No difference between supine and prone position for ERCP in conscious sedated patients: a prospective randomized study

• Introduction
• Patients and methods
• Statistical analysis
• Results
• Discussion
• References

Background and aims: Endoscopic retrograde cholangiopancreatography (ERCP) is usually performed with the patient prone or in the left lateral position. The supine position could be more comfortable and may facilitate airway management. On the other hand, technical difficulties and a greater risk of adverse cardiorespiratory events have been shown when ERCP is performed in a supine patient. Our aim was to assess, in a tertiary referral center, the differences between performing ERCP with the patient supine or prone, in terms of technical features and complications both during and after the procedure.

Patients and methods: Between December 2005 and May 2006, 120 patients (66 female, mean age 62 years) who had an intact papilla and were candidates for therapeutic ERCP were prospectively randomized to undergo ERCP under conscious sedation with midazolam, in the prone (n = 60) or supine (n = 60) position, by an expert endoscopist (tutor) or a trainee. The following parameters were recorded: difficulty of cannulation and difficulty of ECRP procedure, time needed to visualize the papilla, time needed to achieve opacification and cannulation, exam duration, episodes of tachy/bradycardia and desaturation, episodes of duodenoscope displacement into the stomach, and complications.

Results: Ninety-eight patients underwent ERCP for benign disease and 22 for malignant biliary strictures. The ERCP success rate was 98.3 % in the tutor group and 43.3 % in the trainee group. No significant differences were found between the two groups of operators (tutors and trainees) in the recorded parameters and complication rates encountered in prone versus supine patients.

Conclusion: Our results show that ERCP success rates and complications (intraoperative and postoperative) are similar whether ERCP is performed with the patient prone or supine, even when operators are of differing skill levels. Training, technique, and a proper learning phase are recommended in order to perform ERCP with no differences whether the patient is prone or supine.

Introduction

Patients undergoing endoscopic retrograde cholangiopancreatography (ERCP) are usually examined in the left lateral or prone position [1] [2] [3]. Compared to the supine position, the left lateral and prone positions are believed to carry a lower risk of inhalation, to allow easier passage of the scope through the pharynx, and to permit a more comfortable position for the endoscopist ([Fig. 1]). Conversely, the use of the supine position could potentially offer a better view of the biliary and pancreatic radiological anatomy. Intuitively, one might expect that the supine position would also simplify the entire procedure in patients requiring general anesthesia with orotracheal intubation. In our daily clinical practice, the supine position is used for patients with hilar biliary obstruction (display of the intrahepatic anatomy is better), those who have previously undergone a Billroth II gastrectomy (progression of the scope is easier and visualization of the loops better), and in patients under general anesthesia (airway control and resuscitation maneuvers are easier).

The only prospective study [4] comparing the prone and supine positions in patients undergoing ERCP under conscious sedation showed a higher risk of adverse cardiorespiratory events (P = 0.04), and a lower cannulation success rate (P = 0.05) when patients were supine. However, while suggesting to that the supine position be avoided as far as possible in day-to-day practice, the authors concluded that their results needed to be confirmed by other centers with greater experience in performing ERCP in supine patients.

We therefore decided to conduct a randomized, prospective study to compare use of the supine and the prone position in patients undergoing ERCP under conscious sedation, in terms of feasibility, safety and efficacy. To further validate the results, the procedures were also randomly assigned to be performed by expert endoscopists or by trainees.

Patients and methods

Consecutive patients with an intact papilla referred to our unit for biliary and/or pancreatic therapeutic ERCP were enrolled. The study protocol was approved by the Ethical Committee of our university and written informed consent was obtained from every patient entering the study. The following were exclusion criteria: patient age below 18 years; ERCP as a diagnostic procedure; requirement for general anesthesia or deep sedation with propofol; previous partial gastrectomy or other major surgery involving the upper gastrointestinal tract (e. g. total gastrectomy, gastroenteric and bilioenteric anastomosis); previous endoscopic or surgical sphincterotomy; previous endoscopic or percutaneous biliary drainage; ampullary tumor; intended cannulation of the minor papilla duodenal stricture; patient not fit to undergo surgery (ASA 4).

Using a computer-generated randomization list, patients were randomly assigned to start the procedure in either the prone or the supine position (“prone” and “supine” groups). They were also randomly assigned to have the procedure performed by a senior endoscopist (GC, VP, MM) or a trainee (< 250 ERCPs).

