Search tips
Search criteria 


Logo of jigLink to Publisher's site
J Interv Gastroenterol. 2013 Jan-Mar; 3(1): 37–42.
PMCID: PMC3791540
Carbon dioxide insufflation versus air insufflation during endoscopic retrograde cholangiopancreatography: a meta-analysis
Jun Wu and Bing Hucorresponding author
Department of Endoscopy, Eastern Hepatobiliary Hospital, the Second Militar Medical University, Shanghai, China
corresponding authorCorresponding author.
Correspondence to: Bing HU; Email: drhubing/at/
Received November 12, 2012; Revised March 13, 2013; Accepted March 13, 2013.
Background and aims
Adequate visualization during endoscopic retrograde cholangiopancreatography (ERCP) procedure requires distention of the bowel lumen, usually insufflated with room air. Patients often complain of abdominal pain post-procedure. The use of carbon dioxide insufflation in colonoscopy has been shownto result in less post-procedure abdominal pain and distension. Recently, it has been reported the use of carbon dioxide (CO2) insufflation during ERCP procedure is similarly helpful. The purpose of this study was to evaluate the efficiency, safety and comfort of ERCP involving carbon dioxide insufflation through a meta-analysis of published randomized control trials.
Databases including PubMed, EMBASE, the Cochrane Library, the Science Citation Index and momentous meeting abstracts were searched and evaluated by two reviewers independently.
Five randomized control trials involving 446 patients were analyzed. Meta-analysis showed that patientsin the CO2 insufflation group had lower pain score (VAS) at 1-hour [MD −12.37, 95%CI(−20.96,−3.78)], 3-hours [MD −9.81, 95%CI (−17.05, −2.57)) and 6-hours [MD −8.78, 95%CI (−13.71, −3.85)] compared with air insufflation group after procedure. However, there were no significant differences between the two groups regardingtotal procedure time and procedure complication.
Insufflation with carbon dioxide during ERCP may decrease post-procedure abdominal discomfort without any additional adverse reactions. Thus, CO2 insufflation seems appropriate to use during ERCP procedure. Large trials are required to prove any additional advantages to carbon dioxide insufflation during ERCP.
Key words: carbon dioxide, endoscopic retrograde cholangiopancreatography, meta-analysis
Gas insufflation is necessary to obtain optimal visualization during endoscopic retrograde cholangiopancreatography (ERCP). Patients frequently complain of abdominal pain and discomfort after procedures. While these symptoms occasionally herald procedure complications such as pancreatitis or perforation,1 more often these symptoms are related to bowel distension caused by air insufflation. The utility of carbon dioxide insufflation instead of air insufflation during endoscopic procedures was first reported for colonoscopy.2 Subsequent studies have examined the safety and efficacy of carbon dioxide (CO2) insufflation as compared with that of air and most of these studies showed that the use of CO2 insufflation leads to less post procedure abdominal pain and distension.35 This finding is thought to due to faster absorption of CO2 by intestinal mucosa. Recently, studies67 assesse the role of CO2 insufflation during ERCP found that patients also had less pain and distention with CO2 insufflation when compared with air insufflation.
We believe an updated systemic review and meta-analysis would help solidify the conclusion that CO2 insufflation during ERCP results in less post-procedure abdominal pain and distension.
Study identification and eligibility criteria
A comprehensive literature search was done to identify all relevant studies that compared CO2 with air insufflation during ERCP. The PubMed, EMBASE, the Cochrane Library and the Science Citation Index were searched systematically for all articles published up to June 2012, without language restriction, which included the following terms in their titles, abstracts, or keywords lists: “carbon dioxide”, “CO2”, “endoscopic retrograde cholangiopancreatography”, “ERCP”. The references in retrieved articles were also screened manually. The abstracts of United European Gastroenterology week (UEGW) and Digestive Disease Week (DDW) were also searched systematically.
Papers selected from this initial search were then screened for eligibility using the following criteria: (i) RCT study that compared CO2 insufflation with air insufflation in patients undergoing ERCP procedure; (ii) outcome measures: abdominal pain and procedure safety. Any comment, review, or guideline articles were excluded.
Data extraction
Two reviewers (J.W and B.H) abstracted data independently and reached consensus on all items. Data were extracted on: first author; year of publication; country of origin; number of centers; number of patients; age and sex; indication for ERCP; sedation; CO2 delivery system; The measures included abdominal pain scores (VAS) after procedure(1 h, 3 h, 6 h, 24 h); safety of CO2 insufflation; total procedure time and complications of procedures.
