- Email: [email protected]
The use of air or CO2 as insufflation agents for double contrast barium enema (DCBE): Is there a qualitative difference?
Clinical Radiology (1995) 50, 558-561
The Use of Air or CO2 as Insufflation Agents for Double Contrast Barium Enema (DCBE): Is there a Qualitative Difference? D. A. SCULLION, C. W. N. WETTON, C. DAVIES, L. WHITAKER* and P. J. SHORVON
Department of Diagnostic Radiology, Central Middlesex Hospital, London and *Department of Public Health, St Mary's Hospital, London, UK One hundred consecutive out-patients referred for double contrast barium enema (DCBE) were randomized in a double-blind prospective study to receive either air or CO2 as an insuffiation agent. Each examination was performed by the same radiologist and radiographer. The two groups were equally matched for age and sex. Each completed examination was independently assessed by three radiologists for a variety of parameters. The results show that the patients receiving air as an insufflation agent had better overall colonic distension than the group receiving CO2 and that this was statistically significant (P = 0.0004). There was no significant difference in the mean time taken to perform the examination in the two groups nor was there any significant difference in mucosal coating or preparation. In conclusion, our results suggest that poor colonic distension is a potential problem when CO2 is used for DCBE and that this cannot be solely attributed to the time taken to perform the examination. Poor distension could lead to diagnostic errors and this may outweigh the advantages in patient acceptability when using CO2 as an insufllation agent. Scullion, D.A., Wetton, C.W.N., Davies, C., Whitaker, L. & Shorvon, P.J. (1995). Clinical Radiology 50, 558-561. The Use of Air or CO2 as Insufflation Agents for Double Contrast Barium Enema (DCBE): Is there a Qualitative Difference?
Accepted for Publication 28 March 1995
Double contrast barium enema (DCBE) is now universally accepted as the method of Choice for routine barium investigation of the colon. Traditionally air is used as an insufflation agent following infusion of barium suspension and in most cases the use of a muscle relaxant such as hyoscine-butylbromide (Buscopan, Boehringer Ingelheim). The mucosal detail provided by this two-stage technique is far superior to the previously used single contrast studies. The single most important disadvantage of this technique is the pain experienced by patients after the study has been completed. This is thought to be due to colonic and small bowel distension which may produce colic as the muscle relaxant begins to wear off. Although the procedure itself is clearly not a comfortable one, it is of interest that patients may complain of significant abdominal pain some hours after the examination has been completed. In the search for a more patient-acceptable procedure, CO2 has been proposed as an alternative insufflation agent to air. This gas has the advantage over air in that it is absorbed some 35 times faster from the gastrointestinal tract than nitrogen, which is the major component of room air [1]. The absorbed gas is rapidly eliminated in the lungs so that any metabolic effect is negligible and short-lived. The use of CO2 would hopefully minimize some of the more unpleasant delayed effects of DCBE and indeed previous studies have shown this to be the case [2-4]. In view of this recent published evidence we have recently switched to the use of CO2 as the routine insufflation agent for DCBE. However, it was our initial impression that the degree of distension achieved with COz was less than that seen Correspondence to: Dr D.A. Scullion, Department of Diagnostic Radiology, St Mary's Hospital, Praed Street, London W2 INY, UK. 9 1995 Blackwell Science Limited
when air was used for insufflation. Although the recent literature has suggested that there is no difference in the quality of the study when CO2 is substituted for air, there seems to be no firm data to support this statement and in view of this and our initial observations we elected to perform a randomized prospective trial to test this hypothesis. In the past the use of CO2 was largely hampered by the complexity of the delivery system. More recently a simple method of introduction of CO2 via a closed system has been described [5] and this was the method used in our study. It is no more timeconsuming than using air.
