Gastrointest Radiol 17:81-83 (1992)

Gastrointestinal

Radiology 9 Springer-Verlag New York Inc. 1992

Comparison Between Double-Contrast Barium Enema and Colonoscopy to Investigate Lower Gastrointestinal Bleeding Edgar Jaramillo and Premysl Slezak Department of Diagnostic Radiology, Karolinska Hospital, Stockholm, Sweden

Abstract. A retrospective study was performed to c o m p a r e the diagnostic a c c u r a c y of high-quality double-contrast barium e n e m a (DCBE) against gold standard c o l o n o s c o p y in 288 patients with suspected lower gastrointestinal bleeding who went through both examinations. C o l o n o s c o p y detected the potential cause of bleeding in 99 patients (100%); in order of frequency: polyps -> 1 cm (N = 47; 48%), c a r c i n o m a (N = 21; 21%), inflammatory bowel disease (IBD) (N = 15; 15%), solitary ulcers (N = 6; 6%), other types of colitis (N = 5; 5%), angiodysplasia (N = 3; 3%), and stenosis (N = 2; 2%). D C B E diagnosed 88 cases (89%) and missed 11 consisting of I B D (N = 4), angiodysplasia (N = 3), solitary ulcers (N = 3), and polyps (N = 1). The overall sensitivity and specificity of D C B E was 0.89 and 0.97, respectively. The sensitivity for carcinoma, polyps, and I B D was 1.00, 0.98, and 0.73, respectively. We conclude that D C B E is very effective to diagnose c a r c i n o m a and polyps >-- ! cm, the most frequent causes of bleeding, but less effective to diagnose I B D and other nonfrequent causes. If a highquality D C B E does not reveal the cause of bleeding, the contribution of a following c o l o n o s c o p y will be to diagnose causes o f bleeding other than carcinoma and polyps < 1 cm and to offer therapeutic possibilities.

Key words: Rectal b l e e d i n g - - H e m o c c u l t - - C o l o n o s c o p y - - F l e x i b l e s i g m o i d o s c o p y - - B a r i u m enema.

The aim of this study was to clarify the contribution of c o l o n o s c o p y in patients with suspected lower gasAddress offprint requests to: Premysl Slezak, Ph.D., Department

of Diagnostic Radiology, Karolinska Hospital, P.O. Box 60 500, S-104 01 Stockholm, Sweden

trointestinal bleeding in which double-contrast barium e n e m a (DCBE) did not reveal the source of bleeding. When the physical examination and rigid proctosigmoidoscopy have not been of help in revealing the cause of lower gastrointestinal bleeding, there is today no a g r e e m e n t on the best diagnostic approach. Single contrast barium e n e m a has been generally abandoned and replaced by D C B E [1]. D C B E is accepted b y some as the first choice [2-4]. Combining flexible sigmoidoscopy and D C B E gives even a higher diagnostic a c c u r a c y [5, 6], but unfortunately, as with c o l o n o s c o p y , it is not generally available. Although c o l o n o s c o p y can miss some lesions [4, 7], it is claimed to be superior to D C B E [8-10], sometimes based on suboptimal and unreviewed x-ray examinations [3] or randomized studies where both examinations were not p e r f o r m e d in the same patient. We wanted, in patients with suspected lower gastrointestinal bleeding, who had gone through both D C B E and c o l o n o s c o p y , to c o m p a r e the diagnostic a c c u r a c y of high-quality D C B E to colonoscopy with the aim to investigate if, after a negative D C B E , referring such patients for c o l o n o s c o p y could be of benefit.

Methods During the period 1983-1989, 6422 patients were referred to the Karolinska Hospital for colon x-ray. There were 1300 patients (20.6%) with suspected lower gastrointestinal bleeding. In 146 patients, the bleeding source was established with just a barium enema examination and no colonoscopy was performed; therefore, these patients were not included in the study. There were 324 patients who underwent DCBE followed by colonoscopy because of suspected lower gastrointestinal bleeding; 36 were excluded from the study due to reported inadequate DCBE (N = 22; 6.8%) or incomplete colonoscopy (N = 14; 4.3%). Finally, 288 patients were accepted for the study.

