AJR ARRS Member Benefits
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Glick, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Glick, S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
AJR 2000; 174:1529-1537
© American Roentgen Ray Society

Opinion

Double-Contrast Barium Enema for Colorectal Cancer Screening

A Review of the Issues and a Comparison with Other Screening Alternatives

Seth Glick1

1 Department of Radiology, MCP-Hahnemann University, Broad and Vine Sts., Philadelphia, PA 19102.

Received September 2, 1999; accepted after revision November 4, 1999.

 
Address correspondence to S. Glick.


Introduction
Top
 Introduction
Evidence-Based Effectiveness
 Impact of the Screening...
 Screening Test Characteristics
 Conclusion
 References
 
In the last 5 years, multidisciplinary panels have endorsed several approaches and provided guidelines for colorectal cancer screening [1, 2]. Medicare now covers fecal occult blood testing, flexible sigmoidoscopy, or double-contrast barium enema for average-risk patients and double-contrast barium enema or colonoscopy for high-risk patients. Although double-contrast barium enema is an equivalent option for colorectal cancer screening of the general population and high-risk patients, reports in the clinical literature commonly either ignore it or diminish its value. I present a balanced comparison of the available techniques for colorectal cancer screening.

Cancer screening is unlike symptom-based medical practice in that test specificity takes on greater significance and optimization may be required at the expense of sensitivity. A screening test should be repeated over defined intervals, providing an opportunity for the detection of previously overlooked or new lesions before their progression to incurable cancer. However, healthy individuals are exposed to procedures that may be costly and have the potential for physical and psychologic harm. The well-being of those who may not be destined to develop colorectal cancer must be prioritized. A life lost because of screening is not equivalent to a life saved in the process.


Evidence-Based Effectiveness
Top
 Introduction
 Evidence-Based Effectiveness
Impact of the Screening...
 Screening Test Characteristics
 Conclusion
 References
 
For public health policy, factors such as test accuracy, cost, and safety are often irrelevant without proof that testing an asymptomatic population is effective in reducing disease-related mortality. Researchers must show that a screening test that detects disease before clinical presentation can result in a decreased mortality rate. Any procedure that identifies lesions before clinical presentation should be appropriate if it can be performed at intervals that result in acceptable aggregate expense or morbidity. The ideal methodology for determining screening efficacy is the randomized controlled trial. However, such studies usually require a large study population and enormous financial expenditures [3], and when mortality is the measured outcome, a long evaluation period may be required. The necessity for a randomized trial for each screening procedure in the presence of compelling alternative information is questionable, especially in light of the time and resources required.

The strongest evidence for colorectal cancer screening efficacy comes from trials using fecal occult blood tests. The benefit is derived through the detection of cancer at an early pathologic stage. Maximum mortality reduction (33%) was modest, occurring only when stool specimens were rehydrated (more sensitive, less specific) and obtained annually [4]. Because of the number of false-positive examinations, 38% of the patients screened underwent colonoscopy by the 13th year of the study. When screening was biennial [5,6,7] or when nonhydrated samples were used [6, 7], mortality reduction was 15-21%. No trials have assessed the effectiveness of annual nonhydrated testing—the strategy recommended in most guidelines.

Evidence of mortality reduction for sigmoidoscopy is derived from case-control studies. In one study [8], researchers estimated that sigmoidoscopy reduced rectal cancer mortality by approximately 70%. However, on the basis of strict technique-linked criterion, the results are not applicable to flexible sigmoidoscopy because a rigid sigmoidoscope was used. Furthermore, because no cancers were identified in the control group, any mortality reduction should have occurred because of cancer prevention by polypectomy. Although 12 control patients had adenomas removed, no details were given regarding lesion size or precise histology. In another case-control study, approximately two thirds of the procedures were performed with a flexible sigmoidoscope [9]. However, only 27 patients had fatal distal cancer and some were at higher risk for colorectal cancer. In a third case-control study, the study population was symptomatic [10].

In theory, performing fecal occult blood tests and flexible sigmoidoscopy should be supplementary. However, no studies reveal improved mortality reduction when the techniques are combined, and the overlap in identified lesions is undefined. The sensitivity of screening fecal occult blood tests for early cancer proximal to the reach of the flexible sigmoidoscope has never been studied (no investigations break down the results by location and stage). However, in a community-based fecal occult blood program, cancers proximal to the splenic flexure tended to be more advanced: 38% were disseminated compared with 22% in the left colon [11]. Another study compared two screening groups, one undergoing rigid sigmoidoscopy and the other undergoing rigid sigmoidoscopy and fecal occult blood testing [12]. Mortality was reduced with the addition of fecal occult blood tests, but results were of borderline statistical significance. Furthermore, any added contribution from fecal occult blood tests could have resulted from the detection of cancers that could have been visible with a longer instrument. Finally, because the researchers did not have an unscreened control group to document mortality reduction from sigmoidoscopy alone, all the benefit could have been attributed to fecal occult blood testing.

No studies have confirmed mortality reduction from screening colonoscopy in either average- or high-risk patients. However, colonoscopy was the primary method of diagnostic follow-up used in three fecal occult blood trials [4,5,6,7]. Because the direct identification of cancer was actually responsible for the mortality reduction, what is at issue is not whether colonoscopy can reduce mortality but rather how frequently it should be performed. Further support is derived from the National Polyp Study [13], in which the screening population underwent colonoscopic polypectomy at the time of entry and during surveillance. Researchers determined a 76-90% decrease in cancer incidence compared with that expected on the basis of historical controls, suggesting a strong correlation between adenoma removal and cancer reduction. Although colonoscopy served as a diagnostic and therapeutic tool, any technique with similar sensitivity, particularly for large adenomas and early cancer, should attain comparable outcomes. Disease prevalence, cost, and safety need to be considered when choosing a screening or surveillance approach.

The strongest support for double-contrast barium enema is based on the observation that treatment of early cancer in asymptomatic individuals lowers disease-specific mortality and the removal of adenomatous polyps reduces cancer incidence. The double-contrast barium enema reveals most malignant and premalignant lesions. In the first 4 years of one trial, a single-contrast barium enema was performed after positive screening findings [4]. The reported sensitivity of the barium enema for cancer was 80%. In all probability, barium enema contributed to mortality reduction. In another trial, double-contrast barium enema with flexible sigmoidoscopy was used for diagnostic workup [14]. Researchers noted a shift in the tumor distribution to earlier stages and mortality reduction comparable with that of other trials performing biennial screening [15]. The sensitivity of double-contrast barium enema for cancer was 80%, and barium enema alone revealed one third of 81 cancers. In the previously described study in which symptomatic patients had a reduced colorectal cancer mortality, a number of patients also underwent barium enema studies [10]. Researchers noted a statistically significant mortality reduction in patients who underwent previous barium enema (Sonnenberg A, personal communication). A prospective study of high-risk patients with hereditary nonpolyposis colorectal cancer revealed a reduction in cancer incidence for those undergoing screening at 3-year intervals [16]. More than two thirds of the initial evaluations and approximately half of the repeated evaluations were performed with double-contrast barium enema and flexible sigmoidoscopy. Finally, the National Polyp Study reported a reduction in cancer incidence after adenoma removal [13]. However, the relative impact of initial polyp clearance and surveillance cannot be determined. Given the larger size and more ominous histology of the lesions at entry in conjunction with the short follow-up time in the latter study, initial polypectomy probably played the predominant role. Approximately one third of patients with polyps entered the study after receiving positive results on barium enema [17]. A number of patients also had positive findings on double-contrast barium enema during the course of surveillance.