PATIENTS AND M E T H O D S
One hundred consecutive out-patients referred for routine DCBE were randomly assigned to receive either air or CO2 as an insufflation agent. There was no lower age limit and an upper age limit of 80 years was arbitrarily chosen for inclusion in the study. A suspension of barium was made up in a plastic container with attached tubing. The clamp on the plastic tubing was then released and the tubing pushed firmly over the nozzle of a pressurized CO2 cylinder if the patient had been assigned to receive CO2. The bag was then inflated to capacity with CO2 and the tubing clamped and removed. This provided a closed system of gas that should prove sufficient for most enemas. The gas was insufltated by gentle foot pressure. The examination was performed on all occasions by the same radiologist and radiographer, both of whom were blind to the type of gas used. Each DCBE was carried out to a set protocol. The degree of anal tone was assessed and graded by the
559
THE USE OF AIR OR CO2 AS INSUFFLATION AGENTS FOR DCBE
radiologist following digital rectal examination. Buscopan (20 mg iv) was given to all patients when the head of the barium column reached the transverse colon. The gas was insufllated and spot films were taken of the rectum, sigmoid colon, splenic flexure, transverse colon, hepatic flexure and caecum. After the spot films were completed the radiologist insufflated more gas if it was felt that colonic distension was sub-optimal at this stage. Overcouch views were then taken by the radiographer and consisted of a supine film, angled prone (Hampton's) view, left lateral decubitus and fight lateral decubitus view. The overcouch films were taken in the same order on each occasion. The supine film was the first film taken on each occasion. If further overcouch films were required for technical reasons, this was also recorded. Each examination was carefully timed. The start time, time that Buscopan was given, time of first insufflation, time of starting overcouch films and time of finishing overcouch films were all recorded. It was occasionally necessary to give a second bag of the original gas if adequate distension had not been achieved following the first bag. Exchange could be achieved quickly and efficiently and any excess gas requirement was also recorded. The films were reviewed by three radiologists independently. The supine overcouch film was specifically chosen for review of a number of parameters and for measurements of colonic diameter at pre-determined points. The parameters chosen were distension, which was graded as optimal, good, poor or non-diagnostic, mucosal coating, graded as optimal, adequate and poor, and bowel preparation, also graded as optimal, adequate and poor. Each parameter was given a numerical grading of 1-3 (Fig. 1). The scores from each observer were added to give a single score for each parameter for each patient. No supine overcouch film was considered to be so poor as to be non-diagnostic. The degree of mucosal crinkling was also noted and recorded as being either absent, some mucosal crinkling or severe mucosal crinkling. This was also graded numerically 1-3. In addition each observer was asked to assign a numerical value to their diagnostic confidence for each examination on a scale of 1-10. The higher the value the more confident the observer was of making a definitive diagnosis on the basis of the barium investigation. Lastly, each observer was asked to decide whether the study had been performed using air or CO2.
Table 1 - Clinical details of study patients
No. Mean age (range) Sex (M/F)
Air
CO 2
49 57.8 (23-79) 22/27
45 52.2 (21-79) 22/23
examination in the air group was 15.5 (15) min and in the group receiving CO2 16.3 (16) min (P = 0.13, MannWhitney U-test). This is the time taken from the first introduction of barium to completion of the last overcouch film. The time taken to perform the overcouch films was also measured. This we have called the postinsufflation time. The mean (median) post-insufflation time in the group receiving air was 6.2 (6.0) min and 6.3 (6.0) min in the group receiving CO2 (P = 0.9, MannWhitney U-test). There was no significant difference in anal tone between the two groups. When assessing distension, coating and bowel preparation, the sum of the individual scores of all three radiologists for each parameter was used in the statistical analysis, giving a minimum score of 3 and a maximum of 9 for each. There was no statistical difference in the mucosal coating or bowel preparation for the two groups (Table 2). The mean (median) score for distension in the group receiving air for insufflation was 7.6 (8.0) and that of the group receiving CO2 was 5.9 (6.0), and this was highly statistically significant (P = 0.0004, Mann-Whitney U-test). The total number of patients for each numerical value of distension has also been tabulated (Table 3). The overall time taken to perform the investigation was negatively correlated with the degree of colonic distension (R = - 0 . 2 7 , P + 0.0092, Spearman rank correlation), although there was no significant negative correlation between time and distension when the two groups were analysed individually. This is possibly due to the sample size in each group. Mucosal crinkling was more marked in the group receiving CO2 than in the group receiving air and this too was statisTable 2 - Mean (median) score for each parameter in both air and CO2 groups Air
CO 2
(Mann- Whitney U-test)
RESULTS Six patients were excluded from the study, four because extremely poor bowel preparation precluded completion of the study and two in whom a tight colonic stricture prevented further passage of barium. All patients received Buscopan (20 mg iv). Of the 94 patients examined, 45 received CO2 and 49 received air. The two groups were equally matched for age and sex (Table 1). The mean age of all patients was 55 years (range 21-79 years). The mean (median) time taken to perform the DISTENSION
COATING
3 2 1 0
3 2 1 0
= Optimal = Good = Poor = Non-diagnostic
e~
= = = =
Good Adequate Poor Non-diagnostic
PREPARATION 3 = Good 2 = Adequate 1 = Poor 0 = Non-diagnostic
Fig. 1 - Numerical grading system for distension, coating and preparation. 9 1995 Blackwell Science Ltd, Clinical Radiology, 50, 558-561.