82

E. Jaramillo and P. Slezak: Barium Enema and Colonoscopy

Table 1. Comparison between DCBE and colonoscopy in patients with suspected lower gastrointestinal bleeding (N = 288) DCBE/endoscopy

Negative/negative

Negative/positive"

Positive/negative ~'

Positive/positive

Total" (%)

Polyp -> 1 cm Cancer IBD Radiation colitis Unspecified colitis Angiodysplasia Ulcer Stenosis

--------

1 -4 --3 3 --

2 -2 ----1

46 21 11 3 2 -3 2

47 (48%) 21 (21%) 15 (15%) 3 (3%) 2 (2%) 3 (3%) 6 (6%) 2 (2%)

Total (%)d

184 (64%)

11 (4%)

5 (2%)

88 (30%)

--

False negatives; included in total positives b False positives; not included in total positives c Refers to the total number of positive cases (N = 99) a Refers to the total number of analyzed cases (N = 288)

The inclusion criteria were: positive macroscopic or occult blood in the stools, iron-deficiency anemia, an interval of 6 months or less between DCBE and colonoscopy, an adequate DCBE as judged by an experienced radiologist, and a complete and reliable colonoscopy as judged and performed by an experienced endoscopist. Inadequate DCBE was defined as bad coating, improper cleaning of the colon, contrast artifacts due to flocculation, and difficulties to visualize and document all parts of the colon. Accepted potential causes of bleeding were: polyps - 1 cm, malignancy, different types of colitis and proctitis, ulcers, stenosis, and angiodysplasia. Hemorrhoids and diverticula were not accepted as potential causes of bleeding, because rather often there is a concomitant lesion in the colon of major importance [6, 11]. Polyps < 1 cm were also excluded [12]. All DCBE examinations and colonoscopies were performed after routine bowel preparation, starting with a liquid diet plus magnesium oxide and bisacodyl 2 days before the examination and completed (on the day of the examination) with saline enemas until proper cleaning was achieved. Premedication was given at the beginning of the colonoscopy procedure with 5-10 mg IV diazepam and 25-50 mg IV pethidin. No routine premedication was used for the DCBE. Both examinations were performed at our department. X-ray films were read routinely by at least two radiologists, one of them experienced in gastrointestinal radiology. Colonoscopies were performed by members of the gastrointestinal and radiological units. The endoscopists were aware of the DCBE results before the colonoscopy. Data analysis: Colonoscopy. if reported as complete and reliable, was taken as the gold standard. Two-by-two contingency tables were made to calculate the sensitivity, specificity, and positive and negative predictive value of the results.

Results

T h e total n u m b e r o f p a t i e n t s w a s 288, with a m e a n age o f 59 y e a r s ( 1 8 - 8 0 , r a n g e ) ; 158 (55%) f e m a l e s a n d 130 (45%) m a l e s . T h e m e a n i n t e r v a l b e t w e e n the D C B E a n d c o l o n o s c o p y w a s 6 w e e k s (0-26). I n 184 p a t i e n t s (64%), the c a u s e o f b l e e d i n g was n o t a p p a r e n t b y a n y of the two m e t h o d s (Table 1). F i v e p a t i e n t s (2%) w e r e initially r e p o r t e d as p o s i t i v e

