Failures of diagnosis of colorectal cancer by colonoscopy, barium enema, and flexible sigmoidoscopy have been demonstrated using various techniques. A relative assessment of these diagnostic tests for patients with colorectal cancer has not been reported. This study was designed to determine relative rates of failures for these tests when applied to diagnosis of colorectal cancer.
We created a database of patients with colorectal cancer diagnosed between 2000 and 2005. Records were reviewed for the results of colonoscopy, barium enema, and flexible sigmoidoscopy in the 3 years prior to diagnosis. An examination that was negative for cancer with no immediate follow-up was defined as a failure of diagnosis, either from inaccurate observation, failure to examine the entire colon, or failure of timely follow-up. The failure rates were compared.
Three hundred seventy-nine patients, who had 421 examinations, were analyzed. The diagnosis of colorectal cancer failed in 60 of 379 patients (16%). These 60 patients had 71 examinations that failed to make the diagnosis: 25 of 282 colonoscopies (9%), 16 of 79 barium enemas (20%), and 30 of 60 flexible sigmoidoscopies (50%). These differences were statistically significant.
Failure rates for colonoscopy, barium enema, and flexible sigmoidoscopy were 9%, 20%, and 50%.
Colorectal cancer (CRC) is the second leading cause of death due to cancer and the third most common cancer in men and women in the US. In 2008, 148,818 new cases of colorectal cancer are projected with 49,960 deaths . Since 1997, when the US agency for Health Care Policy and Research first advocated widespread colon cancer screening, refinements in techniques of optical colonoscopy (OC), barium enema (BE), and flexible sigmoidoscopy (FS) have evolved [2, 3]. Each examination has a failure rate, ranging from a low of 4% for colonoscopy in right-sided cancers to 46% for FS [4–19]. Failed diagnoses are well known to clinicians, but the relative failure rates for the three most widely used screening examinations in community hospitals are not. We tested the null hypothesis that there is no difference in failure rates between colonoscopy, BE, and FS.
Patient records were obtained from the Scripps Green Hospital Cancer Registry. The study was approved by the Scripps Office for the Protection of Research Subjects, Human Subjects Committee at Scripps Clinic, La Jolla, CA, USA. We created a database for all patients diagnosed with biopsy-proven colorectal cancer at Scripps Green Hospital between January 2000 and December 2005 . Patients were excluded if they had a history of inflammatory bowel disease, familial adenomatous polyposis syndrome, or hereditary nonpolyposis colorectal cancer. Medical records were reviewed for: (1) verification of adenocarcinoma of the colon or rectum, TNM stage, and tumor location within the colon; (2) patient’s age and gender; (3) presenting symptoms; (4) examination that resulted in the diagnosis; (5) results of OC, BE, and FS performed within 3 years of the diagnosis; (6) interval between any diagnostic test and the diagnosis of colon cancer; and (7) reason for failure to detect the cancer. Colonoscopies were performed by staff gastroenterologists and gastroenterology fellows under supervision by staff physicians in the room at all times. All BEs were performed by board-certified radiologists. Flexible sigmoidoscopies were performed by board-certified gastroenterologists, physician’s assistants, and gastroenterology fellows under supervision [21, 22]. The failure rates were a combination of missed lesions on examination and failure to examine the entire colon. The patients were divided into two categories: detected group and failed group. The detected group included patients who were diagnosed with colorectal cancer on the first attempt. The failed group included patients who had one or more diagnostic tests negative for cancer with no immediate follow-up within 3 years of the cancer diagnosis [4, 22–24]. Statistical analysis was performed using the chi-square test.
Three hundred ninety-three patients with colorectal cancer were identified between 2000 and 2005, 379 of whom met the inclusion criteria. Three hundred nineteen (84%) had the diagnosis made at their index diagnostic test, thus constituting the detected group. Sixty (16%) patients had screening with OC, BE, or FS within 3 years prior to their diagnosis that did not identify the cancer and were, thus, the missed group (Fig. 1). A demographic comparison between the two groups revealed no statistical difference in race or gender between the groups, except patients in the missed group were slightly older (Table 1).