Distension Preparation Mucosal coating Mucosal crinkling
7.6 7.5 7.7 3.9
(8.0) (9.0) (9.0) (4.0)
5.9 8.2 8.0 4.8
(6.0) (9.0) (9.0) (5.0)
0.0004 0.16 0.36 0.008
Table 3 - Number of patients with a given total value of distension in both the air and CO2 groups Distension value
Air
C02
3 4 5 6 7 8 9
1 2 5 6 4 7 24
10 6 4 7 4 3 11
560
CLINICAL RADIOLOGY
Table 4 - Rank correlation of diagnostic confidence against distension,
coating and preparation
Distension Preparation Mucosal coating
Spearman Rank Correlation
P =
0.53 0.49 0.52
0.0001 0.0001 0.0001
tically significant (P = 0.008, Mann-Whitney U-test). Overall there was no significant difference in the degree of diagnostic confidence between the two groups. However when diagnostic confidence was correlated individually with distension, mucosal coating and bowel preparation there was a statistically significant correlation with each individual parameter (Table 4). A statistical analysis of inter-observer variation was performed. There was random variation between all observers in terms of confidence level. However when individual parameters such as distension, coating and preparation were assessed there was good agreement between all observers. DISCUSSION For many years barium enema has been the main investigation of disorders of the large bowel. Initially a single contrast study was used but this gradually gave way to the DCBE, using air as an insufflation agent, which is now well established as the main radiological investigation of colonic disorders. There is no doubt that the greater mucosal detail afforded by the doublecontrast technique allows for the detection and characterization of smaller lesions with subsequent increase in diagnostic accuracy. Despite the relatively recent increased use of colonoscopy, this technique is still not universally available and the DCBE still provides a significant proportion of the workload in most radiology departments. Serious complications of the technique are fortunately rare but much attention has been drawn to patient discomfort due to distension of the colon with gas. This can be partly alleviated during the examination by using a smooth muscle relaxant such as Buscopan. Nevertheless patients still complain of significant abdominal pain long after the effect of Buscopan should have worn off, which we postulate may be due to residual colonic gaseous distension. In an effort to improve pain following DCBE, attention has been drawn to the use of CO 2 as an insufltation agent rather than air. The rationale for its use is that CO 2 is absorbed from the large bowel some 35 times faster than nitrogen, the main component of room air. The less gas remaining in the colon after the procedure, the less likely is the patient to experience significant colonic spasm and abdominal discomfort. Refinements of a CO2 delivery system have now made CO2 as easy to use as room air. An early study [6] suggested that pain following DCBE was an idiosyncratic response and that there was no significant difference between air and CO 2. However, numbers in this study were relatively small and patients were only asked about symptoms up to 30 min following the examination. Later studies referred to above used larger numbers of patients and included questions about delayed pain. These concluded that symptoms were less frequent and severe when C02 was
used as the insufflation agent. Some of the previous studies [2-4] have commented that there was no significant difference in colonic distension regardless of the type of gas used. Only one of these studies would appear to have attempted to assess this critically. However this was not the prime purpose of these studies and the results are certainly at variance with ours. Moreover, Marshall et al., comparing the use of CO 2 and air for flexible proctosigmoidoscopy performed on the same day as DCBE, showed that there was less residual colonic gas at the start of the enema when CO2 was used for proctosigmoidoscopy [7]. It might be argued, therefore, that using an easily diffusable gas such as CO2 may well produce enemas of inferior quality and the results of our study bear this out. The use of the same radiographer and radiologist and a standard enema protocol minimized any variation in the technical aspects of the procedure. We did not routinely insuttlate gas between each overcouch film as we felt this would cause unnecessary discomfort in those patients with adequate colonic distension at this stage of the examination. However we have assumed that at the end of insufflation all colons were equally well distended. The difference in time between the two groups does not account for the observed