ERCPs were performed with large-channel therapeutic duodenoscopes (TJF 140R, TJF 145, TJF 160R; Olympus Co. Ltd., Tokyo, Japan). Patients were sedated with midazolam (2 - 10 mg i. v.) and fentanyl (0.05 - 0.1 mg i. v.) titrated according to the patient’s age, comorbidities, and compliance. To inhibit duodenal peristalsis, glucagon (1 - 2 mg i. v.) was routinely given.

To stabilize the tip of the duodenoscope in the duodenum and to achieve correct visualization of the papillary area with the patient supine, the endoscopist has to slightly modify the “standard” position by turning the handle of the scope 90° clockwise ([Fig. 2]) or, alternatively, by turning his/her back on the patient ([Fig. 3]).

All patients were monitored with pulse oximetry. Low oxygen volume (2 - 3 L/min) was continuously administered through the nose during the procedure. Significant oxygen desaturation was defined as an oxyhemoglobin saturation (Sao₂) of less than 90 % for more than 15 seconds. Heart rate was also continuously monitored. Bradycardia was defined as a 25 % heart rate reduction below the baseline level, and tachycardia as a 25 % heart rate increase over the baseline level.

The primary end point of this study was to compare the difficulty of achieving deep cannulation of the desired duct, common bile duct, or main pancreatic duct. The following parameters were also systematically recorded: time needed to visualize the papilla, time needed to inject contrast into the desired duct, time needed to obtain deep cannulation of the desired duct, total procedure time (from insertion of the duodenoscope to its withdrawal), episodes of tachycardia, episodes of bradycardia, episodes of oxygen desaturation, episodes of displacement of the duodenoscope from the duodenum into the stomach, complications. All these parameters, plus American Society of Anesthesiology (ASA) risk class [5], were recorded by a physician not involved in the endoscopic procedure.

The primary end point was measured using a modified Freeman score [6], as follows: 1 - one to five attempts (easy); 2 - six to fifteen attempts (moderately difficult); 3 - more than fifteen attempts (difficult); 4 - success with the help of a guide wire; 5 - success after precut; and 6 - failure. The Schutz score [7] was used to assess the difficulty of the various ERCP procedures ([Table 1]). Grades 1 and 3 did not feature in the present study because diagnostic procedures were not included. Complications were classified according to the criteria described by Cotton et al. [8].

As a rule, after recognition of the papilla the trainee was allowed 10 minutes to achieve deep cannulation of the desired duct. If he/she failed after this time, the tutor took over and all the parameters, with the exception of time needed to visualize the papilla and total procedure time were recorded again for the new operator.

Statistical analysis

Data were collected and analyzed with a dedicated statistical computer program (Statistical Package for the Social Sciences, SPSS 13.0). Differences between groups were evaluated using the Χ² test for parametric data and Student’s t-test for nonparametric data. A sample size of 112 patients (28 for each of the 4 groups, prone/supine, expert/trainee) was calculated in order to demonstrate a statistically significant difference between groups, assuming a β value equal to 0.80 and a probability value of less than 0.05.

Results

One hundred and twenty patients were enrolled in the study and randomized to undergo ERCP in the supine or prone position by a senior endoscopist or a trainee between December 2005 and May 2006. Demographic data and indications for ERCP are listed in [Table 2].

Visualization of the papilla was successful in 100 % of cases in the tutor group and 95 % of those in the trainee group (P = n. s.). No differences were recorded between the prone and the supine groups.

There were no statistically significant differences in the mean difficulty in obtaining deep cannulation of the desired duct, neither between the prone and supine groups, nor between the tutor and trainee groups ([Table 3]). In the trainee group the tutor had to take over in 34 out of 60 examinations because of failure to identify the papilla (n = 3), failure to opacify the desired duct (n = 16), and failure to achieve deep cannulation (n = 15). The overall success rate in the trainee group was 43.3 %. However, switching to the tutor was independent of patient position ([Table 4]). Tutors succeeded in all cases where trainees had failed. Therefore the overall success rate at the first procedure was 99.2 % (119/120 patients). In one patient in the supine group, who had with cancer of the pancreatic head, cholangiography and cannulation failed even after needle-knife precut, due to neoplastic invasion of the duodenal wall. The procedure was successfully completed 2 days later by the same operator with the patient in the same position.

There were no statistically significant differences between the prone and the supine groups for all the remaining recorded parameters ([Table 5]). All the procedures were started and finished under conscious sedation (fentanyl and midazolam) without the need to switch to deep sedation (propofol).

Two cases of post-ERCP acute pancreatitis (1.7 %) were recorded (one moderate and one severe according to Cotton’s criteria). Both patients recovered with conservative management.

