Portal de Periódicos da CAPES

Prevalence of Polyps Greater Than 9 mm in a Consortium of Diverse Clinical Practice Settings in the United States

2005; Elsevier BV; Volume: 3; Issue: 8 Linguagem: Inglês

10.1016/s1542-3565(05)00405-2

1542-7714

David A. Lieberman, Jennifer L. Holub, Glenn M. Eisen, Dale F. Kraemer, Cynthia D. Morris,

Gastric Cancer Management and Outcomes

Background & Aims: Colonoscopy is often performed with the goal of identification of patients with serious colon neoplasia. We determined the prevalence of colon masses or polyps greater than 9 mm on the basis of age, gender, race, and procedure indication in diverse clinical practice settings and compared occurrence in patients receiving colonoscopy for screening, surveillance, or evaluation of symptoms. Methods: We obtained patient demographics, procedure indication, and endoscopic findings from colonoscopy reports in the Clinical Outcomes Research Initiative data repository, which receives endoscopy reports from 73 diverse practice sites in the United States. A multivariate model was developed to measure risk variables for a mass or polyps >9 mm. Absolute risk was calculated in the model on the basis of the number needed to endoscope (NNE) to identify 1 patient with a mass or polyp >9 mm. Results: From 2000–2002, colonoscopies in 141,413 unique patients were analyzed. Sixty-nine percent of the reports came from private practice (nonacademic) settings. Increasing age, male gender, and black race were associated with increased risk of mass or polyps >9 mm. In the 50- to 59-year-old average-risk group, 28 women and 18 men would need to have screening colonoscopy to identify 1 patient with a mass/polyp >9 mm. Patients with positive fecal occult blood test results, hematochezia, and anemia had lower NNE, whereas men older than 60 years receiving adenoma surveillance and patients with irritable bowel symptoms had similar NNE compared with average-risk subjects. Conclusions: The prevalence of a colon lesion >9 mm varies on the basis of age, gender, race, and procedure indication. Understanding utilization and outcomes can lead to more optimal use of colonoscopy. Background & Aims: Colonoscopy is often performed with the goal of identification of patients with serious colon neoplasia. We determined the prevalence of colon masses or polyps greater than 9 mm on the basis of age, gender, race, and procedure indication in diverse clinical practice settings and compared occurrence in patients receiving colonoscopy for screening, surveillance, or evaluation of symptoms. Methods: We obtained patient demographics, procedure indication, and endoscopic findings from colonoscopy reports in the Clinical Outcomes Research Initiative data repository, which receives endoscopy reports from 73 diverse practice sites in the United States. A multivariate model was developed to measure risk variables for a mass or polyps >9 mm. Absolute risk was calculated in the model on the basis of the number needed to endoscope (NNE) to identify 1 patient with a mass or polyp >9 mm. Results: From 2000–2002, colonoscopies in 141,413 unique patients were analyzed. Sixty-nine percent of the reports came from private practice (nonacademic) settings. Increasing age, male gender, and black race were associated with increased risk of mass or polyps >9 mm. In the 50- to 59-year-old average-risk group, 28 women and 18 men would need to have screening colonoscopy to identify 1 patient with a mass/polyp >9 mm. Patients with positive fecal occult blood test results, hematochezia, and anemia had lower NNE, whereas men older than 60 years receiving adenoma surveillance and patients with irritable bowel symptoms had similar NNE compared with average-risk subjects. Conclusions: The prevalence of a colon lesion >9 mm varies on the basis of age, gender, race, and procedure indication. Understanding utilization and outcomes can lead to more optimal use of colonoscopy. Colonoscopy is often performed in patients with and without gastrointestinal (GI) symptoms with the goal of identifying patients with serious neoplasia or colorectal cancer (CRC). Although this is not the only important outcome, it is one of the most common reasons for performing the examination. Guidelines for the use of colonoscopy have been published by a Multi-Society Task Force 1Rex D.K. Bond J.H. Winawer S. et al.Quality in the technical performance of colonoscopy and the continuous quality improvement process for colonoscopy recommendations of the US Multi-Society task force on colorectal cancer.Am J Gastroenterology. 