differences in colonic distension and an average time of between 15 and 16 min to perform each examination is perfectly acceptable. Neither was there any statistical difference in the post-insufflation time between the two groups. This lends weight to the hypothesis that the differences in distension are largely due to the increased rate of diffusion of CO2 as compared with air. It is possible that other factors, such as colonic surface area and vascularity and the effects of Buscopan, may also affect gaseous diffusion, but this is difficult to validate. More detailed analysis of our results did not show a particular post-insufflation time after which distension was judged to be significantly poor, again possibly due to the size of the sample or to large variations between individuals. There is no reason why preparation and mucosal coating should differ significantly in the two groups as all patients were given the same instructions for bowel preparation. The significance of mucosal crinkling is not known. This is a phenomenon that has been described elsewhere [8] and seems to be distinct from colonic spasm. It is thought to be due to the muscularis mucosa being thrown into fine transverse folds. It is possible that this is an early feature of loss of colonic distension and is more prevalent in the CO2 group because of the earlier diffusion of gas. Despite the differences in distension there was no overall difference in diagnostic confidence between the two groups. This is because the overall degree of diagnostic confidence depends on a variety of factors, both technical and interpretative, and one individual may be more sensitive to one of these factors than another observer. This may account for the differences in confidence levels between observers. The results do show, however, that when each parameter is examined in turn, there is a strong correlation between it and the level of diagnostic confidence. This makes optimal colonic distension particularly important when the degree ot coating and preparation are less than perfect. In summary, therefore, our results conflict with those of previous studies and show that there is a significant 9 1995 Blackwell Science Ltd, Clinical Radiology, 50, 558-561.
THE USE OF AIR OR
CO 2
AS INSUFFLATION AGENTS FOR DCBE
difference in colonic distension when C O 2 is used as an insufflation agent and that this may have a bearing on both the quality of the examination and the diagnostic confidence of the radiologist. We believe that the differences can be accounted for by the rapid absorption of CO 2 when compared with air. We are unable, from our results, to give a cut-off time to help determine which insufflation agent should be used or to show a significant negative correlation between time and distension when the two groups were analysed individually. We would, however, stress that distension is negatively correlated with the time taken to perform the examination when the whole patient group is analysed. We agree that there is good evidence that C02 has advantages in terms of patient acceptability, and certainly in some of our patients optimal colonic distension was achieved using this agent, but would suggest that in frail or immobile patients where it is anticipated that the examination will be prolonged, consideration should be given to using air rather than CO 2 to optimize diagnostic confidence.
9 1995 BlackwellScienceLtd, ClinicalRadiology, 50, 558-561.
561
REFERENCES
1 Forster RE. Physiological basis of gas exchange in the gut. Annals of the NY Academy of Science 1968;150:4-12. 2 Taylor PN, Beckly DE. Use of air in double-contrast barium enema - is it still acceptable? Clinical Radiology 1991;44:183-184. 3 Robson NK, Lloyd M, Regan F. The use of carbon dioxide as an insufflation agent in barium enema - does it have a role? British Journal of Radiology 1993;66:197-198. 4 Coblentz CL, Frost RA, Molinaro Vet aL Pain after barium enema: effect of CO2 and air on double-contrast study. Radiology 1985;157: 35-36. 5 Bartram CI. Technical note: a simple method for using carbon dioxide during double contrast barium enema. Clinical Radiology 1989;40:318. 6 Grant DS, Bartram CI, Heron CW. A preliminary study of the possible benefits of using carbon dioxide insut~ation during doublecontrast barium enema. British Journal of Radiology 1986;59:190191. 7 Marshall JB, Hoyt TS, Seger RM et al. Air-contrast barium enema studies after flexible proctosigmoidoscopy: randomised controlled clinical trial. Radiology 1990;176:549-551. 8 Bartram CI, Mootoosamy IM, Lira KH. Washout versus nonwashout (Picolax) preparation for double-contrast barium enemas. Clinical Radiology 1984;35:143-146.