The presenting signs and symptoms of the cancers at diagnosis were not appreciably different between the groups, except patients in the failed group more often presented with anemia than patients in the detected group. The mean and median delay to diagnosis was 574 days (19 months), and 671 days (22 months), respectively. The range of missed examinations prior to diagnosis was 61–1,095 days (2–36 months). The cancers that were missed were more likely to be in the cecum (18 cancers) and ascending colon (16 cancers) or sigmoid colon (13 cancers). In total, 209 cancers (55%) were above the junction of the sigmoid and descending colon. The stages of cancer at diagnosis between the missed group and the detected group were not different. Failure rates for OC, BE, and FS were 25/282 (9%), 16/79 (20%), and 30/60 (50%; Table 2). The differences between failure rates for OC versus BE, OC versus FS, and BE versus FS were significant at p < 0.01, p < 0.001, and p < 0.001, respectively. These results disprove the null hypothesis that there is no difference in failure rates between OC, BE, and FS.
Seventy-one screening examinations failed to make the diagnosis of CRC; the reasons for failure fell into four categories: (1) failure to examine the entire colon (30 cases, 42%); (2) inaccurate observation, i.e., the examination was considered normal when in fact a cancer was present (25 cases, 35%); (3) premature dismissal of suboptimal examinations (11 cases, 15%); and (4) failure to follow up advanced adenomas at shorter intervals (five cases, 7%; Table 3). Of the 11 cases that were missed due to premature dismissal of suboptimal examinations, eight of them were due to poor preparations and four were due to examinations that were difficult to interpret for other reasons (spasm in two, excess folds in one, patient noncompliance in one).
In six patients in the missed group, there were multiple examinations performed: three with OC + BE, two FS + OC, and one FS + BE. In each of these patients, both examinations missed the diagnosis. One patient that had an incomplete OC was then screened with BE that detected colon cancer.
In the 60 patients in the missed group, the cancers were eventually detected by the following modalities: nine patients underwent BE for evaluation of new symptoms (six with anemia and bleeding, three with pain–obstruction). Six patients were diagnosed at surgery (three for bleeding, three for pain–obstruction). Six patients underwent FS (all for bleeding). Thirty-nine underwent OC (ten for bowel habit change, 23 for anemia–bleeding, six for pain–obstruction). The most common presenting symptom in the patients in the missed group was anemia (21 patients, 35%). This was followed by overt bleeding (17 patients, 28%), pain–obstruction (12 patients, 20%), and bowel habit change (ten patients, 16%).
Updated recommendations for colon cancer screening and surveillance have recently been published by the American Gastroenterological Association institute from a joint commission from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology . In patients at average risk for colon cancer, multiple options for colon cancer screening are offered. The recommendations are split into tests that primarily detect cancer (guaiac-based fecal occult blood testing, fecal immunochemical testing, or stool DNA test) and tests that detect adenomatous polyps and cancer (flexible sigmoidoscopy, double-contrast barium enema, colonoscopy, or computed tomography (CT) colonography). Fecal-based testing (guaiac testing or immunochemical testing) should be repeated yearly, with the recommendation that positive tests are followed up with colonoscopy. Stool DNA testing interval remains uncertain, as the test is new and sensitivity is uncertain . Flexible sigmoidoscopy or double-contrast barium enema should be repeated every 5 years, and if positive findings are encountered then patients should be referred for colonoscopy. Colonoscopy remains the mainstay for colon cancer screening and should be performed every 10 years in the absence of positive findings. For the first time, CT colonography is included in the recommendations and currently is recommended to be repeated every 5 years, with colonoscopy recommended if patients are found to have one or more polyps greater than 6 mm in size .
It is notable that for each recommended examination, reader experience and proper training is needed in order to increase the sensitivity of the examination to a level sufficient for screening. The limitations of each examination must be taken into account when discussing colorectal cancer screening with patients. In addition, the effectiveness of any single test or combination of tests depends on high rates of adherence to recommended intervals in addition to quality of examination.
We used a case control method to determine relative failure rates of OC, BE, and FS for diagnosis of CRC. We found failure rates of 9%, 20%, and 50%, respectively. The approach demonstrated appreciable deficiencies in each technique while confirming the expectation that OC is less likely to miss a CRC than BE or FS. The data provide comparative estimates of the failure rates which may aid clinicians in deciding which technique to employ.