Impact of the Screening Program
Top
 Introduction
 Evidence-Based Effectiveness
 Impact of the Screening...
Screening Test Characteristics
 Conclusion
 References
 
Although screening tests should be accurate, safe, inexpensive, widely available, and associated with minimum patient discomfort and disability [18, 19], all these terms are subjective and relative. Their true magnitude is a function of the continued implementation of an entire program, including the screening tests and the additional investigation generated. The necessity of detecting and eliminating every adenoma as opposed to a stratified approach based on lesion size has significant bearing on the cumulative effect.

Models compare the overall implications of each option by integrating the disparate aspects of several strategies. However, the results must be interpreted with caution. The findings are strategy-specific, and the outcome cannot be attributed to the procedure out of context. Proposed screening and surveillance intervals, the definition of a significant lesion (that which triggers diagnostic evaluation and surveillance), and the prevalence (particularly by site) of the latter are crucial. Models addressing the same issues may conflict in their conclusions because they may assess different protocols, vary in design, differ in assumptions, or incorporate uncertain or selective data. Most models assume perfect compliance. However, suboptimal participation maximally impacts the effectiveness of protocols that provide benefit through early cancer detection rather than prevention or are intrinsically less complete or sensitive. Moreover, models do not reflect how, in practice, physicians may request more frequent or extensive investigation than that described in the model's algorithms, escalating cost and harm without necessarily improving outcome.

Some researchers evaluate total cost to determine if projected fiscal requirements are available (cost-identification analysis). One study estimated that screening sigmoidoscopy performed every 3 years would ultimately result in an annual expenditure of $20 billion by 2010 [20]. Other models assess clinical outcomes (decision analysis). Distinct pathways based on a range of strategies and their probable results can quantify all events (e.g., deaths prevented, complications) over a defined time for a given population. Characterizing net benefit in terms of life-years saved is most relevant because a cancer death prevented may add only a few years to a patient's estimated life expectancy. One decision analysis [2] revealed the following results per 100,000 patients for the most commonly recommended regimens (the numerator represents life-years saved and the denominator indicates significant complications): annual fecal occult blood tests with flexible sigmoidoscopy every 5 years (11,760/1069), double-contrast barium enema every 5 years (12,568/345), and colonoscopy every 10 years (12,804/1520). Small adenomas found during sigmoidoscopy did not generate colonoscopic follow-up, and patients with small adenomas identified during endoscopy were not prescribed surveillance. The more aggressive the approach, the greater the number of complications that can be expected.

The most frequently cited model addresses cost-effectiveness. Such an analysis attributes a monetary expense to a specific outcome. The gain is expressed in terms of cancers detected, cancer deaths prevented, or life-years saved. These analyses not only provide a process for assessing various approaches to a specific problem, but also enable comparison with interventions for other diseases, providing a determination regarding where the allocation of funds could have the greatest impact. The most commonly invoked benchmark is $40,000/life-year, which is the value associated with existing screening or therapeutic programs.

Cost-effectiveness figures may be misleading. The cost-effectiveness of a strategy is typically expressed in reference to the alternative of no screening (the average). Two strategies may have similar average cost-effectiveness, but one strategy may produce greater benefit at a proportionally greater total cost. For example, screening mammography of one breast, at a 50% reduction in cost, would presumably be as cost-effective as bilateral mammography. What is more relevant is the incremental cost-effectiveness or the cost per additional life-year saved with a more effective but more costly option. It is this parameter that should be compared with the $40,000 benchmark. Decisions are essentially subjective and depend on the willingness or ability to pay for the maximum benefit.

Few published colorectal screening cost-effectiveness analyses assess the gamut of reasonable strategies for average-risk patients. A study from the United States Congress Office of Technology Assessment [21] revealed the following cost per life-year saved statistics: annual fecal occult blood tests with flexible sigmoidoscopy every 5 years ($13,639/$11,652), double-contrast barium enema every 5 years ($13,495/$9,435), and colonoscopy every 10 years ($22,175/$9,287). The first figure for each protocol assumes a 5-year progression from normal mucosa to invasive cancer, and the second figure is based on a 10-year dwell time. This study incorporated a 70% sensitivity for the double-contrast barium enema, regardless of lesion size or histology. The respective costs used were flexible sigmoidoscopy, $80; double-contrast barium enema, $131; and colonoscopy, $285.

A considerable portion of the financial burden from screening is generated by surveillance, particularly if all patients with previous adenomas are candidates [22]. One estimate of surveillance colonoscopy, assuming 100% effectiveness, predicted a cost of $82,000 for every cancer death prevented [23]. A policy focusing on neoplasms larger than 1 cm is associated with a dramatic improvement in cost-effectiveness while maintaining absolute effectiveness [24]. It is common for patients with positive findings on fecal occult blood tests to undergo colonoscopy. However, as a result of the low positive predictive value of fecal occult blood tests and the relative costs of the procedures, several analyses have revealed double-contrast barium enema follow-up to be more cost-effective [25,26,27]. Modeling has also revealed that, even for high-risk patients (e.g., patients with a family history of malignancy), other procedures such as double-contrast barium enema may be most cost-effective [28].


Screening Test Characteristics
Top
 Introduction
 Evidence-Based Effectiveness
 Impact of the Screening...
 Screening Test Characteristics
Conclusion
 References
 
When the patient, physician, or both are selecting an approach to screening, objective and subjective factors are influential. Some are more concrete and are based on test capability, whereas others are less tangible and are related to the individual experience associated with the procedure. Initial and ongoing patient compliance and the availability of resources may be the most significant factors of effectiveness of a screening program.

Fecal occult blood tests are safe and inexpensive. Several methods of fecal occult blood testing exist, and most methods have sensitivities ranging from approximately 40% to 60%. Sensitivities as low as 25% have been reported [29]. The sensitivity of these tests for early-stage cancer has not been investigated. Fewer than 10% of large adenomatous polyps are detected because such lesions seldom bleed. Because fecal occult blood tests are directed toward detecting early malignancy, adherence to the prescribed regimen is imperative. Suboptimal compliance may result in overlooked malignancies. In research trials, a higher percentage of symptomatic cancers detected in previously screened patients were of a more advanced stage. A report negative for cancer may result in an inappropriate sense of security and subsequent failure to screen or even respond to symptoms. Even when the test reveals positive results, patients may fail to follow up with total colonic evaluation [11, 30], or tests unrelated to the colon may be ordered. [31]. Another problem is low specificity. From 2% to 10% of tests reveal positive findings, but 90% of patients with positive findings do not have colorectal cancer. The false-positive rate in mass population screening may be significantly higher than that reported in controlled studies because adherence to performance requirements may not be as readily achieved [32].