Discussion

This study was designed to evaluate the patient’s position as a determinant of ERCP outcome in terms of success rate and complications. The prone and left lateral positions are generally preferred for patients undergoing ERCP. Both these positions are believed to decrease the risk of aspiration pneumonia and to permit easier introduction of the scope through the pharynx, and they are usually considered by the endoscopists to be more comfortable. Endoscopy textbooks [2] [3] suggest starting the procedure with the patient in a semilateral position and continue with the patient prone once the duodenum has been reached. The supine position might have some practical advantages in patients who require general anesthesia with tracheal intubation, and also in patients undergoing deep sedation, because it allows better control of the airways, easier saliva suction, and quicker resuscitation maneuvers. In our day-to-day practice, we prefer the supine position for examining patients with suspected hilar biliary strictures, those with previous Billroth II gastrectomy, and those whose pancreatic duct anatomy is already known to be difficult to interpret. The supine position allows better interpretation of the biliary and pancreatic anatomy: with the liver and the pancreas lying on the spine under the force of gravity, a more natural display of the anatomical details is achieved; moreover, superimposition of the spine, duodenoscope, or bowel gas on the biliary-pancreatic system can be easily overcome by gently turning the patient’s trunk in either direction. With the patient in the prone position, but also in the left lateral position, this can also be achieved to some extent by rotation of a C-arm X-ray apparatus [9].

We compared the prone and supine positions for therapeutic ERCP as performed both by experts and by trainees. No differences could be found in respect of deep cannulation of the desired duct, duration of the various procedural steps, or complications. These results are in contrast to those of the only published study that has addressed this issue [4]. The authors of that study, Terruzzi et al., prematurely stopped the enrollment of patients due to the strikingly lower cannulation success rate (70 % vs. 100 %) and the higher number of cardiovascular and respiratory adverse events (41 % vs. 6 %) in the supine group. Moreover, the Freeman difficulty score was significantly lower in the prone group. The different premedication (pethidine) and dose of midazolam used by Terruzzi et al. may also explain the difference in adverse cardiorespiratory events encountered in the two studies.

The remarkable discrepancy between the two studies could be explained by the different experience of the two centers in the use of the supine position for patients undergoing ERCP. Terruzzi et al. [4] admit that less than 5 % of ERCPs in their unit are performed with the patient supine, all under general anesthesia. By contrast, counting just patients with hilar strictures or previous Billroth II gastrectomy, more than 10 % of the ERCPs in our unit are routinely performed in supine patients. This figure rises above 20 - 25 % if patients examined under general anesthesia with orotracheal intubation, those with difficult pancreatic anatomy, and those unable to stay prone because they have fresh surgical scars or drains are added.

One limitation of the earlier study was the small number of patients enrolled, which permits a type II error. Thanks to that study, the sample size of this study was calculated to capture any statistically significant difference between groups in terms of cannulation difficulty, with an α value of 0.05 and a β value of 0.80. Another limitation of the previous study was that only patients with biliary indications were included. In our study patients with pancreatic indications were also analyzed, and therefore our results might apply to all patients requiring therapeutic ERCP.

Selection of the easier procedure and the more healthy patient could be a problem encountered during randomization. Probably this did not happen in our study because the levels of difficulty (Schutz scores) of the procedures undertaken were similar in both patient groups (prone/supine) for both sets of operators (see [Table 5]).

The lack of enrolled ASA 4 patients could be another selection bias, but these patients cannot be randomized because they usually have to remain supine owing to advanced age and comorbidities.

According to the results of this study, patient position is not a limiting factor per se, provided the supine position is not used only as an exception. Results of therapeutic ERCP in supine patients are no different from those in prone patients, whether performed by expert endoscopists or by trainees under tutorial control. The supine position seems to be safe, but our study is underpowered to check for differences in the incidence of acute cardiorespiratory complications, and therefore the supine position cannot be recommended in daily practice, especially in less specialized centers with a low case volume of procedures. A learning process, including a few tricks such as adjusting the endoscopist’s position to improve the steadiness of the duodenoscope in front of the papilla, is needed to attain competence in ERCP in the supine patient.