The Use of Air or CO2 as Insufflation Agents for Double Contrast Barium Enema (DCBE): Is there a Qualitative Difference? D. A. SCULLION, C. W. N. WETTON, C. DAVIES, L. WHITAKER* and P. J. SHORVON
Department of Diagnostic Radiology, Central Middlesex Hospital, London and *Department of Public Health, St Mary's Hospital, London, UK One hundred consecutive out-patients referred for double contrast barium enema (DCBE) were randomized in a double-blind prospective study to receive either air or CO2 as an insuffiation agent. Each examination was performed by the same radiologist and radiographer. The two groups were equally matched for age and sex. Each completed examination was independently assessed by three radiologists for a variety of parameters. The results show that the patients receiving air as an insufflation agent had better overall colonic distension than the group receiving CO2 and that this was statistically significant (P = 0.0004). There was no significant difference in the mean time taken to perform the examination in the two groups nor was there any significant difference in mucosal coating or preparation. In conclusion, our results suggest that poor colonic distension is a potential problem when CO2 is used for DCBE and that this cannot be solely attributed to the time taken to perform the examination. Poor distension could lead to diagnostic errors and this may outweigh the advantages in patient acceptability when using CO2 as an insufllation agent. Scullion, D.A., Wetton, C.W.N., Davies, C., Whitaker, L. & Shorvon, P.J. (1995). Clinical Radiology 50, 558-561. The Use of Air or CO2 as Insufflation Agents for Double Contrast Barium Enema (DCBE): Is there a Qualitative Difference?
Accepted for Publication 28 March 1995
Double contrast barium enema (DCBE) is now universally accepted as the method of Choice for routine barium investigation of the colon. Traditionally air is used as an insufflation agent following infusion of barium suspension and in most cases the use of a muscle relaxant such as hyoscine-butylbromide (Buscopan, Boehringer Ingelheim). The mucosal detail provided by this two-stage technique is far superior to the previously used single contrast studies. The single most important disadvantage of this technique is the pain experienced by patients after the study has been completed. This is thought to be due to colonic and small bowel distension which may produce colic as the muscle relaxant begins to wear off. Although the procedure itself is clearly not a comfortable one, it is of interest that patients may complain of significant abdominal pain some hours after the examination has been completed. In the search for a more patient-acceptable procedure, CO2 has been proposed as an alternative insufflation agent to air. This gas has the advantage over air in that it is absorbed some 35 times faster from the gastrointestinal tract than nitrogen, which is the major component of room air [1]. The absorbed gas is rapidly eliminated in the lungs so that any metabolic effect is negligible and short-lived. The use of CO2 would hopefully minimize some of the more unpleasant delayed effects of DCBE and indeed previous studies have shown this to be the case [2-4]. In view of this recent published evidence we have recently switched to the use of CO2 as the routine insufflation agent for DCBE. However, it was our initial impression that the degree of distension achieved with COz was less than that seen Correspondence to: Dr D.A. Scullion, Department of Diagnostic Radiology, St Mary's Hospital, Praed Street, London W2 INY, UK. 9 1995 Blackwell Science Limited
when air was used for insufflation. Although the recent literature has suggested that there is no difference in the quality of the study when CO2 is substituted for air, there seems to be no firm data to support this statement and in view of this and our initial observations we elected to perform a randomized prospective trial to test this hypothesis. In the past the use of CO2 was largely hampered by the complexity of the delivery system. More recently a simple method of introduction of CO2 via a closed system has been described [5] and this was the method used in our study. It is no more timeconsuming than using air.
