Carbon dioxide versus room air insufflation in colonoscopy: a comparative study

With increasing awareness for colon cancer screening, colonoscopy has become the mainstay in detecting precancerous and cancerous lesions [1]. Currently there are two widely accepted means of insufflation, room air (RA) and carbon dioxide (CO2). CO2 as a means of insufflation in laparoscopic surgery has been used for decades [2]. In endoscopy, the use of CO2, has not been widely used. Majority of the endoscopists continue to utilize RA as supplied by the manufacturer. CO2 insufflation requires additional equipment for its utilization. In a study by Janssens et al. in 2009, less than 5% of their study respondents utilized CO2 for luminal distention [3], lower than a previous study conducted in the United States, which reported that only 13% of colonoscopies applied CO2 [4]. A large proportion of those respondents stated that they were either not aware of the ability to use CO2, while others stated challenges in the implementation of equipment [3].


Introduction
With increasing awareness for colon cancer screening, colonoscopy has become the mainstay in detecting precancerous and cancerous lesions [1]. Currently there are two widely accepted means of insufflation, room air (RA) and carbon dioxide (CO 2 ). CO 2 as a means of insufflation in laparoscopic surgery has been used for decades [2]. In endoscopy, the use of CO 2 , has not been widely used. Majority of the endoscopists continue to utilize RA as supplied by the manufacturer. CO 2 insufflation requires additional equipment for its utilization. In a study by Janssens et al. in 2009, less than 5% of their study respondents utilized CO 2 for luminal distention [3], lower than a previous study conducted in the United States, which reported that only 13% of colonoscopies applied CO 2 [4]. A large proportion of those respondents stated that they were either not aware of the ability to use CO 2 , while others stated challenges in the implementation of equipment [3].
One of the major sources of discomfort for patients undergoing colonoscopy is the volume of gas, leading to abdominal distention and increased pain. Air absorption within the colonic lumen is poor. However, the use of CO 2 , can lead to quick absorption intraluminally into the blood stream and exhaled [5]. Due to its rapid absorption, it might be logical to assume that a larger volume of CO 2 would be required, which may mitigate some of its advantages, yet a study conducted by Bretthauer et al. (2003) indicated that similar volumes of RA and CO 2 were utilized by experienced endoscopists [6]. Many studies have demonstrated that the ability of CO 2 to be absorbed by the body reduces intra and post procedural abdominal distention [7] and pain [8], which is important for patient compliance in follow up procedures.
The intent of the study was to analyze the effects of CO 2 compared to RA insufflation amongst endoscopists at our centre with respect to endoscopy time, sedation required, post-procedure pain and recovery time.

Study arms
Four study arms were utilized for this study. The first of the groupings (scenario 1) compared the cases of the primary endoscopist at our centre (physician A) prior to January 29 th (n=112), to the other 6 physicians (physicians B-G) practicing at our facility utilizing RA (n=114). The objective of this arm was to assess the baseline results of all physicians prior to the universal implementation of CO 2 ; this allowed for the detection of inter-practitioner variabilities that may influence the results. The second grouping (scenario 2) consisted of a comparison between all cases conducted by physician A (prior to Jan 29 th & after Jan 29 th ) (n=209) and all cases conducted by the other 6 physicians (utilizing RA and CO 2 ) (n=211). Similarly, this scenario was utilized to detect differences in practice between the main endoscopist and the remainder of the physicians, and thus allowed for the assessment of inter-practitioner variability. The third grouping (scenario 3) compared cases of the other 6 physicians utilizing RA (n=112) versus those that employed CO 2 (n=118); this was used as a direct comparison of RA and CO 2 in practitioners who are new to its use. Scenario 3, allowed for the detection of differences, specifically in those who are new to utilizing CO 2 . The last grouping (scenario 4) compared all cases utilizing CO 2 (n=333) versus all those that utilized RA (n=112), which allowed again for a direct comparison of RA and CO 2 . However, this scenario was utilized to detect differences between the methods that could be generalized to a broader range of physicians and practices.