Assessment of methodological quality
Study quality was assessed by a component approach.8 Three key methodological domains of internal validity that have empirically been shown to be associated with bias in estimates of treatment effect were assessed: allocation concealment, double blinding of outcome assessment and intention-to-treat analysis. We considered an allocation sequence to be adequate if random number tables, computer-generated random numbers, or minimization was mentioned in the report. Sealed, opaque, sequentially numbered assignment envelopes, central randomization, and on-site computerized randomization system were considered to be adequate methods of allocation concealment. Analysis was assumed if the reported number of participants who were randomized and the number who were entered into the analysis were identical.
Statistical analysis
Statistical manipulation was performed with Review Manager Software (Version 5.1, Cochrane Collaboration, UK). Values for analysis were extracted from published reports or calculated from crude data. For summary statistics in meta-analyses, the Relative Risk (RR) is recommended for dichotomous data, and the Mean Difference (MD) is recommended for continuous data. Pooled estimates were presented with a 95% confidence interval. p<0.05 was considered significant. A sensitivity analysis9 determined how the results would be influenced if one study was removed from the analysis for each outcome; This indicates the extent to which the results are(or are not) robust to assumptions and decisions that were made when carrying out the synthesis. Statistical heterogeneity between trials was assessed by calculating I2 measure of inconsistency. Generally, an I2of 0% to 40% excludes heterogeneity, I2 of 30% to 60% may represent moderate heterogeneity, I2 of 50% to 90% may represent substantial heterogeneity, and I2 of 75% to 100% represents considerable heterogeneity. In the presence of statistical heterogeneity, a random effect model was used for the analysis. In the absence of statistically heterogeneity, only the RR by the fixed effect model is given in the results.
Identification of eligible studies
The literature search yielded 55 abstracts for review, and 50 abstracts were excluded because of the reasons shown in Figure 1. Finally, Five RCTs67,1012 were included, all of which were available as full text. The combined studies enrolled 446 patients, and randomized them to either CO2 group or air group. The characteristics of included trials were listed in Table 1. Sedation during ERCP was used in all studies. One study7 only used Propofol for sedation, another 4 studies6,1012 used multiple sedatives; CO2 delivery system was mentioned in all trials. The indication for ERCP included choledocholithiasis, benign or malignant biliary strictures, biliary dilation, bile leak and others. All studies excluded patients with respiratory diseases.
Figure 1
Figure 1
Flow diagram of trials selection. RCT, randomized controlled trials.
Table 1
Table 1
Characteristics of studies included in the meta-analysis
Effect of CO2 insufflation on abdominal pain
All trials assessed abdominal pain related to ERCP procedures. Researchers assessed this measure by means of a linear analogue pain scale (visual analogue scale, VAS) ranging from 0 to 100mm, with 0 being no pain at all and 100 being the worst pain imaginable. The baseline VAS of patients were recorded in 4 trials67,10,12, and there was no significant difference between 2 groups. The VAS score were recorded after procedures at different points (1 h, 3 h, 6 h, 24 h). Owing to the statistical heterogeneity between trials (p<0.1), we used a random effect model for the meta-analysis. The results showed that CO2 insufflation was associated with significant decrease in pain scores for patients at 1-hour [MD −12.37, 95%CI(−20.96,−3.78)], 3-hours [MD −9.81, 95%CI (−17.05, −2.57)) and 6-hours [MD −8.78, 95%CI (−13.71, −3.85)] (Fig.2). There was no significant difference between the two groups at 24-hours (p=0.62) after procedures (Fig.2).
Figure 2
Figure 2
Forrest plot showing patients had lower pain score(VAS) in the CO2 insufflation group at 1-hour [MD −12.37, 95%CI (−20.96, −3.78)]; 3-hours [MD −9.81, 95%CI (−17.05, −2.57)]; 6-hours [MD −8.78, 95%CI (more ...)
