interquartile range; LLD, left lateral decubitus
We then calculated the possible number of cases and fluoroscopy time allowances for
each individual under the two positions. According to the new recommendation of the
International Commission on Radiological Protection (ICRP), which limits the annual
radiation dose for the ocular lens at 20 mSv[2 ], the calculated maximum number of cases per annum for each staff member without
wearing radiation protective eyewear in prone and LLD positions were 1,042 and 651
cases for the primary endoscopist, 2,083 and 1,302 cases for the secondary endoscopist,
and 1,157 and 473 cases for the nurse anesthetist ([Table 3 ]). The annual fluoroscopy time limit in the prone and LLD positions were 71.42 and
46.51 hours for the primary endoscopist, 111.11 and 80 hours for the secondary endoscopist,
and 86.96 and 37.03 hours for the nurse anesthetist ([Table 3 ]).
Table 3
Maximum annual procedures and fluoroscopy time for ERCP personnel (without wearing
eye protection) not to exceed the ocular lens equivalent dose threshold of 20 mSv.
ERCP personnel
Maximum procedures
Maximum annual fluoroscopy time (hours)
Prone
LLD
Prone
LLD
Primary endoscopist
1,042
651
71.42
46.51
Secondary endoscopist
2,083
1,302
111.11
80
Nurse anesthetist
1,157
473
86.96
37.03
ERCP, endoscopic retrograde cholangiopancreatography; LLD, left lateral decubitus
Discussion
This study demonstrated that performing ERCP in a prone position significantly exposed
the primary endoscopist and the nurse anesthetist to lower ocular lens equivalent
doses. By simply changing the patient position from the LLD position to a prone position,
the ocular lens equivalent doses to the primary endoscopist and the nurse anesthetist
were reduced by 37.5 % and 59.0 %, respectively. Although the equivalent dose in the
secondary endoscopist was also reduced by 37.5 %, this did not reach a statistically
significant difference. We speculate that the sample size was too small to have enough
power to demonstrate the difference in equivalent dose to the secondary endoscopist
between the two positions. Of note, the secondary endoscopist is exposed to radiation
at a much lower level than the other two personnel. Interestingly, when we calculated
the ocular lens equivalent dose per the procedure time (dose rate) [17 ], the prone position significantly yielded the lower dose rates in all three personnel
when compared with the LLD position. This calculation eliminated variation in fluoroscopy
time, which was influenced by procedure difficulty [19 ]. This confirmed that, within the same timeframe, the prone position significantly
lowered the ocular-radiation exposure to all personnel.
As we hypothesized earlier, the two factors that might affect the radiation exposure
to ERCP personnel were modulated dose of voltage adjustment by the X-ray tube and
the scattered radiation acquired from the increase in distance between the image intensifier
and the patient [11 ]
[20 ]. Among all baseline characteristics, we showed that only the vertical thickness
of the patient in LLD group was significantly higher than in the prone position group. Interestingly,
radiation doses from the X-ray tube, called the modulator effect (tube voltage and
tube current), were not significantly different. Furthermore, the dose area product
and the patients’ skin dose rates were also not significantly different. These reflected
that the fluoroscopy system did not significantly increase the tube voltage and that
the patient was not exposed to more radiation from the voltage adjustment when changing
the position from prone to LLD. Unlike the previous study in a phantom, which demonstrated
that the radiation dose significantly increased along with the increment of thickness
[11 ], the difference in the phantom was that the density of the medium is more homogeneous
than the real human body and this, in turn, can cause the difference in radiation
penetrance [21 ]. The current study demonstrated that, in a real human body, the increment of vertical
thickness from 20 cm to 27 cm did not significantly increase the exposure by the X-ray
tube. Perhaps the scattered ray from the patients is the only thing responsible for
the increase in radiation exposure to the ERCP personnel. Therefore, to reduce radiation
scatter during ERCP, the image intensifier should be positioned as close to the patient’s
body as possible [5 ].