2002; 97: 1296-1308Crossref PubMed Scopus (869) Google Scholar and the American Society for Gastrointestinal Endoscopy, 2ASGEThe role of colonoscopy in the management of patients with colonic neoplasia.Gastsrointest Endosc. 1999; 50: 921-924Abstract Full Text Full Text PDF PubMed Scopus (17) Google Scholar on the basis of expert consensus. The utilization and outcomes of colonoscopy in clinical practice settings are uncertain. This information could inform decisions on resource utilization. The Clinical Outcomes Research Initiative (CORI) was established in 1995 to study utilization and outcomes of endoscopy in diverse practice settings, representing 73 practice sites in the United States. The purpose of this analysis was to ascertain the prevalence of serious neoplasia on the basis of procedure indication, age, gender, race/ethnicity, and indication within this consortium. We defined serious neoplasia as a suspected malignant mass or colon polyp or growth greater than 9 mm in diameter. This was the first analysis of colonoscopy outcomes in diverse clinical practice settings. The CORI consortium includes 73 practice sites in 24 states. Participating physicians use a structured, computerized endoscopic report generator to produce their endoscopic reports. The data file from that report is transmitted electronically to a central data repository. All complete colonoscopy reports received between January 1, 2000 and August 15, 2002 were included in this analysis. We excluded reports in patients younger than 20 years old. Race and ethnicity became a mandatory field in January 2001. Before that date, report of race was not always completed. During this period, 12.3% of women and 12.5% of men did not have recorded race or ethnicity and were excluded from this analysis. Patients with known or suspected inflammatory bowel disease were excluded. Also excluded during this period were 1.7% of reports because of incomplete information, including the absence of indications, findings, or demographic information. Indications for procedures were captured in the database. In many cases, there was more than one indication for the procedure. There was considerable overlap among the indications of abdominal pain, change in bowel habits, diarrhea, and constipation. We defined an irritable bowel syndrome (IBS) cluster to include 1 or more of the following symptoms: abdominal pain, bloating, change in bowel habits, diarrhea, or constipation, excluding Crohn's disease, ulcerative colitis, weight loss, or GI bleeding (anemia/iron deficiency, positive fecal occult blood test [FOBT] result, hematochezia, or melena). The primary analysis focused on patients in whom a key end point of colonoscopy would be the discovery of serious neoplasia. Patients were assigned to 1 of 3 categories: screening (including asymptomatic screening, family history of CRC or polyps, and positive FOBT result); surveillance (including surveillance after removal of prior adenomas or cancer); and evaluation of signs or symptoms (including hematochezia, melena, anemia, weight loss, and IBS cluster). CORI reports do not systematically include final pathology results, which arrive several days after the completion of the procedure. In this analysis, we have assumed that one important reason for performing colonoscopy is to identify or rule out serious colonic neoplasia. Prior literature has suggested that colonic polyps or masses greater than 9 mm in size are likely to be neoplastic adenomas and are regarded as significant findings. 3Lieberman D.A. Weiss D.G. Bond J.H. et al.Use of colonoscopy to screen asymptomatic adults for colorectal cancer.N Engl J Med. 2000; 343: 162-168Crossref PubMed Scopus (1626) Google Scholar, 4Imperiale T.F. Wagner D.R. Lin C.Y. et al.Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings.N Engl J Med. 2000; 343: 169-174Crossref PubMed Scopus (926) Google Scholar On the basis of this rationale, we defined the key end point as a suspected malignant mass or polyp greater than 9 mm. A multivariate logistic regression analysis was performed to test indications for colonoscopy and confounding variables for prediction of suspected neoplasia. Backwards stepwise selection was used with a retention criterion of .01. The adjusted relative risk (RR) of each outcome was separately calculated with 95% confidence intervals (CIs). With the exception of age, each of the predictor variables was categorized as a dichotomous variable, and the significance of each was assessed by using a likelihood ratio test statistic obtained from a logistic regression model. Comparison of demographic data was performed with χ2 tests. All analyses were performed by using SAS software (SAS Institute, Inc, Cary, NC). We estimated the risk of a mass or polyps greater than 9 mm and calculated the number needed to endoscope (NNE) to identify 1 patient with this end point on the basis of the logistic multivariate model. This regression model is used to generate the estimated probability of the outcome occurring given specified levels of the variables in the model. The NNE is the reciprocal of the estimated probability. This calculation permits meaningful comparisons of risk by age, gender, and procedure indication. Between January 1, 2000 and August 15, 2002, colonoscopy examination reports from 141,413 unique adult patients were received during the 30-month period. Of these patients, 69% were seen in private practice settings, 18% in academic universities, and 13% in Veterans Affairs (VA) medical centers. The demographic characteristics of this cohort are described in Table 1. In non-VA sites, 55% of all colonoscopies were performed in women. Overall, 19.2% of patients in this colonoscopy cohort were age 50 years or younger at the time of the procedure, including 21.3% of women and 17.2% of men; 6.6% of all patients were 80 years or older, including 7.0% of women and 6.3% of men. White non-Hispanics accounted for 84% of patients. As compared with the US census, significantly fewer colonoscopies were performed in all racial or ethnic minority groups including black non-Hispanics, Hispanics, and Asian/Pacific Islanders (P < .0001 for each group).Table 1Patient Demographics (n = 141,413)FemaleMaleProportion female excluding VA2000 US censusaIn the 2000 US Census, Hispanic ethnicity was collected and analyzed independently from race (except for white non-Hispanic and multiracial non-Hispanic). Therefore the data are not directly comparable, except for Hispanic, but represent the proportion by race, not ethnicity.N% of total femaleN% of total maleAge (y) 9 mm< 50 y50–59 y60–69 y70–79 y≥ 80 yFemale (N = 14,936)Male (N = 12,230)Female (N = 20,116)Male (N = 20,862)Female (N = 16,440)Male (N = 18,591)Female (N = 13,823)Male (N = 15,213)Female (N = 4916)Male (N = 4466)Screening Routine Screening (N = 14,249)7/157, 4.5%9/286, 3.2%86/2529, 3.4%159/3015, 5.3%119/2149, 5.5%191/2448, 7.8%86/1526, 8.3%142/1608, 8.8%22/274, 8.0%31/257, 12.1% Family history of CRC (N = 20,408)119/3302, 3.6%139/2442, 5.7%193/4026, 4.8%232/3201, 7.3%142/2441, 5.8%192/2021, 9.5%113/1479, 7.6%99/1004, 9.9%31/314, 9.9%17/178, 9.6% Family history of polyps (N = 2862)17/541, 3.1%21/331, 6.3%31/732, 4.2%32/499, 6.4%21/334, 6.3%18/180, 10.0%8/126, 6.4%8/73, 11.0%4/31, 12.9%3/15, 20.0% Positive FOBT (N = 16,491)67/1351, 5.0%91/1320, 6.9%119/1890, 6.3%318/2775, 11.5%196/1673, 11.7%432/2523, 17.1%208/1505, 13.8%408/2092, 19.5%115/698, 16.5%133/664, 20.0%Surveillance Surveillance of adenomatous polyps (N = 23,585)51/844, 6.0%69/974, 7.1%118/1948, 6.1%209/3013, 6.9%212/2687, 7.9%373/4684, 8.0%224/2861, 7.8%413/4424, 9.3%109/1006, 10.8%124/1144, 10.8% Surveillance of CRC (N = 4489)6/129, 4.7%8/131, 6.1%14/274, 5.1%31/359, 8.6%31/468, 6.6%64/702, 9.1%60/693, 8.7%89/917, 9.7%37/414, 8.9%38/402, 9.5%Evaluation of signs and symptoms Anemia or iron deficiency (N = 10,290)42/1146, 3.7%48/602, 8.0%72/1023, 7.0%95/1044, 9.1%94/1005, 9.4%157/1220, 12.9%132/1276, 10.3%231/1544, 15.0%102/756, 13.5%104/674, 15.4% Hematochezia (N = 31,691)191/4750, 4.0%317/4815, 6.6%260/4395, 5.9%529/5214, 10.2%255/2891, 8.8%438/3381, 13.0%233/2165, 10.8%335/2467, 13.6%116/850, 13.7%125/763, 16.4% Melena (N = 1499)2/112, 1.8%12/164, 7.3%3/110, 2.7%11/183, 6.0%6/112, 5.4%21/185, 11.4%16/152, 10.5%23/257, 9.0%9/95, 9.5%16/129, 12.4% Weight loss (N = 2793)8/264, 3.0%24/274, 8.8%19/239, 8.0%37/275, 13.5%23/266, 8.7%45/301, 15.0%28/379, 7.4%68/409, 16.6%26/181, 14.4%37/205, 