Several studies have attempted to estimate failure rates of OC from large-population databases, tandem colonoscopies, and comparison to CT colonography. In a population-based Canadian study, the miss rate for right-sided colon cancer by OC was 4% . Controlled studies of back-to-back OC revealed a miss rate of 6% for advanced lesions equal to or greater than 1 cm [5, 6]. Comparisons of OC to CT colonography reported 12–17% missed advanced adenomas or cancers ≥1 cm [7–10], with our study finding a failure rate of 9% for OC in detecting cancer. Failure rates for BE have been studied by comparison to OC, with failure rates of 8–52% for adenomas and cancer [11–14, 28]. Some of the shortcomings of BE may be overcome by CT colonography [7–10]. Studies of FS reveal that 32–46% of advanced neoplasia failed diagnosis when the colon was only examined to the junction of the sigmoid and descending colon [15, 16, 29, 30]. It is an inherent shortcoming of flexible sigmoidoscopy that right-sided cancers may be missed on these examinations, as they depend on the presence of distal adenomas to trigger an examination of the entire colon. We still considered FS to have missed the cancer when it was normal since it has been recommended by experts as a possible screening tool .
Four main reasons for failure of diagnosis were identified in this study and provide a framework for improvement strategies. First, it is apparent that the entire colon should be examined with the most sensitive screening test available [7–16]. Although proctosigmoidoscopy screening was demonstrated to decrease mortality compared to no screening [29, 30], this observation should not be considered justification for routine screening as the only diagnostic approach when full-colon examinations are available. Second, 25 cancers were missed due to inaccurate interpretation, i.e., the examination was considered normal when a cancer was present. A variety of new technologies and novel techniques to improve the accuracy of diagnostic tests have been reported [31–35]. Third, if the examination is limited by poor preparation or patient difficulty with cooperation, the examination should be repeated after more preparation or with more sedation. Lastly, advanced adenomas that are incompletely removed or removed piecemeal should be reexamined in 2–6 months, as recommended by the American Society for Gastrointestinal Endoscopy and others [36–39].
Additional factors associated with risks for missing lesions include right-sided and transverse cancers, advanced age, female gender, diverticular disease, examination in a private office, and examination by an internist or family physician [40, 41]. Our study showed anemia to be the most common presentation when a cancer was missed. There were no significant associated factors in our study for missed lesions other than location.
Weaknesses of this study include that it is observational and retrospective. The primary criterion for entry into the study was a pathologic diagnosis of colorectal cancer; thus, the inclusion of some patients could reflect a selection bias. The greatest concern with the possible selection bias is that some cancers defined as missed lesions may have in fact been de novo cancers. However, we do not feel that de novo lesions would have made up a large enough percentage to significantly alter the results. Also, multiple examiners performed these procedures, which made it impossible to determine if the quality of the examination contributed to the missed diagnosis. Nevertheless, this study provides a quantitative estimate of the relative failure rates of community-available diagnostic tests and identification of four major causes of diagnostic failure.
In conclusion, our study found failure rates of 9%, 20%, and 50% for OC, BE, and FS, respectively, in diagnosis of colorectal cancer in a community hospital. Diagnosis of CRC is not nearly as reliable as needed.
Fecal occult blood test
American Cancer Society. Colorectal cancer facts and figures special edition 2008. Atlanta: American Cancer Society; 2008.
Winawer S, Fletcher R, Rex D, et al. Colorectal cancer screening and surveillance: clinical guidelines and rationale—update based on new evidence. Gastroenterology 2003;124:544–60. doi:10.1053/gast.2003.50044.
Ransohoff D. Colon cancer screening in 2005: status and challenges. Gastroenterology 2005;128:1685–95. doi:10.1053/j.gastro.2005.04.005.
Bressler B, Paszat L, Vinden C, et al. Colonoscopic miss rates for right-sided colon cancer: a population-based analysis. Gastroenterology 2004;127:452–6. doi:10.1053/j.gastro.2004.05.032.
Rex D, Cutler C, Lemmel G, et al. Colonoscopic miss rates of adenomas determined by back-to-back colonoscopies. Gastroenterology 1997;112:24–8. doi:10.1016/S0016-5085(97)70214-2.