Safety of CO2 insufflation
Three RCTs6,10,12 evaluated the safety of CO2 insufflation (Table 2). Two studies6,10 evaluated the safety with transcutaneous pCO2 and one12 with PaCO2 measures. Baseline pCO2 levels of the two groups in these 3 studies were equivalent. There was no significant difference in PCO2 level between the CO2 and air groups during procedure and post-procedure. Luigiano12 reported that patients with high peak values of PaCO2(>55 mmHg) in the CO2 group were more than in the air group, but this high value was easily reduced by increasing ventilation. Bretthauer6 reported that a rise in pCO2 levels was observed in both groups during the procedure, and this rise was more apparent in the air group, with pCO2 level slightly above the reference range during the procedure. But there were no significant differences between the two groups. There were no adverse respiratory complications reported in all of these studies.
Table 2
Table 2
Safety of CO2 insufflation
Procedure time and complication
All studies reported the total procedure time in both groups. Meta-analysis showed no significant difference in procedure time between CO2 and air insufflation groups (p=0.31) (Fig.3). All studies reported procedure related complications. Metaanalysis showed that complications in the CO2 group were not significantly different from those in the air group (p=0.10) (Fig.4). There were no adverse respiratory complications, and the most common complications associated with ERCP were pancreatitis and bleeding. All complications were treated conservatively.
Figure 3
Figure 3
Meta-analysis showed total procedure time used in ERCP showing no significant difference between two groups (p=0.31).
Figure 4
Figure 4
Forrest plot showing RR with 95%CI for complications during procedure with CO2 versus air. There was no significant difference between 2 groups (p=0.10).
Sensitivity analysis and study quality assessment
We conducted sensitivity analysis by excluding one study (Evan S.Dellon/201010) to see how the results would influence the pooled estimate. Sensitivity analysis showed that CO2 insufflation results in reduced abdominal pain scores [MD −17.05, 95%CI (−25.49,−8.6)] 1-hour post-procedure. We had further sensitivity analysis by restricting the studies to high-quality studies (Jadad score=5) and multiple sedative (n≥2). Restricting the studies to these parameters did not alter the results. Data were also analyzed by random effects. Sensitivity analysis showed that the results were robust for the CO2 insufflation with significantly reduced abdominal pain. Measures of allocation concealment were described in all studies (100%). All of the studies were double-blinded. Only one11 of the five studies (20%) had an intention-totreat analysis without missing data (Table 3).
Table 3
Table 3
Methodological quality of studies included in the meta-analysis
ERCP examinations can be prolonged, and there may be greater volumes of gas insufflated to maintain visualization of the bowel lumen. Usually air is used to distend the bowel lumen, but air is not well absorbed and must either be suctioned or passed through the bowel as flatus. Carbon dioxide has the benefit of rapid absorption through the intestinal mucosa into the bloodtbefore it is eliminated via respiration. The goal of this study was to establish the safety and efficacy of CO2 insufflation during ERCP through meta-analysis deserve careful evaluation.
While Carbon dioxide is a common metabolic product that usually can be eliminated from body through respiration. If all products accumulate, CO2 can cause acidosis. Several studies35 confirmed the safety of CO2 insufflation during colonoscopy, but a study13 involving endoscopic submucosal dissection (ESD) in the colon found a slightly increase in the PaCO2 when insufflated with CO2. Three studies included in our meta-analysis assessed safety of CO2 use by monitoring transcutaneous pCO2 and PaCO2 during and after endoscopic procedures, and no significant differences were found between the two groups. The pCO2 levels of most patients in three studies remained in normal range. Dellon10 reported more than half of all patients in both the CO2 and air groups developed CO2 retention probably because this may be that sedation affects respiration function, causing retention of CO2. This CO2 retention was easily treated by increasing the ventilation.
Not surprisingly, meta-analysis demonstrated that the total procedure duration in CO2 group was comparable with its counterpart in all studies. There were no adverse respiratory complications, and most complications associated with ERCP were pancreatitis and bleeding. Complications in the CO2 group were not significantly different from that in the air group, and all complications were treated conservatively.
Several limitations of the present study need to be considered. First, there was significant heterogeneity for all major pain related analyses. The source of heterogeneity may also include the different publishing time of papers, the research object, the characteristics of patients etc. Though the data was treated with random effect models, there was still some influence to final results. Secondly, only three trials6,10,12 evaluated the safety of CO2 insufflation, and all trials excluded patients with respiratory disease which could introduce bias into our meta-analysis. Thirdly, most RCTs were involved relatively small samples and a single center, which mightinfluence the results. Fourthly, sedation in most studies was provided with multiple sedatives, so some drugs may influence the patient's feeling post-procedure and cause some bias to our meta-analysis.