Sigmoidoscopy can detect adenomas and cancers. It is associated with some expense ($116 or $140 with biopsy) [33], a low risk of perforation (one in 10,000), and a moderate degree of patient discomfort. Approximately 40% of patients have some form of polyp that requires biopsy and 12% have an adenoma, potentially requiring colonoscopy [34]. One study showed that patients considered flexible sigmoidoscopy significantly more painful than barium enema or colonoscopy [35]. Unlike full colonic investigations, the patient preparation for sigmoidoscopy is not as intensive (one or two enemas); however, prospective studies reveal that poor preparation precludes adequate rectosigmoid evaluation in 20% of examinations [36, 37]. Examiner skills may vary and not all may use the 60-cm sigmoidoscope, limiting its potential yield. Even when passed to the splenic flexure, the 60-cm endoscope can reach only 40-50% of neoplasms. However, visualization is confined to the rectum and sigmoid colon in about 75-80% of attempts [38,39,40], resulting in the potential identification of 30-40% of lesions. Furthermore, neoplasms within the traversed bowel may be overlooked. Effective sensitivity can be viewed in terms of all lesions detected, both initially and subsequently at follow-up colonic examination performed as a result of positive findings at sigmoidoscopy. Even with this protocol, at least 30% of proximal adenomas and carcinomas will be over-looked because of a lack of association with a distal adenoma [41, 42]. This percentage may be higher depending on the actual reach and yield of sigmoidoscopy, or whether a small adenoma generates a colonoscopic evaluation (Figs. 1A and 1B). This is particularly relevant in populations such as African Americans, in whom the prevalence of neoplastic disease proximal to the sigmoid is significantly higher [43,44,45,46,47,48].



View larger version (183K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1A. —62-year-old man with constipation. Spot radiograph obtained during barium enema of sigmoid colon shows 5-mm sessile polyp (arrows).

 


View larger version (142K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1B. —62-year-old man with constipation. Overhead radiograph obtained during barium enema shows subtle flat elevation in proximal transverse colon (arrows). Patient had carcinoma confined to muscularis propria. If patient were screened with flexible sigmoidoscopy and colonoscopy only after finding adenoma larger than 1 cm, early cancer would not have been discovered.

 

Although sigmoidoscopy is widely available because it is performed by several types of specialists, financial incentives and relevant logistics (time, facilities) may be inadequate for physicians who are not primarily endoscopists [49]. When performed by fulltime endoscopists, sigmoidoscopy screening or colonoscopic surveillance intervals may be shorter than recommended in most guidelines, and pathologic criteria for follow-up colonoscopy or surveillance may be less restrictive, generating increased cost and complications. When "sigmoidoscopy" is performed with a colonoscope [50], the observation of an elevation of any size in the distal colon usually triggers immediate full colonic evaluation rather than reserving colonoscopy until a biopsy has confirmed distal neoplasia.

Another poorly defined but potentially significant problem is the fact that endoscopes require a time-consuming antimicrobial cleansing process. With increasing procedural volume, a greater number of instruments are required. In an attempt to minimize expense or maximize volume, cleansing or removal of the disinfectant may be compromised, resulting in the iatrogenic transmission of infection [51,52,53,54,55] or significant chemical proctocolitis [56,57,58,59]. The latter occurred in 3% of patients who participated in a sigmoidoscopy screening program [59]. Research shows that patients undergoing endoscopy have contracted bacterial [52] and viral (hepatitis B and C) infections as a result [52,53,54]. Appreciation of the degree of risk of the latter disease is problematic because the etiologic relationship may be unapparent (because of the remote temporal proximity of the procedure and the clinical diagnosis). Two studies revealed that even after vigorous cleaning, as many as 25% of endoscopes had bacterial and viral contamination [51, 55]. Moreover, HIV can remain viable for days on a cleansed endoscope [60].

An approach that combines annual fecal occult blood testing with flexible sigmoidoscopy incorporates all the weaknesses and limitations of the former test. No studies quantify early cancers identified but not otherwise detected through a program of sigmoidoscopy with selective colonoscopic follow-up. Cross-sectional studies fail to reveal an enhanced neoplastic yield from simultaneous exposure [61,62,63,64,65]. These studies were not repeated over time, so cumulative potential cannot be assessed.

Although portions of the colon may not be adequately visualized in approximately 5% of studies because of residual stool, the barium enema evaluates the entire colon in almost all patients. Other technical factors may also detract from study quality. Both the laxative preparation and the procedure itself may be associated with mild to moderate, and occasionally severe, discomfort. Similar to the way in which sigmoidoscopy often requires colonoscopy, barium enemas having positive results commonly require a second investigation. Therefore, the definition of a significant positive result, prevalence of such lesions, and probability of a positive result (true and false) are all factors that influence the overall cost and logistic burden of cancer screening techniques.

The expense of a double-contrast barium enema is slightly higher than that of a sigmoidoscopy ($48 professional + $97 technical = $145) [33]; however, the double-contrast barium enema is safer. The only study focusing on double-contrast barium enema complications reveals a perforation rate of one in 25,000 and an associated mortality rate from perforation of one in 250,000 [66]. The most significant question is that of the accuracy of a double-contrast barium enema in a screening population. No studies use the double-contrast barium enema as the primary procedure in a screening population. However, adenomatous polyps do not usually produce symptoms. Their prevalence and size distribution are comparable in symptomatic (with the exception of those with rectal bleeding) and asymptomatic patients [67, 68]. Investigations of symptomatic patients are essentially a search for cancer with serendipitous adenoma detection.

A number of problems affect the interpretation of published performance data. Sensitivity citations in the clinical literature are frequently too low. The number of lesions analyzed is often small, with weak statistical power producing wide confidence intervals. Furthermore, verification methodology is usually biased. Radiologic sources tend to overestimate the capabilities of a barium enema. In several studies, study design has consisted of a review of neoplasms obtained from pathology registries that were recorded during a specified time interval in which temporally correlated barium enemas were identified [69,70,71,72,73,74,75]. Under these circumstances, many lesions were initially detected on barium enemas; these findings represent selection bias. Patients with normal results did not usually undergo further evaluation, precluding a determination of overlooked neoplasms with a probable overestimation of sensitivity. This deficiency is compensated for when assessing cancer detection, assuming a geographically stable population and that retrospective barium enema retrieval extends for an appropriate time period (e.g., >2 years) before diagnosis. Studies using this methodology report a sensitivity for cancer of 85-90% [69,70,71,72,73,74,75]. Less adequately evaluated is the sensitivity for early cancer (Figs. 2A and 2B). Because survival statistics are highly favorable in the absence of nodal or distant spread, such lesions should be considered early cancer. However, early cancer has been variably defined, sometimes limited to tumors confined to the submucosa or bowel wall, or inappropriately including noninvasive lesions with no metastatic potential. Reports describing sensitivity for localized disease are few, reporting ranges of 58-94% [70, 72, 76]. In the fecal occult blood trials, in which lesion population was skewed toward earlier disease, reported sensitivity was 75-80% [12, 14].