18.1% IBS clusteraDefined as 1 or more of the following symptoms: diarrhea, constipation, abdominal pain/bloating, and change in bowel habits, but not weight loss or bleeding (anemia or iron deficiency, positive FOBT, hematochezia, or melena). (N = 26,855)85/4539, 1.9%70/2314, 3.0%157/4728, 3.3%136/2577, 5.3%167/3845, 4.3%178/2222, 8.0%152/3092, 4.9%148/1859, 8.0%76/1026, 7.4%61/653, 9.3%a Defined as 1 or more of the following symptoms: diarrhea, constipation, abdominal pain/bloating, and change in bowel habits, but not weight loss or bleeding (anemia or iron deficiency, positive FOBT, hematochezia, or melena). Open table in a new tab To account for the effect of demographic characteristics and indications for examination on the presence of mass or polyp >9 mm, we performed a multivariate logistic regression analysis. The adjusted RRs for demographic variables are presented in Figure 1. Increasing age was strongly associated with an increased risk of a mass or polyp >9 mm; this appeared to increase linearly with each decade through ≥80 years. Compared with adults younger than 50 years old, the risk was increased 1.55 times (adjusted RR, 1.55; 95% CI, 1.43–1.68) in patients age 50–59 years. By age >80 years, the RR was 3.25 (adjusted RR, 3.25; 95% CI, 2.93–3.62). Male gender was associated with increased risk in the entire cohort (adjusted RR, 1.57; 95% CI, 1.50–1.64). Overall, blacks had a 15% increased risk of mass or polyp >9 mm compared with white, non-Hispanics (adjusted RR, 1.15; 95% CI, 1.07–1.24), whereas Asian/Pacific Islanders had a lower risk (adjusted RR, .67; 95% CI, .56–.81). Hispanics, Native Americans, and multiracial patients had risk equivalent to non-Hispanic whites. As an expression of the absolute risk of demographic characteristics and examination indications, we used this multivariate logistic regression model to calculate the NNE to identify 1 patient with a mass or polyp >9 mm (Table 3). This analysis permits a comparison of patients receiving colonoscopy for screening, surveillance, and evaluation of symptoms, as well as by sex and age.Table 3NNE to Identify 1 Patient With Mass or Polyp >9 mmaThe final logistic regression model included age, gender, race and ethnicity, and the indications of asymptomatic screening, family history of CRC, family history of polyps, positive FOBT, surveillance of adenomatous polyps, surveillance of CRC, anemia or iron deficiency, hematochezia, melena, weight loss, and IBS cluster.FemaleMale<50 y50–59 y60–69 y70–79 y≥ 80 y<50 y50–59 y60–69 y70–79 y≥ 80 yScreening Asymptomatic screening42282018142818131210 Family history of CRC3422161411221511108 Positive FOBT Alone24159881610765 + Family history CRC251610981611766Surveillance Adenoma surveillance Alone1717141310151513129 + Family history CRC17181513111616141210 CRC surveillance3523171512231511108Evaluation of signs and symptoms Hematochezia2517121191711886 Anemia or iron deficiency Alone2718131291812987 + Hematochezia2214111081410775 Melena3926191713251712119 Weight loss2517121191611886 IBS clusterbDefined as 1 or more of the following symptoms; diarrhea, constipation, abdominal pain/bloating, and change in bowel habits, but not weight loss or bleeding (anemia or iron deficiency, positive FOBT, hematochezia, or melena).47312220153020151310a The final logistic regression model included age, gender, race and ethnicity, and the indications of asymptomatic screening, family history of CRC, family history of polyps, positive FOBT, surveillance of adenomatous polyps, surveillance of CRC, anemia or iron deficiency, hematochezia, melena, weight loss, and IBS cluster.b Defined as 1 or more of the following symptoms; diarrhea, constipation, abdominal pain/bloating, and change in bowel habits, but not weight loss or bleeding (anemia or iron deficiency, positive FOBT, hematochezia, or melena). Open table in a new tab Across all indications, the NNE decreases with increasing age. Men have a lower NNE than women within every age group. Among patients 9 mm. The NNE for asymptomatic women does not reach 20 until the 60- to 69-year-old decade. Comparing asymptomatic men and women, the NNE for men is roughly equivalent to that in women one decade older; this pattern holds across the age spectrum. FOBT is very strongly associated with an increased risk across every age group compared with