after the D C B E , a finding that c o u l d not be confirmed b y c o l o n o s c o p y ; t h e r e f o r e , t h e y were c o n s i d ered as false p o s i t i v e s . I n 99 p a t i e n t s (34%), the pot e n t i a l c a u s e o f b l e e d i n g w a s e s t a b l i s h e d : 88 b y b o t h D C B E a n d c o l o n o s c o p y ; 11 o n l y b y c o l o n o s c o p y (false n e g a t i v e s ) . T h e c a u s e s of b l e e d i n g (N = 99) were: p o l y p s -> 1 c m (N = 47; 48%), c a r c i n o m a (N = 21; 21%), I B D (N = 15; 15%), solitary ulcers (N = 6; 6%), o t h e r t y p e s o f colitis (N = 5; 5%), a n g i o d y s p l a s i a (N = 3; 3%), a n d s t e n o s i s (N = 2; 2%). T h e d i a g n o s e s m i s s e d b y D C B E (N = 11) w e r e I B D (N = 4), a n g i o d y s p l a s i a (N = 3), solitary ulcers (N = 3), a n d p o l y s (N = 1). T h e false positives (N = 5) w e r e p o l y p s (N = 2), I B D (N = 2), a n d stenosis (N = 'l). F o r the D C B E , the o v e r a l l s e n s i t i v i t y a n d specificity w e r e 0.89 a n d 0.97, r e s p e c t i v e l y (Table 2). T h e p o s i t i v e a n d n e g a t i v e p r e d i c t i v e v a l u e were 0.95 a n d 0.94, r e s p e c t i v e l y . D u e t o the fact that D C B E failed to d i a g n o s e a p o l y p (1 o u t of 47) a n d o v e r d i a g n o s e d t w o cases, the s e n s i t i v i t y a n d specificity of D C B E to d e t e c t p o l y p s was 0.98 a n d 0.99, r e s p e c t i v e l y . N o m a l i g n a n c i e s were m i s s e d a n d there were n o false p o s i t i v e s , so the s e n s i t i v i t y a n d specificity of D C B E to d e t e c t c a r c i n o m a w e r e the s a m e as for c o l o n o s c o p y . O u t o f 15 c a s e s o f I B D , there were four false n e g a t i v e s a n d t w o false p o s i t i v e s , giving a s e n s i t i v i t y a n d specificity for D C B E of 0.73 a n d 0.99, r e s p e c tively.

Discussion

I n p a t i e n t s with s u s p e c t e d g a s t r o i n t e s t i n a l b l e e d i n g , D C B E w a s as good as c o l o n o s c o p y in the diagnosis o f c a n c e r a n d p o l y p s -> 1 cm, w h i c h in o u r s t u d y w e r e the m o s t f r e q u e n t c a u s e s o f b l e e d i n g (repre-

E. Jaramillo and P. Slezak: Barium Enema and Colonoscopy

83

Table 2. Sensitivity and specificity of DCBE in patients with suspected lower gastrointestinal bleeding

Patients with acute bleeding would probably benefit from initial colonoscopy if angiography or scintigraphy have not been of help or are not indicated [8, 15]. One should also remember that if colonoscopy is used as the first diagnostic choice but is incomplete, then the patient should be referred for DCBE. The selective use of DCBE and colonoscopy in the diagnosis of the cause of lower gastrointestinal bleeding leads to an effective utilization of re-

Sensitivity

All cases

Polyps Cancer

IBD

0.89 0.98 1.00 0.73

Specificity

0.97 0.99 1.00 0.99

Positive predictive value

Negative predictive

0.95 0.96 1.00 0.85

0.94 0.99 1.00 0.98

value

sources.