In conclusion, the meta-analysis reveals that, compared with room air, CO2 likely decreases symptom of abdominal pain and discomfort, without affecting procedure duration and procedure-related complications. Large trials are required to prove the conclusions the effectiveness of carbon dioxide insufflation during ERCP in future.
ERCPendoscopic retrograde cholangiopancreatography
CO2carbon dioxide
UEGWUnited European Gastroenterology week
DDWDigestive Disease Week
VASvisual analogue scale

Previously published online:
1. Masci E, Toti G, Mariani A, Curioni S, Lomazzi A, Dinelli M, et al. Complications of diagnostic and therapeutic ERCP: a prospective multicenter study. Am J Gastroenterol. 2001;96:417–423. [PubMed]
2. Hussein AM, Bartram CI, Williams CB. Carbon dioxide insufflation for more comfortable colonoscopy. Gastrointest Endosc. 1984;30:68–70. [PubMed]
3. Stevenson GW, Wilson JA, Wilkinson J, Norman G, Goodacre RL. Pain following colonoscopy: elimination with carbon dioxide. Gastrointest Endosc. 1992;38:564–567. [PubMed]
4. Sumanac K, Zealley I, Fox BM, Rawlinson J, Salena B, Marshall JK, et al. Minimizing post colonoscopy abdominal pain by using CO(2) insufflation: a prospective, randomized, double blind, controlled trial evaluating a new commercially available CO(2) delivery system. Gastrointest Endosc. 2002;56:190–194. [PubMed]
5. Bretthauer M, Thiis-Evensen E, Huppertz-Hauss G, Gisselsson L, Grotmol T, Skovlund E, et al. NORCCAP (Norwegian colorectal cancer prevention): a randomized trial to assess the safety and efficacy of carbon dioxide versus air insufflation in colonoscopy. Gut. 2002;50:604–607. [PMC free article] [PubMed]
6. Bretthauer M, Seip B, Aasen S, Kordal M, Hoff G, Aabakken L. Carbon dioxide insufflation for more comfortable endoscopic retrograde cholangiopancreatography: a randomized, controlled, double-blind trial. Endoscopy. 2007;39:58–64. [PubMed]
7. Maple JT, Keswani RN, Hovis RM, Saddedin EZ, Jonnalagadda S, Azar RR, et al. Carbon dioxide insufflation during ERCP for reduction of postprocedure pain: a randomized, double-blind, controlled trial. Gastrointest Endosc. 2009;70:278–283. [PubMed]
8. Mazaki T, Ebisawa K. Enteral versus parenteral nutrition after gastrointestinal surgery: a systematic review and meta-analysis of randomized controlled trials in the English literature. J Gastrointest Surg. 2008;12:739–755. [PubMed]
9. Sterne JA, Egger M, Smith GD. Systematic reviews in health care: Investigating and dealing with publication and other biases in meta-analysis. BMJ. 2001;323:101–105. [PMC free article] [PubMed]
10. Dellon ES, Velayudham A, Clarke BW, Isaacs KL, Gangarosa LM, Galanko JA, et al. A randomized, controlled, double-blind trial of air insufflation versus carbon dioxide insufflation during ERCP. Gastrointest Endosc. 2010;72:68–77. [PMC free article] [PubMed]
11. Kuwatani M, Kawakami H, Hayashi T, Ishiwatari H, Kudo T, Yamato H, et al. Carbon dioxide insufflation during endoscopic retrograde cholangiopancreatography reduces bowel gas volume but does not affect visual analogue scale scores of suffering: a prospective, double-blind, randomized, controlled trial. Surg Endosc. 2011;25:3784–3790. [PubMed]
12. Luigiano C, Ferrara F, Pellicano R, Fabbri C, Cennamo V, Bassi M, et al. Carbon dioxide insufflation versus air insufflation during endoscopic retrograde cholangiopancreatography under general anesthesia. Minerva Med. 2011;102:261–269. [PubMed]
13. Saito Y, Uraoka T, Matsuda T, Emura F, Ikehara H, Mashimo Y, et al. A pilot study to assess the safety and efficacy of carbon dioxide insufflation during colorectal endoscopic submucosal dissection with the patient under conscious sedation. Gastrointestinal Endosc. 2007;65:537–542. [PubMed]
Articles from Journal of Interventional Gastroenterology are provided here courtesy of
Landes Bioscience