The ocular lens is composed of radiosensitive tissues that are at risk of developing
cataracts after receiving significant ionizing radiation [22 ]. Since April 2011, the ICRP has lowered the equivalent dose limit for the lens of
the eyes during occupational exposure from 150 mSv/year to 20 mSv/year as averaged
over the period of 5 years, with no single year exceeding 50 mSv [2 ]. However, radiation protection for eyes has not yet been mandated by major international
guidelines on radiation protection during ERCP. Radiation-protective eyewear is recommended
as an optional measure by the American Society for Gastrointestinal Endoscopy (ASGE)
[23 ] and is recommended only when using over-couch fluoroscopy by the European Society
of Gastrointestinal Endoscopy (ESGE) [5 ]. Patient position in ERCP can be either prone, supine, or LLD. Selection of position
is dependent upon patient factors (e. g., neck mobility, presence of abdominal drains
or wounds), airway management [24 ], or endoscopist’s preference. LLD is considered to be easier on airway management
and scope intubation/positioning; however, the examination is limited for extra-hepatic
bile duct indications, because the anatomical orientation is suboptimal for pancreatic
duct or biliary bifurcation [24 ].
Although LLD is not a common position for ERCP in the United States [12 ], in Thailand, LLD is the most common position, which accounts for 50 %, followed
by prone (32.7 %) and then supine (17.3 %; personal unpublished survey from ERCP endoscopists
across Thailand). Furthermore, many endoscopists have overlooked radiation protection
to their eyes. In the same survey, only 38.2 % reported availability of radiation-protective
eyewear and only 7.5 % reported wearing that eyewear at all times. Likewise, the survey
from Korea [25 ] revealed that radiation-protective eyewear was used by endoscopists only 37.8 %
of the time, while a lead apron and thyroid shield were used 98.7 % and 94.7 % of
the time, respectively.
This study emphasized that not only the primary endoscopist but also the nurse anesthetist
is at risk of developing cataracts and eyewear use should be the standard of practice
because of the potential for exceeding the allowance of annual radiation exposure
to the ocular lens [2 ] (if the annual radiation exposure exceeds 1,000 procedures in a prone position or
600 procedures in an LLD position). The calculated procedure limit was based on the
C-arm fluoroscopy system, and the ALARA approach applied in this study, for example
using pulsed fluoroscopy, had the lowest possible pulsed rate, rather than continuous
fluoroscopy, stored as the “last image hold” rather than taking radiographs, and avoidance
of the magnification mode [5 ]. When the ALARA protocol is not in effect or when facing a complex ERCP case that
requires a longer fluoroscopy time, the limited number of procedures per annum could
have been lower.
This study had some limitations. First, we could not blind the endoscopist and the
assistant who controlled the fluoroscopy. However, the attending staff members had
experience in ERCP of more than 200 cases/year and were well trained on radiation
safety and complied with the ALARA principle. Because the results of fluoroscopy parameters,
especially fluoroscopy times, were no different between the two groups and were comparable
with other studies [26 ]
[27 ], there was low bias for the fluoroscopy control in this study. Second, we excluded
complex cases, especially hilar cholangiocarcinoma, because of the need for an antero-posterior
view of fluoroscopy, as those indications might require a longer fluoroscopy time
and result in greater radiation exposure [5 ]
[28 ]. We then calculated an equivalent dose rate to eliminate variation in fluoroscopy
time and this might be appropriate for radiation monitoring rather than the mean dose
per procedure [17 ]. Lastly, because it was not practical to place the PDM near the eyes as that might
obscure the visual field of the personnel, the ocular lens doses were calculated based
on calibration from the measured doses at the neck level. This adjustment was suggested
by the previous study that demonstrated that placement of the PDM at the thyroid collar
was suitable for the ocular lens dose assessment when compared with direct measurement
close to the eyes [20 ]. Furthermore, the ocular lens doses in the current study were in line with the previous
study that made measurements directly between the eyes [17 ]. Regarding the involvement of trainees, the switch-over rate from primary to secondary
endoscopist was comparable in both groups and we always swapped the PDMs when the
primary and secondary endoscopists changed their positions. This ensured that the
correct exposure measurement was based on the standing position (not based on the
individual). Of note, we observed that the trainees spent most of their time on biliary
cannulation but that would not have much of an effect on the radiation exposure because
it required proportionally less fluoroscopy.
Conclusion
In conclusion, performing ERCP with a patient in a prone position with the image intensifier
positioned as close as possible to the patient body and using a lead curtain reduces
by one-third the ocular-radiation exposure to ERCP personnel from the LLD position.
Therefore, more annual ERCPs can be performed in patients in the prone position under
the recommended dose limit.