senting 69% of the total). The sensitivity and specificity of DCBE to detect polyps decreases with the size of the polyp [4], but polyps < 1 cm are very unlikely to bleed [12], although a great proportion can be adenomas. In the diagnosis of IBD, DCBE had a lower sensitivity compared to colonoscopy, but it is known that in order to detect IBD by DCBE there must be considerable mucosal surface changes [13]. The onset of IBD is primarily in young people, and these cases are already clinically suspect, so they would probably benefit from an initial colonoscopy; not forgetting that the statistically increased risk for cancer in IBD is based on radiological studies. Surprisingly, all other cases of colitis (N = 5) were diagnosed by DCBE. Three out of six solitary ulcers were missed by DCBE, demonstrating the difficulties in detecting such lesions. Angiodysplasia cannot be diagnosed by DCBE, but it is not a frequent cause of bleeding and it only represented 3% of the cases in our study; a value similar to that reported by others [8]. Our results support the idea that a high-quality DCBE has a high sensitivity and specificity to detect the most frequent causes of lower gastrointestinal bleeding. It requires well-trained and patient radiologists to analyze the films, preferably with a doublechecking routine to diminish perceptive errors, which have been one of the major drawbacks of DCBE [2, 14]. After these considerations, we conclude that DCBE is a good first diagnostic choice in the diagnosis of the cause of lower gastrointestinal bleeding. Due to resource limitations, colonoscopy is not as widely available as DCBE; therefore, remissions for colonoscopy should be well chosen. If, after a negative or inadequate DCBE, there is still a suspicion of bleeding and other more proximal sources of gastrointestinal bleeding have been excluded, then colonoscopy may be of benefit: mainly in order to diagnose other causes of bleeding other than carcinoma and polyps < 1 cm, and also with the p0ssibjlity to offer therapeutic treatment. Colonoscopy should first be considered if there is a suspicion o f IBD.

Acknowledgments. We want to thank Ms. Margareta Ibring and the personnel from the radiological archives for their patience in collecting the clinical cases.

References 1. Thoeni RF, Margulis AR. Colon. Radiological examination. In: Margulis AR, Burhenne J, eds. Alimentary tract radiology, vol. 1. Baltimore: CV Mosby, 1989:893-931 2. Simpkins KC. Large intestine. Radiology. In: Fazio VW,

Phillips SF, eds. Current opinions in gastroenterology 1990;6:50-53 3. Investigation of rectal bleeding [Editorial]. Lancet 1989;1:195-197 4. Stewart ET, Dodds WJ. Colon, polyps. In: Margulis AR, Burhenne HJ, eds. Alimentary tract radiology, vol. 1. Baltimore: CV Mosby, 1989:1017-1049 5. Saito Y, Slezak P, Rubio C. The diagnostic value of combining flexible sigmoidoscopy and d0uble-contrast barium enema as one-stage procedure. Gastrointest Radiol 1989;14:357-359 6. Irvine EJ, O'Connor J, Frost RA, Shorvon P, Some~;s S, Stevenson GW, Hunt RH. Prospective comparison of double contrast barium enema plus flexible sigmoidoscopy v colonoscopy in rectal bleeding: barium enema v colonoscopy in rectal bleeding. Gut 1988;29:1188-1193 7. Miller RE, Lehman G. Polypoid colonic lesions undetected by endoscopy. Radiology 1978;129:295-297 8. Williams CB. Colonoscopy. Br Med Bull 1986;42:265-269 9. Barry MJ, Mulley AG, Richter JM. Effect of workup strategy on the cost-effectiveness of fecal occult blood screening for colorectal cancer. Gastroenterology 1987;93:301-310 10. Knutson CO, Max MH. Value of colonoscopy in patients with

rectal blood loss unexplained by rigid proctosigmoidoscopy and .barium contrast enema examinations. Am J Surg 1980;139:84-87 11. Funch DP. Diagnostic delay in symptomatic colorectal cancer. Cancer 1985;56:2120-2124 12. Macrae F A , St John JB. Relationship between patterns of bleeding and hemoccult sensitivity in patients with colorectal cancers or adenomas. Gastroenterology 1982;82:891-898 13. Thoeni RF, Margulis AR. Colon. Inflammatory diseases. In:

Margulis AR, Burhenne J, eds. Alimentary tract radiology, vol. 1. Baltimore: CV Mosby, 1989:963-1015 14. Fork FT. Radiographic findings in overlooked colon carcinomas. Acta Radiol 1988;3:331-336 15. Schuman BM. When should colonoscopy be the first study for active lower intestinal hemorrage? [Editorial]. Gastrointest Endosc 1984;6:372-373

Received: April 15, 1991; accepted May 18, 1991