Van Rijn J, Reitsma J, Stoker J, et al. Polyp miss rate determined by tandem colonoscopy: a systematic review. Am J Gastroenterol. 2006;101:343–50. doi:10.1111/j.1572-0241.2006.00390.x.
Van Gelder R, Nio CY, Florie J, et al. Computed tomographic colonography compared with colonoscopy in patients at increased risk for colorectal cancer. Gastroenterology 2004;127:41–8. doi:10.1053/j.gastro.2004.03.055.
Pickhardt P, Nugent P, Mysliwiec P, et al. Location of adenomas missed by optical colonoscopy. Ann Intern Med. 2004;141:352–9.
Kim D, Pickhardt P, Taylor A, et al. CT colonography versus colonoscopy for the detection of advanced neoplasia. N Engl J Med. 2007;357:1403–12. doi:10.1056/NEJMoa070543.
Pickhardt P, Choi J, Hwang I, et al. Computed tomography virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med. 2003;349:2191–200. doi:10.1056/NEJMoa031618.
Rex D, Rahmani E, Haseman J, et al. Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice. Gastroenterology 1997;112:117–23.
Schrock T. Colonoscopy versus barium enema in the diagnosis of colorectal cancers and polyps. Gastrointest Endosc Clin N Am. 1993;3:585–610.
Winawer SJ, Stewart ET, Zauber AG, et al. A comparison of colonoscopy and double-contrast barium enema for surveillance after polypectomy. N Engl J Med. 2000;342:1766–72. doi:10.1056/NEJM200006153422401.
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;14:357–9. doi:10.1007/BF01889237.
Lieberman D, Weiss D, Bond J, et al. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. N Engl J Med. 2000;343:162–8. doi:10.1056/NEJM200007203430301.
Imperiale T, Wagner D, Lin C, et al. Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med. 2000;343:169–74. doi:10.1056/NEJM200007203430302.
Leaper M, Johnston MJ, Barclay M, et al. Reasons for failure to diagnose colorectal carcinoma at colonoscopy. Endoscopy 2004;36:499–503. doi:10.1055/s-2004-814399.
Shehadeh I, Rebala S, Kumar R, et al. Retrospective analysis of missed advanced adenomas on surveillance colonoscopy. Am J Gastroenterol. 2002;97:1143–7. doi:10.1111/j.1572-0241.2002.05680.x.
Regula J, Rupinski M, Kraszewska E, et al. Colonoscopy in colorectal-cancer screening for detection of advanced neoplasia. N Engl J Med. 2006;355:1863–72. doi:10.1056/NEJMoa054967.
Strum WB. Impact of a family history of colorectal cancer on age at diagnosis, anatomic location, and clinical characteristics of colorectal cancer. Int J Gastrointest Cancer. 2005;35:121–6. doi:10.1385/IJGC:35:2:121.
Strum WB. Impact of a family history of colorectal cancer on the prevalence of advanced adenomas of the rectosigmoid colon at flexible sigmoidoscopy in 3147 asymptomatic patients. Dig Dis Sci. 2006;51:2048–52. doi:10.1007/s10620-006-9660-9.
Strum WB. Impact of adenoma size in distal colon on risk for advanced adenoma of the proximal colon. Dig Dis Sci. 2006;51:2064–7. doi:10.1007/s10620-005-9066-0.
Chen T, Yen M, Lai M, et al. Evaluation of a selective screening for colorectal carcinoma: the Taiwan Multicenter Cancer Screening (TAMCAS) project. Cancer 1999;86:1116–28. doi:10.1002/(SICI)1097-0142(19991001)86:7<1116::AID-CNCR4>3.0.CO;2-D.
Winawer S, Zauber A, Ho M, et al. Prevention of colorectal cancer by colonoscopic polypectomy. N Engl J Med. 1993;329:1977–81. doi:10.1056/NEJM199312303292701.
Levin B, Lieberman DA, McFarland B, Smith RA, Brooks D, et al. Screening and surveillance for early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 2008;134:1570–95. doi:10.1053/j.gastro.2008.02.002.