Patient selection
Patients selected underwent complete colonoscopies utilizing either CO 2 or RA insufflation in the gastrointestinal department at the Brandon Regional Health Centre, a regional referral centre in Brandon, Manitoba, Canada. Patients completed a standard patient consent form for colonoscopy. The study was reviewed and approved by the Gastrointestinal Endoscopy department administration at Brandon Regional Health Centre. All colonoscopies were scheduled and were performed between November 28, 2014 and May 29, 2015.This regional referral site conducts over 3800 colonoscopies each year; the main endoscopist at this site implemented CO 2 into his practice several years ago, however the other physicians only recently began utilizing it for luminal distention on January 29 th . Physicians included in this study were either general surgeons or gastroenterologists. Physician A and D have greater than 25 years' experience in colonoscopy; physicians' B and G have greater than 20 years' experience; physician F has greater than 15 years, physician C has greater than 10 years and physician E has 3 years' experience in colonoscopy. Consecutive patients were chosen from each of the study arms, defined above, to be representative of the number of patients scoped by each of the physicians over a 1-year period. Table 1 outlines the percentage of total colonoscopies performed by each of the physicians at our site over the most recent year.

Exclusion criteria
Patients were excluded from the study due to the following: previous large bowel resection, scheduled for endoscopic mucosal resection, incomplete colonoscopy, incomplete recording of data parameters, addition of gastroscopy and those done on an emergent basis.

Parameters assessed
The charts of all patients selected were reviewed and the required data was extracted and recorded. Patient demographic data was obtained to ensure that all groups were similar with respect to age, gender (male: female ratio) and body mass index (BMI) (calculated utilizing the formula mass/height 2 , where mass is in kilograms and height is in centimeters) ( Table 2). It should be noted that both height and weight were self-reported by the patient and were not directly measured by the nurse upon patient admission to the Gastrointestinal Department.
Intra-procedure measurements assessed included amount of sedation utilized, oxygen (O 2 ) requirement by nasal prong and endoscopy time. Within the test facility the two medications that are employed for patient sedation during colonoscopies are fentanyl and midazolam. Endoscopy time was measured from the time the scope was introduced, to the time the scope was completely withdrawn.
Post-procedure the patients were assessed for the length of the recovery time and the presence of post procedure pain. For this study, recovery time was defined as the time from admission to the recovery area to time of discharge. The requirements that patients needed to meet prior to discharge were return to pre-procedural baseline of vital signs and sedation/responsiveness. Post-procedure pain was assessed by the nursing staff at three intervals, on admission to recovery, 15 and 30 minutes after admission to recovery. Due to variation in nurse interpretation of pain, for the purposes of this study pain was noted only as present or absent at the intervals.

Data analysis
Data from all three of the previously mentioned scenarios was analyzed utilizing 2 sample unpaired t-tests on Minitab statistical analysis software by the authors to assess for any differences in means between the two samples involved in each scenario. For each of the scenarios the following parameters were assessed utilizing 2 sample unpaired t-tests: age, gender, BMI, amount of fentanyl and midazolam, O 2 requirement, endoscopy time, and recovery time. Post-procedure pain was assessed more thoroughly for each of the scenarios. Pain was assessed in the following ways: overall difference in pain experienced by patients, difference in pain experienced by gender, as well as difference in pain at the assessed time intervals, both overall and by gender.