PATIENTS AND M E T H O D S
One hundred consecutive out-patients referred for routine DCBE were randomly assigned to receive either air or CO2 as an insufflation agent. There was no lower age limit and an upper age limit of 80 years was arbitrarily chosen for inclusion in the study. A suspension of barium was made up in a plastic container with attached tubing. The clamp on the plastic tubing was then released and the tubing pushed firmly over the nozzle of a pressurized CO2 cylinder if the patient had been assigned to receive CO2. The bag was then inflated to capacity with CO2 and the tubing clamped and removed. This provided a closed system of gas that should prove sufficient for most enemas. The gas was insufltated by gentle foot pressure. The examination was performed on all occasions by the same radiologist and radiographer, both of whom were blind to the type of gas used. Each DCBE was carried out to a set protocol. The degree of anal tone was assessed and graded by the
559
THE USE OF AIR OR CO2 AS INSUFFLATION AGENTS FOR DCBE
radiologist following digital rectal examination. Buscopan (20 mg iv) was given to all patients when the head of the barium column reached the transverse colon. The gas was insufllated and spot films were taken of the rectum, sigmoid colon, splenic flexure, transverse colon, hepatic flexure and caecum. After the spot films were completed the radiologist insufflated more gas if it was felt that colonic distension was sub-optimal at this stage. Overcouch views were then taken by the radiographer and consisted of a supine film, angled prone (Hampton's) view, left lateral decubitus and fight lateral decubitus view. The overcouch films were taken in the same order on each occasion. The supine film was the first film taken on each occasion. If further overcouch films were required for technical reasons, this was also recorded. Each examination was carefully timed. The start time, time that Buscopan was given, time of first insufflation, time of starting overcouch films and time of finishing overcouch films were all recorded. It was occasionally necessary to give a second bag of the original gas if adequate distension had not been achieved following the first bag. Exchange could be achieved quickly and efficiently and any excess gas requirement was also recorded. The films were reviewed by three radiologists independently. The supine overcouch film was specifically chosen for review of a number of parameters and for measurements of colonic diameter at pre-determined points. The parameters chosen were distension, which was graded as optimal, good, poor or non-diagnostic, mucosal coating, graded as optimal, adequate and poor, and bowel preparation, also graded as optimal, adequate and poor. Each parameter was given a numerical grading of 1-3 (Fig. 1). The scores from each observer were added to give a single score for each parameter for each patient. No supine overcouch film was considered to be so poor as to be non-diagnostic. The degree of mucosal crinkling was also noted and recorded as being either absent, some mucosal crinkling or severe mucosal crinkling. This was also graded numerically 1-3. In addition each observer was asked to assign a numerical value to their diagnostic confidence for each examination on a scale of 1-10. The higher the value the more confident the observer was of making a definitive diagnosis on the basis of the barium investigation. Lastly, each observer was asked to decide whether the study had been performed using air or CO2.
Table 1 - Clinical details of study patients
No. Mean age (range) Sex (M/F)
Air
CO 2
49 57.8 (23-79) 22/27
45 52.2 (21-79) 22/23
examination in the air group was 15.5 (15) min and in the group receiving CO2 16.3 (16) min (P = 0.13, MannWhitney U-test). This is the time taken from the first introduction of barium to completion of the last overcouch film. The time taken to perform the overcouch films was also measured. This we have called the postinsufflation time. The mean (median) post-insufflation time in the group receiving air was 6.2 (6.0) min and 6.3 (6.0) min in the group receiving CO2 (P = 0.9, MannWhitney U-test). There was no significant difference in anal tone between the two groups. When assessing distension, coating and bowel preparation, the sum of the individual scores of all three radiologists for each parameter was used in the statistical analysis, giving a minimum score of 3 and a maximum of 9 for each. There was no statistical difference in the mucosal coating or bowel preparation for the two groups (Table 2). The mean (median) score for distension in the group receiving air for insufflation was 7.6 (8.0) and that of the group receiving CO2 was 5.9 (6.0), and this was highly statistically significant (P = 0.0004, Mann-Whitney U-test). The total number of patients for each numerical value of distension has also been tabulated (Table 3). The overall time taken to perform the investigation was negatively correlated with the degree of colonic distension (R = - 0 . 2 7 , P + 0.0092, Spearman rank correlation), although there was no significant negative correlation between time and distension when the two groups were analysed individually. This is possibly due to the sample size in each group. Mucosal crinkling was more marked in the group receiving CO2 than in the group receiving air and this too was statisTable 2 - Mean (median) score for each parameter in both air and CO2 groups Air
CO 2
(Mann- Whitney U-test)
RESULTS Six patients were excluded from the study, four because extremely poor bowel preparation precluded completion of the study and two in whom a tight colonic stricture prevented further passage of barium. All patients received Buscopan (20 mg iv). Of the 94 patients examined, 45 received CO2 and 49 received air. The two groups were equally matched for age and sex (Table 1). The mean age of all patients was 55 years (range 21-79 years). The mean (median) time taken to perform the DISTENSION
COATING
3 2 1 0
3 2 1 0
= Optimal = Good = Poor = Non-diagnostic
e~
= = = =
Good Adequate Poor Non-diagnostic
PREPARATION 3 = Good 2 = Adequate 1 = Poor 0 = Non-diagnostic
Fig. 1 - Numerical grading system for distension, coating and preparation. 9 1995 Blackwell Science Ltd, Clinical Radiology, 50, 558-561.