Imperiale TF, Ransohoff DF, Itzhowitz SH, et al. Fecal DNA versus fecal occult blood for colorectal-cancer screening in an average-risk population. N Engl J Med. 2004;351:2704–14. doi:10.1056/NEJMoa033403.
Moravec M, Lieberman D, Holub J, et al. Rate of advanced pathologic features in 6–9 mm polyps in patients referred for colonoscopic screening. Gastrointest Endosc. 2007;65:822. doi:10.1016/j.gie.2007.03.1064.
Johnson C, Carlson H, Taylor W, et al. Barium enemas of carcinoma of the colon: sensitivity of double- and single-contrast studies. Am J Roentgenol. 1983;140:1143–9.
Selby JV, Friedman GD, Quesenberry CP, et al. A case-control study of screening sigmoidoscopy and mortality from colorectal cancer. N Engl J Med. 1992;326:653–7.
Doria-Rose VP, Levin TR, Selby JV, et al. The incidence of colorectal cancer following a negative screening sigmoidoscopy: implications for screening interval. Gastroenterology 2004;127:714–22. doi:10.1053/j.gastro.2004.06.048.
Rex DK. Colonoscopic withdrawal technique is associated with adenoma miss rates. Gastrointest Endosc. 2000;51:33–6. doi:10.1016/S0016-5107(00)70383-X.
Barclay RL, Vicari JJ, Doughty AS, et al. Colonoscopic withdrawal times and adenoma detection during screening colonoscopy. N Engl J Med. 2006;355:2533–41. doi:10.1056/NEJMoa055498.
Lieberman D. A call to action—measuring the quality of colonoscopy. N Engl J Med. 2006;355:2588–9. doi:10.1056/NEJMe068254.
East J, Suzuki N, Arebi N, et al. Position changes improve visibility during colonoscope withdrawal: a randomized, blinded crossover trial. Gastrointest Endosc. 2007;65:263–9. doi:10.1016/j.gie.2006.04.039.
East J, Saunders B, Burling D, et al. Surface visualization at CT colonography simulated colonoscopy: effect of varying field of view and retrograde view. Am J Gastroenterol. 2007;102:2529–935. doi:10.1111/j.1572-0241.2007.01429.x.
Soetikno R, Kaltenbach T, Rouse R, et al. Prevalence of nonpolypoid (flat and depressed) colorectal neoplasms in asymptomatic and symptomatic adults. J Am Med Assoc. 2008;299:1027–35. doi:10.1001/jama.299.9.1027.
Davila RE, Rajan E, Baron TH, Adler DG, Egan JV, Faigel DO, et al. ASGE guideline: colorectal cancer screening and surveillance. Gastrointest Endosc. 2006;63:546–57. doi:10.1016/j.gie.2006.02.002.
Winawer S, Zauber A, Fletcher R, et al. Guidelines for colonoscopic surveillance after polypectomy: a consensus update by the US multi-society task force on colorectal cancer and the American Cancer Society. Gastroenterology 2006;130:1872–85. doi:10.1053/j.gastro.2006.03.012.
Atkin WS, Morson BC, Cuzick J. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas. N Engl J Med. 1992;326:658–62.
Bressler B, Paszat LF, Chen Z, et al. Rates of new or missed colorectal cancers after colonoscopy and their risk factors: a population-based analysis. Gastroenterology 2007;132:96–102. doi:10.1053/j.gastro.2006.10.027.
Shah HA, Paszat LF, Saskin R, et al. Factors associated with incomplete colonoscopy: a population-based study. Gastroenterology 2007;132:2297–303. doi:10.1053/j.gastro.2007.03.032.
We gratefully acknowledge Sharon Reid and Clarence Sandbakken of the Scripps Clinic Cancer Registry and Dr. James Koziol and Anne Feng of the Biostatistics Department of The Scripps Research Institute for their invaluable assistance.
About this article
Cite this article
Frenette, C.T., Strum, W.B. Relative Rates of Missed Diagnosis for Colonoscopy, Barium Enema, and Flexible Sigmoidoscopy in 379 Patients with Colorectal Cancer. J Gastrointest Canc 38, 148–153 (2007). https://doi.org/10.1007/s12029-008-9027-x
- colon cancer
- barium enema
- flexible sigmoidoscopy