Discussion
The use of colonoscopy is the main method of detection of colon polyps and cancer. Many individuals are reluctant to undergo colonoscopy due to their perception of pain associated with the procedure. CO 2 has been suggested as a means of reducing patient discomfort for several years [4,5]. A barrier to the implementation of any intervention in health care is the cost of it use. However, CO 2 is relatively inexpensive, in a study conducted by Wong et al. in 2008 calculated that its use cost less than 0.84 euros extra per patient when compared to the use of RA [9]. The use of CO 2 has also been found to be safe for use with electrosurgical instruments, as such it is a viable method for use in all colonoscopic procedures [10].
Within all scenarios, the patient profile was similar between all groups. Intra-procedure O 2 requirement did not change with the introduction of CO 2 insufflation, thus indicating that the respiratory drive of patients did not change. CO 2 is likely safe in all patients, as O 2 requirements were not increased, and patients with chronic lung disease were not excluded. However, the data for end-tidal CO 2 (ETCO 2 ) was not assessed in this study as it was not readily available on all patients. To confirm these findings, it would be necessary to study the ETCO 2 of patients undergoing colonoscopy. Previous studies assessing ETCO 2 have been mixed with some demonstrating an increase, while others demonstrating no change in levels [11,12].
Endoscopy time was largely dependent on the individual endoscopist. Physicians new to the use of CO 2 did not significantly increase their endoscopy time when compared to their baseline. Physicians adapted quickly to the use of CO 2 method of insufflation.
The amount of administration of medications appear to be largely physician dependent during the procedure. Some physicians prefer to utilize higher amounts of midazolam, while others prefer fentanyl. The use of CO 2 did not alter significantly their prescribing practices, however, the impression by the nursing staff is that physicians utilize less medication with CO 2 . At our institution, nursing staff are under the impression that patients who underwent colonoscopy with CO 2 had shorter recovery times. However, the difference was not significant when analyzing the data. Many confounding factors may have skewed the data, including patients waiting for their family to pick them up, lab work that needed to be completed or waiting to speak with their endoscopist. Discharge criteria utilized by the unit was not altered after the implementation of CO 2 . Some nursing staff were still waiting for patients to pass flatus and were monitoring for bowel sounds. As CO 2 is readily absorbed into the bloodstream and expired, these criteria may no longer be appropriate. There was no difference in recovery time between patients undergoing colonoscopy with or without biopsy versus polypectomy. A difference in recovery time was also postulated by Belle et al., but they were also unable to detect a difference in their study [7]. They attributed this inconsistency to differences between clinical assessment done by nursing staff and guidelines [7].
Overall, pain was improved with the use of CO 2 . This can be explained by the mechanism of clearance of the CO 2 , as it can be rapidly absorbed which means less abdominal distention and thus less discomfort. A study conducted by Belle et al., demonstrated less abdominal distention with CO 2 by measuring abdominal girth before and after colonoscopy [7]. Similarly, two randomized control trials demonstrated minimal colonic gas in the large intestine was discovered by abdominal radiograph 1-hour post-procedure with CO 2 (94% CO 2 vs. 2% RA), and minimal residual gas in the small intestine (87% CO 2 vs. 55% RA) [13,14]. In our study, significantly less patients experienced pain at 15 minutes in to recovery, but at 30 minutes this difference did not reach statistical significance. Studies have demonstrated that the benefits of CO 2 with respect to pain can last from 6 to 24 hours [8,15]. Matyja et al. found that CO 2 in unsedated colonoscopy did not reduce pain immediately or 15 minutes after the procedure, but a slightly lower pain intensity was observed 60 minutes after the procedure, although did not reach statistical significance [16]. Procedures conducted in their study were also very short (an average of 11 minutes each), postulating that in shorter procedures the use of CO 2 did not impact the post-procedure pain [16].
Overall decreased pain is seen in both genders in those undergoing colonoscopy with CO 2 . The level of physician experience is an important factor in predicting patient discomfort, with patients experiencing less pain during and after the procedure was noted in our study.

Limitations
Due to the small sample size and ability to measure pain, which is subjective in nature, the extent of the benefits of CO 2 over RA insufflation in post-procedure discomfort are difficult to assign a metric especially in a sedated patient with different indications for colonoscopy. Data from multiple endoscopists, with different skill sets and training, were included which may have influenced the outcomes. Lastly, as data was collected from charts, data directly from the patients could not be obtained.

Future directions
Further studies should focus on assessing the extent to which pain is decreased in patients when CO 2 is utilized, as well as to investigate intra-procedure discomfort experienced by patients. The advantage of CO 2 should also be assessed in gastroscopy and endoscopic retrograde cholangiopancreatography.

Conclusion
Data obtained in this study suggests that utilizing CO 2 for insufflation leads to a better overall patient experience, with less discomfort during and after the procedure. With less discomfort, patients may be more likely to present for follow up procedures, furthermore cost saving could be realized with earlier discharge as suggested by clinical impression from nursing staff. It also suggests that physicians do not need to significantly change their practice, with respect to medication administration and endoscopy time.

Author contributions
Both authors contributed equally to following: conception and design; analysis and interpretation of the data; drafting of the article; critical revision and final approval of the article