Distension Preparation Mucosal coating Mucosal crinkling
7.6 7.5 7.7 3.9
(8.0) (9.0) (9.0) (4.0)
5.9 8.2 8.0 4.8
(6.0) (9.0) (9.0) (5.0)
0.0004 0.16 0.36 0.008
Table 3 - Number of patients with a given total value of distension in both the air and CO2 groups Distension value
Air
C02
3 4 5 6 7 8 9
1 2 5 6 4 7 24
10 6 4 7 4 3 11
560
CLINICAL RADIOLOGY
Table 4 - Rank correlation of diagnostic confidence against distension,
coating and preparation
Distension Preparation Mucosal coating
Spearman Rank Correlation
P =
0.53 0.49 0.52
0.0001 0.0001 0.0001
tically significant (P = 0.008, Mann-Whitney U-test). Overall there was no significant difference in the degree of diagnostic confidence between the two groups. However when diagnostic confidence was correlated individually with distension, mucosal coating and bowel preparation there was a statistically significant correlation with each individual parameter (Table 4). A statistical analysis of inter-observer variation was performed. There was random variation between all observers in terms of confidence level. However when individual parameters such as distension, coating and preparation were assessed there was good agreement between all observers. DISCUSSION For many years barium enema has been the main investigation of disorders of the large bowel. Initially a single contrast study was used but this gradually gave way to the DCBE, using air as an insufflation agent, which is now well established as the main radiological investigation of colonic disorders. There is no doubt that the greater mucosal detail afforded by the doublecontrast technique allows for the detection and characterization of smaller lesions with subsequent increase in diagnostic accuracy. Despite the relatively recent increased use of colonoscopy, this technique is still not universally available and the DCBE still provides a significant proportion of the workload in most radiology departments. Serious complications of the technique are fortunately rare but much attention has been drawn to patient discomfort due to distension of the colon with gas. This can be partly alleviated during the examination by using a smooth muscle relaxant such as Buscopan. Nevertheless patients still complain of significant abdominal pain long after the effect of Buscopan should have worn off, which we postulate may be due to residual colonic gaseous distension. In an effort to improve pain following DCBE, attention has been drawn to the use of CO 2 as an insufltation agent rather than air. The rationale for its use is that CO 2 is absorbed from the large bowel some 35 times faster than nitrogen, the main component of room air. The less gas remaining in the colon after the procedure, the less likely is the patient to experience significant colonic spasm and abdominal discomfort. Refinements of a CO2 delivery system have now made CO2 as easy to use as room air. An early study [6] suggested that pain following DCBE was an idiosyncratic response and that there was no significant difference between air and CO 2. However, numbers in this study were relatively small and patients were only asked about symptoms up to 30 min following the examination. Later studies referred to above used larger numbers of patients and included questions about delayed pain. These concluded that symptoms were less frequent and severe when C02 was
used as the insufflation agent. Some of the previous studies [2-4] have commented that there was no significant difference in colonic distension regardless of the type of gas used. Only one of these studies would appear to have attempted to assess this critically. However this was not the prime purpose of these studies and the results are certainly at variance with ours. Moreover, Marshall et al., comparing the use of CO 2 and air for flexible proctosigmoidoscopy performed on the same day as DCBE, showed that there was less residual colonic gas at the start of the enema when CO2 was used for proctosigmoidoscopy [7]. It might be argued, therefore, that using an easily diffusable gas such as CO2 may well produce enemas of inferior quality and the results of our study bear this out. The use of the same radiographer and radiologist and a standard enema protocol minimized any variation in the technical aspects of the procedure. We did not routinely insuttlate gas between each overcouch film as we felt this would cause unnecessary discomfort in those patients with adequate colonic distension at this stage of the examination. However we have assumed that at the end of insufflation all colons were equally well distended. The difference in time between the two groups does not account for the observed