View larger version (144K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2A. —59-year-old man with rectal bleeding. Anteroposterior image of rectum from barium enema shows enface appearance of 3.5-cm flat discoid filling defect (arrows).

 


View larger version (167K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2B. —59-year-old man with rectal bleeding. Lateral image shows lesion in tangent (arrows). Patient had early cancer confined to muscularis propria.

 

In some studies, analysis is restricted to a specific colonic segment, such as the rectosigmoid or the colon proximal to the rectosigmoid. The problem with drawing conclusions for a defined region is that, with examiner skills so crucial to outcome, it is difficult to determine whether the operating characteristics in a particular report are definitive for that segment or whether, in that setting, similar results would be obtained in the remainder of the colon. Some contend that the rectosigmoid is inadequately evaluated, thereby requiring supplemental sigmoidoscopy. Although the combination of two procedures should detect more total lesions, the magnitude, if any, of such improvement on outcome is uncertain; and justification for the routine use of a second procedure with its associated inconvenience, discomfort, cost, and risk is questionable. Furthermore, it is possible that patients who might benefit from supplemental sigmoidoscopy can be selected on the basis of double-contrast barium enema results. For example, sensitivity has been correlated with the number of sigmoid diverticula [77, 78].

Determining the capability for adenoma detection is more difficult than that for cancer. Sensitivity data may not distinguish adenomas by size, or the size categories under evaluation may vary. The perceived significance of small adenomas has tremendous implications because sensitivity and specificity improve markedly with increasing size. Fewer than 1% of adenomas smaller than 1 cm contain, or ever progress to, invasive cancer. Even most adenomas larger than 1 cm are associated with a benign natural history. In an in vivo observation of such lesions [79], only 25% of patients developed invasive cancer over 20 years, with a cumulative incidence of 2.5% at 5 years and 8% at 10 years. When a double-contrast barium enema is performed at 5-year intervals, previously overlooked neoplasms larger than 1 cm have a reasonable window of opportunity for subsequent detection without detrimentally affecting a patient's outcome. Moreover, sensitivity is usually determined using the total number of lesions rather than the identification of patients with neoplastic disease. Because one fourth to one third of patients with adenomas have multiple lesions, such a methodology would underestimate patient sensitivity. In the context of screening, and even in a diagnostic context, it is not essential to identify every lesion, but rather to select patients to undergo the definitive procedure.

Specificity data are limited because negative findings are rarely confirmed. Furthermore, in biostatistical terminology, specificity is a function of the false-positive rate. However, the attributed specificity has frequently been underestimated because of inappropriate referrals to published statistics for false-positive results in which the source studies used the concept of false-positive as a function of the positive predictive valve. Although the positive predictive value carries implications analogous to those of specificity, an important distinction exists. The former indicates the percentage of positive results that are correct, whereas specificity reflects the percentage of individuals without disease correctly identified as healthy. One study reports a false-positive rate of 10% (positive predictive value = 90%) for polyps larger than 1 cm [80]. Additionally, the fact that "false-positive" examinations may be endoscopic false-negative findings [81,82,83,84,85] cannot be overlooked (Fig. 3). Typically, researchers make no attempt to adequately resolve discrepancies when endoscopy overlooks a lesion revealed on barium enema studies. This practice has considerable bearing on sensitivity and specificity. In two studies addressing this concern [81, 82], approximately half the lesions that were overlooked on colonoscopy were ultimately verified as neoplastic. A false-positive diagnosis of cancer is extremely rare, although some reports inappropriately include significant disease (e.g., large benign neoplasms, diverticulitis) initially reported as cancer as a false-positive finding. Therefore, the specificity for malignant disease approaches 100%. The negative predictive value integrates sensitivity and lesion prevalence, allowing a broader appreciation of the degree of certainty and confidence level after a negative report. In one surveillance study, researchers reported a negative predictive value of 93% for adenomas larger than 1 cm [86].



View larger version (160K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3. —65-year-old woman with right lower quadrant pain. Anteroposterior image of rectum shows 1.3-cm lobulated sessile polyp (arrow) with wide irregular base. Lesion was not visible during initial colonoscopy. After subsequent polypectomy, lesion was identified as tubulovillous adenoma containing carcinoma extending into submucosa.

 

Another poorly appreciated association is that of factors that adversely affect performance (diverticulosis, poor preparation, limited mobility, inability to retain contrast material) and increasing patient age, particularly in the very elderly. The older the patient, the less likely a missed adenoma, regardless of size, will progress to a fatal cancer in their lifetime. Optimal performance and outcomes would more likely be obtained in a screeningage population.

A reasonable estimate of accuracy for large polyps can be determined by focusing on studies in which an adequate number of patients are studied and all patients undergo endoscopy. In one study, 190 patients who underwent a double-contrast barium enema were randomly selected for colonoscopy [87]. The sensitivity was 81% and specificity 96% for adenomas larger than 1 cm. In a study in which colonoscopy and double-contrast barium enema were used for surveillance in patients with previous adenomas, the sensitivity for lesions larger than 7 mm was 71% and the specificity was 98% [88]. Most overlooked lesions measured between 7 and 10 mm. Preliminary data from the National Polyp Study, another surveillance population, have been widely quoted as showing a sensitivity of less than 50% for adenomas larger than 1 cm. However, the total number of large lesions was small (<25), and almost all measured approximately 1 cm, enabling measurement variation to potentially falsely elevate this already minimal population. The method used to assess polyp size has been shown to consistently overestimate the true dimension by several millimeters [89]. The fact that sensitivity was higher for polyps smaller than 1 cm gives credence to the likelihood that many of these lesions were smaller than the 1-cm threshold. Even if the assigned diameters were accurate, it is questionable whether they were representative of the size distribution of neoplasms larger than 1 cm likely to be encountered in a screening population. Studies focusing on the rectosigmoid have reported sensitivities of 75-81% for adenomas larger than 1 cm [14, 90, 91]. One report described a detection rate of 100% with a specificity of 96% [92]. These results are not significantly dissimilar to those for the overall performance of the double-contrast barium enema. Therefore, the double-contrast barium enema can probably detect at least three quarters of adenomas larger than 1 cm, and possibly an even higher percentage of patients with such lesions. The cumulative benefits of periodic screening and the influence of the long natural history should also be considered.

A more significant issue is the wide variation of skill in performance and interpretation. Although this issue also exists with endoscopy, the magnitude of the problem is probably not as great. However, a significant number of endoscopic procedures (including colonoscopy) are performed by nongastroenterologists, and evidence suggests that their performance is not superior to the performance of barium enemas performed by radiologists from community or academic practices [69]. Standardization and quality improvement can be accomplished, as has been shown with mammography, through educational initiatives and accreditation programs. Because most overlooked lesions are visible on radiographs, attention to observational skills can enhance lesion detection [75, 93, 94]. A theoretic concern regarding the barium enema is radiation exposure. However, the cumulative dose, when the procedure is performed every 5 years, is lower than that from annual mammography over a comparable time interval.