asymptomatic patients with asymptomatic screening colonoscopy. Patients with a family history of CRC had only a slightly increased risk (lower NNE) of polyp(s) >9 mm compared with patients without a family history. The NNE for positive FOBT result with a family history of CRC is not different from that of a positive FOBT result alone; the risks are not additive. Patients undergoing surveillance after adenoma removal or cancer represent more than 20% of all colonoscopies in patients older than age 50 years. We found that the NNE for young patients ( 9 mm. These results confirm that there is a low risk of polyps >9 mm or masses in both asymptomatic men and women younger than 50 years of age. In patients younger than 50 years of age undergoing colonoscopy for a positive FOBT result, 6.9% of men and 5.0% of women had a mass or polyp >9 mm. These results endorse the performance of colonoscopy if the FOBT result is positive. Prior studies have shown that the risk of advanced neoplasia is increased 3–4-fold in patients with a positive FOBT result compared with those with a negative test result. 6Lieberman D.A. Weiss D.G. VACSP #380 Study GroupOne-time screening for colorectal cancer with combined fecal occult-blood testing and examination of the distal colon.N Engl J Med. 2001; 345;: 555-560Crossref PubMed Scopus (534) Google Scholar However, the value of performing FOBT in patients younger than 50 years is uncertain, and this practice is not recommended by the United States Preventive Services Task Force, the American Cancer Society, or the Multi-Society Task Force. 7Pignone M, Rich M, Teutsch SM, et al. Screening for colorectal cancer in adults at average risk: a summary of the evidence for the US Preventive Services Task Force. Ann Intern Med 137:132-141.Google Scholar, 8Smith R.A. Cokkinides V. Eyre H.J. American Cancer Society guidelines for early detection of cancer, 2003.CA Cancer J Clin. 2003; 53: 27-43Crossref PubMed Scopus (395) Google Scholar, 9Winawer S. Fletcher R. Rex D. et al.Colorectal cancer screening and surveillance: clinical guidelines and rationale: update based on new evidence.Gastroenterology. 2003; 124: 544-560Abstract Full Text PDF PubMed Scopus (1992) Google Scholar IBS symptoms without signs of bleeding or weight loss accounted for 25.2% of examinations in patients younger than 50 years old. Only 2.3% of these patients had a mass or polyps >9 mm. Women accounted for most of these examinations (66%), and the prevalence of a mass or polyps >9 mm in women with IBS symptoms was 1.9%. Further data are needed to determine whether these younger patients with IBS benefit from colonoscopy. Despite the low risk of large polyps or malignancy in individuals younger than 50 years, there might be benefits to ruling out malignancy and colonic inflammation. In patients with abdominal pain or changes in bowel habits, it is possible that a negative procedure could relieve anxiety and inform subsequent management decisions. If physicians believe that the diagnosis of IBS requires the confident exclusion of any GI pathology, colonoscopy could play an important role, even if negative. This benefit is unproven, and further study is needed to measure the benefits of colonoscopy in young patients with IBS symptoms. Hematochezia was an indication in 35.2% of colonoscopies performed in patients younger than 50 years. The severity of bleeding is unknown in our study. Several studies that have used colonoscopy to determine the risk of malignancy in patients with rectal bleeding are plagued by inconsistent methods and reporting. 10Helfand M. Marton K.I. Zimmer-Gembeck M.J. et al.History of visible rectal bleeding in a primary care population initial assessment and 10-year follow-up.JAMA. 1997; 277: 44-48Crossref PubMed Google Scholar, 11Mulcahy H.E. Patel R.S. Postic G. et al.Yield of colonoscopy in patients with nonacute rectal bleeding a multicenter database study of 1766 patients.Am J Gastroenterol. 2002; 97: 328-333Crossref PubMed Google Scholar, 12Talley N.J. Jones M. Self-reported rectal bleeding in a United States community prevalence, risk factors and health care seeking.Am J Gastroenterol. 1998; 93: 2179-2183Crossref PubMed Scopus (85) Google Scholar, 13Church J.M. Analysis of the colonoscopic findings in patients with rectal bleeding according to the pattern of their presenting symptoms.Dis Colon Rectum. 