Competing interests: None

References

• 1 Taylor A J, Bohorfoush III (eds) A G. Interpretation of ERCP. Philadelphia; Lippincott-Raven 1997
  Search in Google Scholar
• 2 Yakshe P N, Vennes J A. Technique of endoscopic retrograde cholangiopancreatography. In: Sivak MV Jr (ed). Gastroenterologic endoscopy. 2nd edn., chap 57. Philadelphia; Saunders 2000: 845-862
  Search in Google Scholar
• 3 Leung J. Fundamentals of ERCP. In: Leung J, Cotton PB (eds). ERCP Advanced Endoscopy E-book/Annual. 2004 http://www.gastrohep.com/advancedendo/book3.asp
  Search in Google Scholar
• 4 Terruzzi V, Radaelli F, Meucci G, Minoli G. Is the supine position as safe and effective as the prone position for endoscopic retrograde cholangiopancreatography? A prospective randomized study.  Endoscopy. 2005;  37 1211-1214
  Thieme ConnectPubMedSearch in Google Scholar
• 5 Owens W D, Felts J A, Spitznagel E L. ASA physical status classification: a study of consistency of ratings.  Anesthesiology. 1978;  43 239-243
  PubMedSearch in Google Scholar
• 6 Freeman M L, Nelson D B, Sherman S. et al . Complications of endoscopic biliary sphincterotomy.  N Engl J Med. 1996;  335 909-919
  CrossrefPubMedSearch in Google Scholar
• 7 Schutz S M, Abbott R M. Grading ERCPs by degree of difficulty: a new concept to produce more meaningful outcome data.  Gastrointest Endosc. 2000;  51 535-539
  CrossrefPubMedSearch in Google Scholar
• 8 Cotton P B, Lehman G, Vennes J. et al . Endoscopic sphincterotomy complications and their management: an attempt at consensus.  Gastrointest Endosc. 1991;  37 383-393
  CrossrefPubMedSearch in Google Scholar
• 9 Froehlich F. Patient position during ERCP: prone versus supine. What about the left lateral throughout?.  Endoscopy. 2006;  38 755
  Thieme ConnectPubMedSearch in Google Scholar

G. Costamagna, MD 

Digestive Endoscopy Unit
Università Cattolica del Sacro Cuore
“A. Gemelli” University Hospital

Largo A. Gemelli 8
00168 Rome
Italy

Fax: +39-06-30156581

Email: gcostamagna@rm.unicatt.it

References

• 1 Taylor A J, Bohorfoush III (eds) A G. Interpretation of ERCP. Philadelphia; Lippincott-Raven 1997
  Search in Google Scholar
• 2 Yakshe P N, Vennes J A. Technique of endoscopic retrograde cholangiopancreatography. In: Sivak MV Jr (ed). Gastroenterologic endoscopy. 2nd edn., chap 57. Philadelphia; Saunders 2000: 845-862
  Search in Google Scholar
• 3 Leung J. Fundamentals of ERCP. In: Leung J, Cotton PB (eds). ERCP Advanced Endoscopy E-book/Annual. 2004 http://www.gastrohep.com/advancedendo/book3.asp
  Search in Google Scholar
• 4 Terruzzi V, Radaelli F, Meucci G, Minoli G. Is the supine position as safe and effective as the prone position for endoscopic retrograde cholangiopancreatography? A prospective randomized study.  Endoscopy. 2005;  37 1211-1214
  Thieme ConnectPubMedSearch in Google Scholar
• 5 Owens W D, Felts J A, Spitznagel E L. ASA physical status classification: a study of consistency of ratings.  Anesthesiology. 1978;  43 239-243
  PubMedSearch in Google Scholar
• 6 Freeman M L, Nelson D B, Sherman S. et al . Complications of endoscopic biliary sphincterotomy.  N Engl J Med. 1996;  335 909-919
  CrossrefPubMedSearch in Google Scholar
• 7 Schutz S M, Abbott R M. Grading ERCPs by degree of difficulty: a new concept to produce more meaningful outcome data.  Gastrointest Endosc. 2000;  51 535-539
  CrossrefPubMedSearch in Google Scholar
• 8 Cotton P B, Lehman G, Vennes J. et al . Endoscopic sphincterotomy complications and their management: an attempt at consensus.  Gastrointest Endosc. 1991;  37 383-393
  CrossrefPubMedSearch in Google Scholar
• 9 Froehlich F. Patient position during ERCP: prone versus supine. What about the left lateral throughout?.  Endoscopy. 2006;  38 755
  Thieme ConnectPubMedSearch in Google Scholar

G. Costamagna, MD 

Digestive Endoscopy Unit
Università Cattolica del Sacro Cuore
“A. Gemelli” University Hospital

Largo A. Gemelli 8
00168 Rome
Italy

Fax: +39-06-30156581

Email: gcostamagna@rm.unicatt.it