From the standpoint of efficiency, colonoscopy would appear to be the most attractive approach. Its major advantage is that it represents a single procedure that can evaluate the entire colon and is diagnostic and therapeutic. However, more than 95% of patients screened do not have a large adenoma or cancer. Because of its "superior" sensitivity, particularly for small adenomas, longer screening intervals (10 years) are acceptable. Although higher sensitivity for small adenomas may be beneficial when future screening or surveillance compliance is questionable, the ongoing detection and removal of small polyps may actually be detrimental, because costs escalate and complications increase without evidence that such an aggressive approach significantly improves outcome. Many colonoscopy patients are unnecessarily assigned to surveillance regimens. This problem is compounded when the intervals are shorter than recommended. Colonoscopy has the highest cost of all procedures ($285 professional + $384 technical = $669) [33]. The cost increases when polypectomy is performed. The perforation rate of colonoscopy is at least 25 times greater than that of the double-contrast barium enema [66, 95] (one in 1000 for diagnostic colonoscopy and one in 500 after polypectomy). Polypectomy is also associated with major hemorrhage (three in 1000). In one prospective series, 5% of patents undergoing polypectomy experienced complications and 2% required hospitalization [96].

Other serious complications are common. Respiratory depression from sedation occurs in one in 200 [2], and the mortality rate is one to three in 10,000 [2]. Complications of lesser severity tend to be underreported; for example, in one prospective series, 17% of patients reported some adverse procedure-related event after a 30-day follow-up survey [97]. An emergency department or physician visit was precipitated in 1.7% of all patients undergoing colonoscopy. The potential for transmission of infection or colitis from the cleansing agents, as with sigmoidoscopy, is also applicable to colonoscopy.

Discomfort from the preparation and the procedure may be associated with colonoscopy. Studies report conflicting information regarding the magnitude of the negative experience relative to the double-contrast barium enema [98, 99]. Use of sedation necessitates a recovery period, precluding usual activities (e.g., working) for that day. One report showed that 20% of patients missed at least one additional workday after undergoing colonoscopy [100]. The presence of a traveling companion or some arranged means of transportation is also required because of the lingering effects of the medications. Although the colonoscope can assess the entire colon, at best, the cecum is reached in approximately 85-95% of attempts [101]. Some reported completion rates have been lower, reflecting experience or type of specialty. Overall practice may be more consistent with such figures. Of even greater concern is that, on occasion, the endoscopist may incorrectly believe that the entire colon has been viewed. Suboptimal preparation is an occasional problem but has rarely been mentioned or quantified. Incomplete examinations and overlooked lesions take on increased significance in view of the recommended 10-year span between screening examinations. This long interval may also be compromised by cases in which the rate of progression exceeds the usual estimate.

Although it is often assumed that colonoscopy is infallible in visualized portions of the colon, even large lesions may be overlooked on complete examinations [85]. The precise sensitivity of colonoscopy for cancer and various-sized adenomas is difficult to determine because colonoscopy has been accepted as the standard of reference. This generalization tends to result in overestimates. Methods used to determine sensitivity have included tandem or back-to-back colonoscopies performed by two endoscopists during a single visit or repeated colonoscopies performed by the same or a different individual after a short interval. Although operative specimens represent the ideal reference, current findings are mostly from segmental resections, thus limiting general applicability. Retrospective observational studies suggest a sensitivity of about 90-95% for cancer when performed by gastroenterologists [69, 73, 74, 102]. However, sensitivities of 85% have been reported for colonoscopies performed by nongastroenterologists or colonoscopies performed in general community practice [69, 103]. A review of Medicare submissions indicates that almost half of colonoscopies are not performed by gastroenterologists [104]. Tandem or follow-up study designs indicate a sensitivity of 90% for adenomas larger than 1 cm and 75% for smaller adenomas [105, 106]. After resection correlation, one study indicated a sensitivity for adenomas larger than 1 cm as low as 76% [107]. Whatever the true numbers, the sensitivity of colonoscopy is probably no more than 10% greater than that of the double-contrast barium enema for cancer and 15% for large adenomas. The relevant consideration is to what degree this improvement affects outcome and whether this justifies the difference in cost and harm. The specificity of colonoscopy is often assumed to be 100%. However, nonneoplastic abnormalities may be considered false-positive results because their removal is associated with increased cost (procedure and pathologic interpretation) and complications, but no benefit. Such lesions comprise 28-50% of all polyps [108].

Another limitation of colonoscopy is its inferiority to double-contrast barium enema for lesion localization [109, 110]. This limitation may be particularly important for neoplasms that are difficult to detect during surgery, particularly laparoscopic colectomy [111]. Finally, when a double-contrast barium enema reveals multiple adenomas and endoscopy fails to correlate precisely with all lesions, repeated studies can resolve the discrepancy.


Conclusion
Top
 Introduction
 Evidence-Based Effectiveness
 Impact of the Screening...
 Screening Test Characteristics
 Conclusion
References
 
No "best test" for colorectal cancer screening exists; however, patients can choose from several techniques that can be variably effective. Each technique has features that make it more or less attractive. Advocates for fecal occult blood testing, flexible sigmoidoscopy, or both combined emphasize the costs and morbidity of colonoscopy, and promoters of colonoscopy emphasize the diagnostic limitations of the former. However, the double-contrast barium enema offers the positive features of both options. The double-contrast barium enema is as safe or safer than flexible sigmoidoscopy, has comparable costs, provides a complete screening of the entire colon, and has a potential sensitivity similar to that of colonoscopy.

If the goal of screening is to maximize the identification of high-risk patients while minimizing expense and morbidity, the double-contrast barium enema most closely achieves this task. Its inclusion by Medicare as the only accepted screening test for both average-and high-risk patients reflects these characteristics. What remains is for radiologists to accept the challenge by developing the skills necessary to obtain results. In doing so, a significant contribution to this major health care problem can be made.


References
Top
 Introduction
 Evidence-Based Effectiveness
 Impact of the Screening...
 Screening Test Characteristics
 Conclusion
 References
 