1991; 34: 391-395Crossref PubMed Scopus (54) Google Scholar, 14Goulston K.J. Cook I. Dent O.F. How important is rectal bleeding in the diagnosis of bowel cancer and polyps?.Lancet. 1986; 2: 261-265Abstract PubMed Scopus (117) Google Scholar, 15Fine K.D. Nelson A.C. Ellington T. et al.Comparison of the color of fecal blood with the anatomical location of gastrointestinal bleeding lesions potential misdiagnosis using only flexible sigmoidoscopy for bright red blood per rectum.Am J Gastroenterol. 1999; 94: 3202-3210Crossref PubMed Google Scholar, 16Segal W.N. Greenberg P.D. Rockey D.C. et al.The outpatient evaluation of hematochezia.Am J Gastroenterol. 1998; 93: 179-182Crossref PubMed Scopus (51) Google Scholar, 17Korkis A.M. McDougall C.J. Rectal bleeding in patients less than 50 years of age.Dig Dis Sci. 1995; 40: 1520-1523Crossref PubMed Scopus (45) Google Scholar There are no uniform definitions to describe the amount or frequency of bleeding. We found that the NNE for polyps >9 mm in patients with hematochezia who are younger than 50 years was 17 in men and 25 in women, considerably lower than the NNE in patients receiving asymptomatic screening. Further study is needed to determine whether colonoscopy is the optimal procedure to evaluate hematochezia in patients younger than 50 years without a family history of CRC, or if sigmoidoscopy would have been adequate. 17Korkis A.M. McDougall C.J. Rectal bleeding in patients less than 50 years of age.Dig Dis Sci. 1995; 40: 1520-1523Crossref PubMed Scopus (45) Google Scholar Above 50 years of age, the risk of masses or polyps >9 mm increases progressively with age in both men and women (Figure 1, Table 3). The NNE to identify 1 average-risk, asymptomatic patient with mass or polyp(s) >9 mm declines in men from 18 (age, 50–59 years) to 10 (age, >80 years); in women, the NNE declines from 28 (age, 50–59 years) to 14 (>80 years). These findings from diverse clinical practice settings are consistent with prior reports of screening colonoscopy in asymptomatic populations. 3Lieberman D.A. Weiss D.G. Bond J.H. et al.Use of colonoscopy to screen asymptomatic adults for colorectal cancer.N Engl J Med. 2000; 343: 162-168Crossref PubMed Scopus (1626) Google Scholar, 4Imperiale T.F. Wagner D.R. Lin C.Y. et al.Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings.N Engl J Med. 2000; 343: 169-174Crossref PubMed Scopus (926) Google Scholar, 18Schoenfeld P. Cash B. Dobhan R. et al.Colorectal neoplasia screening with colonoscopy in average-risk women at regional Naval medical centers the CONCERN trial.Gastrointest Endosc. 2002; 55: AB99Google Scholar The results highlight age-related and gender differences. There is a 10-year gap between risk in asymptomatic men and women on the basis of these data, which might have implications for future screening recommendations. These results confirm the value of performing screening colonoscopy in patients with a positive family history of CRC, who have a lower NNE across all age groups, compared with patients without a family history. In addition, there is a striking reduction in NNE (almost 50%) in patients who receive colonoscopy for a positive FOBT result compared with patients undergoing screening colonoscopy without FOBT. This confirms prior studies demonstrating the positive predictive value of FOBT and emphasizes the importance of performing colonoscopy in all patients with a positive test result. A recent survey found that many physicians either repeat FOBT or perform sigmoidoscopy alone after positive FOBT result. 19Nadel M.R. Shapiro J.A. Klabunde C.N. et al.A national survey of primary care physicians' methods for screening for fecal occult blood.Ann Intern Med. 2005; 142: 86-94Crossref PubMed Scopus (158) Google Scholar Patients receiving surveillance have a lower NNE than patients with screening colonoscopy, but this difference decreases with increasing age. In men aged 60 years and older, the NNE for surveillance of adenomas is similar to asymptomatic screening. Colon surveillance for prior adenomas accounts for more than 20% of all procedures performed in patients older than 50 years. In this analysis, we do not know the findings on the baseline colonoscopy or the interval for surveillance. The most important surveillance examination might be the first one after detection