differences in colonic distension and an average time of between 15 and 16 min to perform each examination is perfectly acceptable. Neither was there any statistical difference in the post-insufflation time between the two groups. This lends weight to the hypothesis that the differences in distension are largely due to the increased rate of diffusion of CO2 as compared with air. It is possible that other factors, such as colonic surface area and vascularity and the effects of Buscopan, may also affect gaseous diffusion, but this is difficult to validate. More detailed analysis of our results did not show a particular post-insufflation time after which distension was judged to be significantly poor, again possibly due to the size of the sample or to large variations between individuals. There is no reason why preparation and mucosal coating should differ significantly in the two groups as all patients were given the same instructions for bowel preparation. The significance of mucosal crinkling is not known. This is a phenomenon that has been described elsewhere [8] and seems to be distinct from colonic spasm. It is thought to be due to the muscularis mucosa being thrown into fine transverse folds. It is possible that this is an early feature of loss of colonic distension and is more prevalent in the CO2 group because of the earlier diffusion of gas. Despite the differences in distension there was no overall difference in diagnostic confidence between the two groups. This is because the overall degree of diagnostic confidence depends on a variety of factors, both technical and interpretative, and one individual may be more sensitive to one of these factors than another observer. This may account for the differences in confidence levels between observers. The results do show, however, that when each parameter is examined in turn, there is a strong correlation between it and the level of diagnostic confidence. This makes optimal colonic distension particularly important when the degree ot coating and preparation are less than perfect. In summary, therefore, our results conflict with those of previous studies and show that there is a significant 9 1995 Blackwell Science Ltd, Clinical Radiology, 50, 558-561.
THE USE OF AIR OR
CO 2
AS INSUFFLATION AGENTS FOR DCBE
difference in colonic distension when C O 2 is used as an insufflation agent and that this may have a bearing on both the quality of the examination and the diagnostic confidence of the radiologist. We believe that the differences can be accounted for by the rapid absorption of CO 2 when compared with air. We are unable, from our results, to give a cut-off time to help determine which insufflation agent should be used or to show a significant negative correlation between time and distension when the two groups were analysed individually. We would, however, stress that distension is negatively correlated with the time taken to perform the examination when the whole patient group is analysed. We agree that there is good evidence that C02 has advantages in terms of patient acceptability, and certainly in some of our patients optimal colonic distension was achieved using this agent, but would suggest that in frail or immobile patients where it is anticipated that the examination will be prolonged, consideration should be given to using air rather than CO 2 to optimize diagnostic confidence.
9 1995 BlackwellScienceLtd, ClinicalRadiology, 50, 558-561.
561
REFERENCES
1 Forster RE. Physiological basis of gas exchange in the gut. Annals of the NY Academy of Science 1968;150:4-12. 2 Taylor PN, Beckly DE. Use of air in double-contrast barium enema - is it still acceptable? Clinical Radiology 1991;44:183-184. 3 Robson NK, Lloyd M, Regan F. The use of carbon dioxide as an insufflation agent in barium enema - does it have a role? British Journal of Radiology 1993;66:197-198. 4 Coblentz CL, Frost RA, Molinaro Vet aL Pain after barium enema: effect of CO2 and air on double-contrast study. Radiology 1985;157: 35-36. 5 Bartram CI. Technical note: a simple method for using carbon dioxide during double contrast barium enema. Clinical Radiology 1989;40:318. 6 Grant DS, Bartram CI, Heron CW. A preliminary study of the possible benefits of using carbon dioxide insut~ation during doublecontrast barium enema. British Journal of Radiology 1986;59:190191. 7 Marshall JB, Hoyt TS, Seger RM et al. Air-contrast barium enema studies after flexible proctosigmoidoscopy: randomised controlled clinical trial. Radiology 1990;176:549-551. 8 Bartram CI, Mootoosamy IM, Lira KH. Washout versus nonwashout (Picolax) preparation for double-contrast barium enemas. Clinical Radiology 1984;35:143-146.