  1. Byers T, Levin B, Rothenberger D, Dodd GD, Smith RA. American Cancer Society guidelines for screening and surveillance for early detection of colorectal polyps and cancer: update 1997. CA Cancer J Clin 1997;154:154 -160
  2. Winawer SJ, Fletcher RH, Miller L, et al. Colorectal cancer screening: clinical guidelines and rationale. Gastroenterology 1997;112:594 -642[Medline]
  3. Smart CR. Limitations of the randomized trial for the early detection of cancer: a clinical perspective. Cancer 1997;79:1740 -1746[Medline]
  4. Mandel JS, Bond JH, Church TR, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993;328:1365 -1371[Abstract/Free Full Text]
  5. Mandel JS, Church TR, Ederer F, Bond JH. Colorectal cancer mortality: effectiveness of biennial screening for occult blood. J Natl Cancer Inst 1999;91:434 -437[Abstract/Free Full Text]
  6. Hardcastle JD, Chamberlain JO, Robinson MHE, et al. Randomised controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet 1996;348:1472 -1477[Medline]
  7. Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sandergaard O. Randomised study of screening for colorectal cancer with faecal-occult-blood test. Lancet 1996;348:1467 -1471[Medline]
  8. Selby JV, Friedman GD, Quesenberry CP, Weiss NS. A case-control study of screening sigmoidoscopy and mortality from colorectal cancer. N Engl J Med 1992;326:653 -657[Abstract]
  9. Newcomb PA, Norfleet RG, Storer BE, Surawica TS, Marcus PM. Screening sigmoidoscopy and colorectal cancer mortality. J Natl Cancer Inst 1992;84:1572 -1575[Abstract/Free Full Text]
  10. Muller AD, Sonnenberg A. Protection by endoscopy against death from colorectal cancer: a case-control study among veterans. Arch Intern Med 1995;155:1741 -1748[Abstract]
  11. Rae LC. Community screening for colorectal cancer in north-eastern New South Wales, 1987-1996. Med J Aust 1998;168:382 -385[Medline]
  12. Winawer SJ, Flehinger BJ, Schottenfeld D, Miller DG. Screening for colorectal cancer with fecal occult blood testing and sigmoidoscopy. J Natl Cancer Inst 1993;85:1311 -1318[Abstract/Free Full Text]
  13. Winawer SJ, Zauber AG, Ho MN, et al. Prevention of colorectal cancer by colonoscopic polypectomy. N Engl J Med 1993;329:1977 -1981[Abstract/Free Full Text]
  14. Kewenter J, Brevinge H, Engaras B, Haglind E. The yield of flexible sigmoidoscopy and double-contrast barium enema in the diagnosis of neoplasms in the large bowel in patients with a positive hemoccult test. Endoscopy 1995;27:159 -163[Medline]
  15. Towler B, Irwig L, Gasziou P, et al. A systematic review of the effects of screening for colorectal cancers using the fecal occult blood test, hemoccult. Br Med J 1998;315:559 -565[Free Full Text]
  16. Jarvinen HJ, Mecklin J-P, Sistonen P. Screening reduces colorectal cancer rate in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 1996;108:1405 -1411[Medline]
  17. Winawer SJ, Zauber AG, O'Brien MJ, et al. The national polyp study: design, methods, and characteristics of patients with newly diagnosed polyps. Cancer 1992;70:1236 -1245[Medline]
  18. Black WC, Welch HG. Screening for disease. AJR 1997;168:3 -11[Abstract/Free Full Text]
  19. Parsonnet J, Axon TR. Principles of screening and surveillance. Am J Gastroenterol 1996;91:847 -849[Medline]
  20. Marshall JR, Fay D, Lance P. Potential costs of flexible sigmoidoscopy-based colorectal cancer screening. Gastroenterology 1996;111:1411 -1417[Medline]
  21. U.S. Congress, Office of Technology Assessment.Cost-effectiveness of colorectal cancer screening in average-risk adults . BP-H-146 Washington, DC: April 1995
  22. Koretz RL. Universal colonic polypectomy: a rocky cost for smoothing bumps in the road (abstr). Gastroenterology 1993;104[suppl]:A15
  23. Ransohoff DF, Lang CA, Kuo SH. Colonoscopic surveillance after polypectomy: considerations of cost effectiveness. Ann Intern Med 1991;114:177 -182
  24. Glick S, Wagner JL, Johnson CD. Cost-effectiveness of double-contrast barium enema in screening for colorectal cancer. AJR 1998;170:629 -636[Free Full Text]
  25. England WL, Halls JJ, Hunt VB. Strategies for screening for colorectal carcinoma. Med Decis Making 1989;9:3 -13
  26. 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[Medline]
  27. Weller D, Moss J, Hiller J, Thomas D, Edwards J. Screening for colorectal cancer: what are the costs? Int J Technol Assess Health Care 1995;11:26 -39[Medline]
  28. Eddy DM, Nugent FW, Eddy JF, et al. Screening for colorectal cancer in a high-risk population: results of a mathematical model. Gastroenterology 1987;92:682 -692[Medline]
  29. Ahlquist DA. Fecal occult blood testing for colorectal cancer: can we afford to do this? Gastroenterol Clin North Am 1997;26:41 -55[Medline]
  30. Myers RE, Balshem AM, Wolf TA, Ross EA, Millner L. Screening for colorectal neoplasia: physicians' adherence to complete diagnostic evaluation. Am J Public Health 1993;83:1620 -1622[Abstract/Free Full Text]
  31. Lurie JD, Welch HG, Diagnostic testing following fecal occult blood screening in the elderly. J Natl Cancer Inst 1999;91:1641 -1646[Abstract/Free Full Text]
  32. Levin B, Hess K, Johnson C. Screening for colorectal cancer: a comparison of 3 fecal occult blood tests. Arch Intern Med 1997;157:970 -976[Abstract]
  33. Medicare Special Bulletin. 1999 Medicare physician fee schedule report. Available at:www.xact.org/pdf/par992k.pdf. Accessed November 4, 1999
  34. Cannon-Albright LA, Bishop T, Samowitz W, DiSario JA, Lee R, Burt RW. Colonic polyps in an unselected population: prevalence, characteristics and associations. Am J Gastroenterol 1994;89:827 -831[Medline]
  35. Kim LS, Rockey DC. Prospective comparison of patient's perception of pain during barium enema, flexible sigmoidoscopy and colonoscopy (abstr). Gastroenterology 1997;112:A22
  36. Sharma VK, Chockalingham S, Clark V, et al. Randomized controlled comparison of two forms of preparation for screening flexible sigmoidoscopy. Am J Gastroenterol 1997;92:809 -811[Medline]
  37. Osgard E, Jackson JL, Strong J. A randomized trial comparing three methods of bowel preparation for flexible sigmoidoscopy. Am J Gastroenterol 1998;93:1126 -1130[Medline]
  38. Lehman GA, Buchner DM, Lappas JC. Anatomical extent of fiberoptic sigmoidoscopy. Gastroenterology 1983;84:803 -808[Medline]