of neoplasia, to confirm the absence of a missed lesion on the baseline study. Further study is needed to determine the appropriate interval for adenoma surveillance based on findings of the baseline examination. Patients with anemia, hematochezia, or weight loss had lower NNE across all age groups compared with patients receiving screening colonoscopy, suggesting that colonoscopy is appropriate for these indications. Patients with melena or IBS cluster as the primary reason for colonoscopy did not have a lower NNE, compared with patients with screening colonoscopy. Therefore, these latter symptoms do not identify an individual more likely to have serious neoplasia than asymptomatic patients. There is considerable evidence that black non-Hispanics have a higher risk of death from CRC than whites and Hispanics. Prior studies have suggested that blacks represent a high-risk group and should be targeted in screening efforts. 20Jemal A. Murray T. Samuels A. et al.Cancer statistics 2003.CA Cancer J Clin. 2003; 53: 5-26Crossref PubMed Scopus (3380) Google Scholar, 21Theuer C.P. Wagner J.L. Taylor T.H. et al.Racial and ethnic colorectal cancer patterns affect the cost-effectiveness of colorectal cancer screening the United States.Gastroenterology. 2001; 120: 848-856Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar, 22Nelson R.L. Dollear T. Freels S. et al.The relation of age, race and gender to the subsite location of colorectal carcinoma.Cancer. 1997; 80: 193-197Crossref PubMed Scopus (140) Google Scholar, 23Cordice J.W. Johnson H. Anatomic distribution of colonic cancers in middle-class black Americans.J Natl Med Assoc. 1991; 83: 730-732PubMed Google Scholar, 24Chen V.W. Fenoglio-Preiser C.M. Wu X.C. et al.Aggressiveness of colon carcinoma in blacks and whites.Cancer Epidemiol Biomarkers Prev. 1997; 6: 1087-1093PubMed Google Scholar We found important differences in procedure utilization and outcomes on the basis of race/ethnicity. Black non-Hispanics undergoing colonoscopy for any reason have a higher risk of mass or polyp(s) >9 mm than whites (non-Hispanic) or Hispanics (Figure 1). There are many possible explanations for these results. The biology of colorectal neoplasia might differ in blacks, resulting in more rapid progression to advanced neoplasia at a younger age. It is also possible that blacks do not seek medical care when they develop GI symptoms, or that they have less access to medical care. This could lead to delays in diagnosis and discovery of colon lesions in later stages of disease. We found that blacks, compared with whites, underwent colonoscopy more commonly for anemia and rectal bleeding (36.3% vs 27.6%; P < .001). These symptoms are associated with a higher risk of colon neoplasia (Table 3). After adjusting for other variables, our results suggest that blacks represent a group at higher risk for advanced neoplasia and are under-represented among patients receiving colonoscopy in the CORI cohort, compared with the 2000 United States census. Further study is needed to determine factors that affect colonoscopy utilization among blacks. There are several important limitations of CORI data. Although the consortium was designed to mirror endoscopic practice in the United States, we cannot exclude the possibility of site selection bias. Practice sites that participate in CORI are comfortable using computers for endoscopic reporting and sharing data from their practice. Such practices might differ in important ways from those that will not share data or do not use electronic records to monitor quality in their practice. Some of the practices participating in CORI use the computerized report generator for procedures performed in an ambulatory surgical center, but not in the hospital, because of hospital policies regarding information technologies. Therefore, these data might not reflect the full spectrum of inpatient hospital procedures performed by the participants. Because most colonoscopy procedures are performed in the outpatient setting, the current data should provide a valid reflection of outpatient colonoscopy. The prevalence of pathology in this cohort might differ from rates in the general population, even though the study cohort is quite diverse. Some data might be incomplete or inaccurate. Indications for colonoscopy are provided by the physician by checking an item on a computer screen. Physicians might be parsimonious and enter some, but not all, reported symptoms or problems. The endoscopy report also lacks clinical perspective; for example, we do not know what prior diagnostic tests or treatments were performed before colonoscopy. With regard to colon adenoma surveillance, we do not know whether the baseline lesion was an advanced adenoma, or whether the previous examination was complete or compromised by poor bowel preparation. This information could strongly influence the decision to perform colonoscopy. Last, race and ethnicity data are not self-reported but are provided by the physician or nurse. Although CORI is a large database, cell sizes in some minority groups were limited. In this study, we used mass or polyp >9 mm as a surrogate end point for serious pathology and assumed that more than 90% of such lesions would be adenomatous. 3Lieberman D.A. Weiss D.G. Bond J.H. et al.Use of colonoscopy to screen asymptomatic adults for colorectal cancer.N Engl J Med. 2000; 343: 162-168Crossref PubMed Scopus (1626) Google Scholar, 4Imperiale T.F. Wagner D.R. Lin C.Y. et al.Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings.N Engl J Med. 2000; 343: 169-174Crossref PubMed Scopus (926) Google Scholar Lesions of this size are considered clinically important because of the relatively high risk of developing cancer. 25Muto T. Bussey H.J.R. Morson B.C. The evolution of cancer or the colon and rectum.Cancer. 1975; 36: 2251-2270Crossref PubMed Scopus (1911) Google Scholar, 26Atkin W.S. Morson B.C. Cuzick J. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas.N Engl J Med. 1992; 326: 658-662Crossref PubMed Scopus (968) Google Scholar, 27Stryker S.J. Wolff B.G. Culp C.E. et al.National history of untreated colonic polyps.Gastroenterology. 1987; 93: 1009-1013Abstract PubMed Google Scholar, 28O'Brien M.J. Winawer S.J. Zauber A.G. et al.The National Polyp Study patient and polyp characteristics associated with high-grade dysplasia in colorectal adenomas.Gastroenteorlogy. 1990; 98: 371-379PubMed Google Scholar Recent studies have suggested that visual estimates of polyp size less than or greater than 9 mm are accurate more than 90% of the time. 29Schoen R.E. Gerber L.D. Margulies C. The pathologic measurement of polyp size is preferable to the endsocopic estimate.Gastrointest Endosc. 1997; 46: 492-496Abstract Full Text Full Text PDF PubMed Scopus (163) Google Scholar, 30Gopalswamy N. Shenoy V.N. Choudhry U. et al.Is in vivo measurement of size of polyps during colonoscopy accurate?.Gastrointest Endosc. 1997; 46: 497-502Abstract Full Text Full Text PDF PubMed Scopus (150) Google Scholar Although this analysis has focused on one key end point, we recognize that other findings at colonoscopy might be clinically important. Despite these important limitations, the CORI repository is a valuable hypothesis-generating tool. Observational snapshots of clinical practice identify important research questions, which will require more precise data collection in prospective studies. In conclusion, this is the first practice-based analysis of utilization and outcomes of colonoscopy in the United States. The current data demonstrate that in diverse practice settings, rates of detecting mass or polyp >9 mm vary on the basis of age, gender, race, and procedure indication. There are several important findings. Patients younger than 50 years account for 19% of colonoscopies and generally have a low risk of polyps >9 mm. Evaluation of hematochezia and IBS symptoms represents the most common indications in this age group. Further research is needed to determine whether colonoscopy is the most appropriate test for these symptoms in patients younger than 50 years. In patients older than 50 years, surveillance after adenoma or cancer and evaluation of positive FOBT result, anemia, hematochezia, and weight loss are associated with a higher yield of polyps >9 mm compared with patients receiving screening colonoscopy. Finally, black Americans have a higher risk of polyps >9 mm and should be the targets of intensive screening effort before the development of symptoms. Understanding utilization and outcomes can lead to further study to measure optimal use of colonoscopy and provide a basis for shifting resources to the screening of asymptomatic individuals.