  39. Painter J, Saunders DB, Williams CB, Pitt R, Bladen J. Depth of insertion at flexible sigmoidoscopy: implications for colorectal cancer screening and instrument design. Endoscopy 1999;31:227 -231[Medline]
  40. Jensen J, Kewenter J, Swedenbord J. The anatomic range of examination of fiberoptic rectosigmoidoscopy (60 centimetres). Scand J Gastroenterol 1992;27:842 -844[Medline]
  41. Kadakia SC, Wrobleski CS, Kadakia AS, Meier NJ. Prevalence of proximal colonic polyps in average-risk asymptomatic patients with negative fecal occult blood tests and flexible sigmoidoscopy. Gastrointest Endosc 1996;44:112 -117[Medline]
  42. Lemmel TG, Haseman JH, Rex DK, Rahmani E. Neoplasia distal to the splenic flexure in patients with proximal colon cancer. Gastrointest Endosc 1996;44:109 -111[Medline]
  43. Demers RY, Severson RK, Schottenfeld D, Lazar L. Incidence of colorectal adenocarcinoma by anatomic subsite: an epidemiologic study of time trends and racial differences in the Detroit, Michigan area. Cancer 1997;79:441 -447[Medline]
  44. Nelson RL, Dollear T, Freels S, Persky V. The relation of age, race, and gender to the subsite location of colorectal carcinoma. Cancer 1997;80:193 -197[Medline]
  45. Hahn DL. Sex and race are risk factors for colorectal cancer within reach of the sigmoidoscope. J Fam Pract 1990;30:409 -416[Medline]
  46. Johnson H, Margolis I, Wise L. Site-specific distribution of large-bowel adenomatous polyps: emphasis on ethnic differences. Dis Colon Rectum 1988;31:258 -260[Medline]
  47. Offerhaus GJA, Giardiello FM, Tersmette KWF, et al. Ethnic differences in the anatomical location of colorectal adenomatous polyps. Int J Cancer 1991;49:641 -644[Medline]
  48. Ozick LA, Jacob L, Donelson SS, Agarwal SK, Freeman HP. Distribution of adenomatous polyps in African-Americans. Am J Gastroenterol 1995;90:758 -760[Medline]
  49. Lewis JD, Asch DA. Barriers to office-based screening sigmoidoscopy: does reimbursement cover costs? Ann Intern Med 1999;130:525 -530[Abstract/Free Full Text]
  50. Rex DW, Lehman GA, Hawes RH, O'Connor KW, Smith JJ. Performing screening flexible sigmoidoscopy using colonoscopes: experience in 500 patients. Gastrointest Endosc 1990;36:486 -488[Medline]
  51. Kaczmarek RG, Moore RM, McCrohan J, et al. Multi-state investigation of the actual disinfection/sterilization of endoscopes in health care facilities. Am J Med 1992;92:257 -261[Medline]
  52. Spach DH, Silverstein FE, Stamm WE. Transmission of infection by gastrointestinal endoscopy and bronchoscopy. Ann Intern Med 1993;118:117 -128[Abstract/Free Full Text]
  53. Bronowicki J-P, Venard V, Botte C, et al. Patient-to-patient transmission of hepatitis C virus during colonoscopy. N Engl J Med 1997;337:237 -240[Free Full Text]
  54. Ponchon T. Transmission of hepatitis C and prior diseases through digestive endoscopy: evaluation of risk and recommended practices. Endoscopy 1997;29:199 -202[Medline]
  55. Deva AK, Vickery K, Zou J, et al. Detection of persistent vegetative bacteria and amplified viral nucleic acid from in-use testing of gastrointestinal endoscopes. J Hosp Infect 1998;39:149 -157[Medline]
  56. Rozen P, Somjen GJ, Baratz M, Kimel R, Arber N, Gilat T. Endoscope-induced colitis: description, probable cause by glutaraldehyde, and prevention. Gastrointest Endosc 1994;40:547 -553[Medline]
  57. West AB, Kuan S-F, Bennick M, Lagarde S. Glutaraldehyde colitis following endoscopy: clinical and pathological features and investigation of an outbreak. Gastroenterology 1995;108:1250 -1255[Medline]
  58. Birnbaum BA, Gordon RB, Jacobs JE. Glutaraldehyde colitis: radiologic findings. Radiology 1995;195:131 -134[Abstract/Free Full Text]
  59. Hanson JM, Plusa SM, Bennett MK, Browell DA, Cunliffe WJ. Glutaraldehyde as a possible cause of diarrhea after sigmoidoscopy. Br J Surg 1998;85:1385 -1387[Medline]
  60. Lewis DL, Arens M. Resistance of microorganisms to disinfection in dental and medical devices. Nat Med 1995;1:956 -958[Medline]
  61. Berry DP, Clarke P, Hardcastle JD, Vellacott KD. Randomized trial of the addition of flexible sigmoidoscopy to faecal occult blood testing for colorectal neoplasia population screenings. Br J Surg 1997;84:1274 -1276[Medline]
  62. Foley DP, Dunne P, Dervan PJ, Callaghan TWO, Crowe J, Lennon JR. Left-sided colonoscopy and haemoccult screening for colorectal neoplasia. Eur J Gastroenterol Hepatol 1992;4:925 -936
  63. Rozen P, Ron E, Fireman Z, et al. The relative value of fecal occult blood tests and flexible sigmoidoscopy in screening for large bowel neoplasia. Cancer 1987;60:2553 -2558[Medline]
  64. Brevinge H, Lindholm E, Buntzen S, Kewenter J. Screening for colorectal cancer with faecal occult blood testing compared with flexible sigmoidoscopy directly in a 55-56 years' old population. Int J Colorectal Dis 1997;12:291 -295[Medline]
  65. Rasmussen M, Kronberg O, Fenger C, Jorgensen OD. Possible advantages and drawbacks of adding flexible sigmoidoscopy to hemoccult-II in screening for colorectal cancer. Scand J Gastroenterol 1999;34:73 -78[Medline]
  66. Blakeborough A, Sheridan MB, Chapman AH. Complications of barium enema examinations: a survey of UK consultant radiologists 1992 to 1994. Clin Radiol 1997;52:142 -148[Medline]
  67. Rex DK. Colonoscopy: a review of its yield for cancers and adenomas by indication. Am J Gastroenterol 1995;90:353 -365[Medline]
  68. Neugut AI, Garbowski GC, Waye JD, et al. Diagnostic yield of colorectal neoplasia with colonoscopy for abdominal pain, change in bowel habits, and rectal bleeding. Am J Gastroenterol 1993;88:1179 -1183[Medline]
  69. Rex DK, Rahmani EY, Haseman JH, Lemmel GT, Kaster S, Buckley JS. Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice. Gastroenterology 1997;112:17 -23[Medline]
  70. Johnson CD, Carlson HC, Taylor WF, Weiland LP. Barium enemas of carcinoma of the colon: sensitivity of double- and single-contrast studies. AJR 1983;140:1143 -1149[Abstract/Free Full Text]
  71. Fork F-T, Lindstom C, Ekelund G. Double contrast examination in carcinoma of the colon and rectum: a prospective clinical series. Acta Radiol 1983;24:177 -187
  72. Thorpe CD, Grayson DJ, Wingfield PB. Detection of carcinoma of the colon and rectum by air contrast enema. Surg Gynecol Obstet 1981;152:307 -309[Medline]
  73. Beggs J, Thomas BM. Diagnosis of carcinoma of the colon by barium enema. Clin Radiol 1983;140:1143 -1149
  74. Reiertsen O, Bakka A, Tronnes S, Gauperaa T. Routine double contrast barium enema and fiberoptic colonoscopy in the diagnosis of colorectal carcinoma. Acta Chir Scand 1988;154:53 -56[Medline]
  75. Kelvin FM, Gardiner R, Vas W, Stevenson GW. Colorectal carcinoma missed on double contrast barium enema study: a problem in perception. AJR 1981;137:307 -313[Abstract/Free Full Text]
  76. Kalra L, Hamlyn AN. Comparative evaluation of investigations for colorectal carcinoma in symptomatic patients. Postgrad Med J 1988;64:666 -668[Abstract]
  77. Stefansson T, Bergman A, Ekbom A, Nyman R, Pahlman L. Accuracy of double contrast barium enema and sigmoidoscopy in the detection of polyps in patients with diverticulosis. Acta Radiol 1994;35:442 -446[Medline]
  78. Baker SR, Alterman DD. False-negative barium enema in patients with sigmoid cancer and coexistent diverticula. Gastrointest Radiol 1985;10:171 -173[Medline]
  79. Stryker SJ, Wolff BG, Culp CE, Libbe SD, Ilstrup DM, MacCarty RL. Natural history of untreated colonic polyps. Gastroenterology 1987;93:1009 -1013[Medline]
  80. Ott DJ, Scharling ES, Chen YM, Gelfand DW, Wu WC. Positive predictive value and posttest probability of diagnosis of colonic polyp on single- and double-contrast barium enema. AJR 1989;153:735 -739[Abstract/Free Full Text]
  81. Weyman PJ, Koehler RE, Zuckerman GR. Resolution of radiographic-endoscopic discrepancies in colon neoplasms. J Clin Gastroenterol 1981;3[suppl 1]:89 -93
  82. Gelfand DW, Chen MYM, Ott DJ. Benign colorectal neoplasms undetected by colonoscopy. Gastrointest Radiol 1992;17:344 -346[Medline]
  83. Miller RE, Lehman G. Polypoid colonic lesions undetected by endoscopy. Radiology 1978;129:295 -297[Abstract]
  84. Laufer I, Smith NCW, Mullens JE. The radiological demonstration of colorectal polyps undetected by endoscopy. Gastroenterology 1976;70:167 -170[Medline]
  85. Glick SN, Teplick SK, Balfe EM, et al. Large colonic neoplasms missed by endoscopy. AJR 1989;152:513 -517[Abstract/Free Full Text]
  86. Kjaergard H, Nordkild P, Hennild V, Pedersen M, Geerdsen J. Follow-up study after colorectal polypectomy: the predictive value of a negative double-contrast barium enema. Scand J Gastroenterol 1986;21:353 -356[Medline]
  87. Steine S, Stordahl A, Lunde OC, Loken K, Laerum E. Double-contrast barium enema versus colonoscopy in the diagnosis of neoplastic disorders: aspects of decision-making in general practice. Fam Pract 1993;10:288 -291[Abstract/Free Full Text]
  88. Williams CB, Macrae FA, Bartram CI. A perspective study of diagnostic methods in adenoma follow-up. Endoscopy 1982;14:74 -78[Medline]
  89. Schoen RE, Gerber LD, Margulies C. The pathologic measurement of polyp size is preferable to the endoscopic estimate. Gastrointest Endosc 1997;46:492 -496[Medline]
  90. Saito Y, Slezak P, Rubio C. The diagnostic value of combining flexible sigmoidoscopy and double-contrast barium enema as a one-stage procedure. Gastrointest Radiol 1989;14:357 -359[Medline]
  91. Norfleet RG, Ryan ME, Wyman JB, et al. Barium enema versus colonoscopy for patients with polyps found during flexible sigmoidoscopy. Gastrointest Endosc 1991;37:531 -534[Medline]
  92. Rex DK, Lehman GA, Lappas JC, Miller RE. Sensitivity of double-contrast barium study for left-colon polyps. Radiology 1986;158:69 -72[Abstract/Free Full Text]
  93. Anderson N, Cook HB, Coates R. Colonoscopically detected colorectal cancer missed on barium enema. Gastrointest Radiol 1991;16:123 -127[Medline]
  94. Markus JB, Somers S, O'Malley BP, Stevenson GW. Double-contrast barium enema studies: effect of multiple reading on perception error. Radiology 1990;175:155 -156[Abstract/Free Full Text]
  95. Waye JD, Kahn O, Auerbach ME. Complications of colonoscopy and flexible sigmoidoscopy. Gastrointest Endosc Clin North Am 1996;6:343 -377[Medline]
  96. Waye JD, Lewis BS, Yessayan S. Colonoscopy: a prospective report of complications. J Clin Gastroenterol 1992;15:347 -351[Medline]
  97. Zubarik R, Fleischer DE, Mastropietro C, et al. Prospective analysis of complications 30 days after outpatient colonoscopy. Gastrointest Endosc 1999;50:322 -328[Medline]
  98. Steine S. Which hurts the most? A comparison of pain rating during double-contrast barium enema examination and colonoscopy. Radiology 1994;191:99 -101[Abstract/Free Full Text]
  99. Van Ness MM, Chobanian SJ, Winters C Jr, Diehl AM, Esposito RL, Cattau EL Jr. A study of patient acceptance of double-contrast barium enema and colonoscopy: which procedure is preferred by patients? Arch Intern Med 1987;147:2175 -2176[Abstract]
  100. Eisen GM, Zubarik R, Carroll J, et al. Work time lost subsequent to outpatient colonoscopy as determined by thirty day follow-up study (abstr). Am J Gastroenterol 1987;92:1742
  101. Obrecht WF, Wu WC, Gelfand DW, Ott DJ. The extent of successful colonoscopy: a second assessment using modern equipment. Gastrointest Radiol 1984;9:161 -162[Medline]
  102. Abrams JS. A second look at colonoscopy: indications, failures, costs. Arch Surg 1982;117:913 -917[Abstract]
  103. Brady AP, Stevenson GW, Stevenson I. Colorectal cancer overlooked at barium enema and colonoscopy: a continuing perceptual problem. Radiology 1994;192:373 -378[Abstract/Free Full Text]
  104. Meyer GS, Jacoby I, Krakauer H, et al. Gastroenterology workforce modeling. JAMA 1996;276:689 -694[Abstract]
  105. Rex DK, Cutler CS, Lemmel GT, et al. Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 1997;112:24 -28[Medline]
  106. Hixon LJ, Fennerty MB, Sampliner RE, McGee DL, Garewal H. Two-year incidence of colon adenomas developing after tandem colonoscopy. Am J Gastroenterol 1994;89:687 -691[Medline]
  107. Warneke J, Petrelli N, Herrera L, Nava H. Accuracy of colonoscopy for the detection of colorectal polyps. Dis Colon Rectum 1992;35:981 -985[Medline]
  108. Glick SN. What is the evidence-based specificity of colonoscopy? AJR 1998;170:1104[Medline]
  109. Hancock JH, Talbot RW. Accuracy of colonoscopy in localization of colorectal cancer. Int J Colorect Dis 1995;10:140 -141[Medline]
  110. Frager DH, Frager JD, Wolf EL, Beneventano TC. Problems in the colonoscopic localization of tumors: continued value of the barium enema. Gastrointest Radiol 1987;12:343 -346[Medline]
  111. Leung KL, Yiu RYC, Lai PBS, et al. Laparoscopic-assisted resection of colorectal carcinoma: five year audit. Dis Colon Rectum 1999;42:327 -333[Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us