BACKGROUND. Screening is effective in reducing the incidence and mortality of

Transcription

1 2093 Patterns and Predictors of Colorectal Cancer Test Use in the Adult U.S. Population Laura C. Seeff, M.D. 1 Marion R. Nadel, Ph.D., M.P.H. 1 Carrie N. Klabunde, Ph.D. 2 Trevor Thompson, B.S. 1 Jean A. Shapiro, Ph.D. 1 Sally W. Vernon, Ph.D. 3 Ralph J. Coates, Ph.D. 1 1 Division of Cancer Prevention and Control, Centers for Disease Control and Prevention, Atlanta, Georgia. 2 Applied Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Bethesda, Maryland. 3 University of Texas-Houston School of Public Health, Center for Health Promotion and Prevention Research, Houston, Texas. BACKGROUND. Screening is effective in reducing the incidence and mortality of colorectal cancer. Rates of colorectal cancer test use continue to be low. METHODS. The authors analyzed data from the National Health Interview Survey concerning the use of the home-administered fecal occult blood test (FOBT) and sigmoidoscopy/colonoscopy/proctoscopy to estimate current rates of colorectal cancer test use and to identify factors associated with the use or nonuse of tests. RESULTS. In 2000, 17.1% of respondents reported undergoing a home FOBT within the past year, 33.9% reported undergoing an endoscopy within the previous 10 years, and 42.5% reported undergoing either test within the recommended time intervals. The use of colorectal cancer tests varied by gender, race, ethnicity, age, education, income, health care coverage, and having a usual source of care. Having seen a physician within the past year had the strongest association with test use. Lack of awareness and lack of physician recommendation were the most commonly reported barriers to undergoing such tests. CONCLUSIONS. Less than half of the U.S. population age 50 years underwent colorectal cancer tests within the recommended time intervals. Educational initiatives for patients and providers regarding the importance of colorectal cancer screening, efforts to reduce disparities in test use, and ensuring that all persons have access to routine primary care may help increase screening rates. Cancer 2004;100: American Cancer Society. KEYWORDS: colorectal cancer (CRC), mass screening, fetal occult blood test (FOBT), sigmoidoscopy, colonoscopy. Address for reprints: Laura C. Seeff, M.D., Centers for Disease Control and Prevention, DCPC, 4770 Buford Highway, NE, Mailstop K-55, Atlanta, GA ; Fax: (770) ; lvs3@cdc.gov Received December 31, 2003; revision received March 4, 2004; accepted March 17, In accordance with evidence that screening reduces colorectal cancer (CRC) incidence and mortality, 1 7 several sets of national guidelines now recommend regular CRC screening for average-risk persons age 50 years using 1 or more of the following options: annual home fecal occult blood test (FOBT), flexible sigmoidoscopy every 5 years, a combination of home FOBT and flexible sigmoidoscopy, colonoscopy every 10 years, and/or double-contrast barium enema every 5 years However, the adoption of these guidelines is occurring slowly and test use remains low. We analyzed data from the 2000 Cancer Control Supplement of the National Health Interview Survey (NHIS) regarding the use of the home-administered FOBT and sigmoidoscopy/colonoscopy/proctoscopy to estimate the most current rates of CRC test use, to evaluate factors previously described in association with CRC test use, and to describe first-time national estimates of barriers to CRC testing from the perspective of the general public American Cancer Society DOI /cncr Published online 20 April 2004 in Wiley InterScience (

2 2094 CANCER May 15, 2004 / Volume 100 / Number 10 MATERIALS AND METHODS The NHIS, conducted by the National Center for Health Statistics (NCHS), the Centers for Disease Control and Prevention (CDC), is an in-person survey that collects health behavior and sociodemographic information from the civilian, noninstitutionalized U.S. population. The current survey oversamples black and Hispanic populations. The 2000 NHIS included a Cancer Control Module. 11 Response Rates In 2000, the household response rate was 88.9% for the core NHIS survey and 72.1% for the Cancer Control Module. Survey responses were weighted to reflect nonresponse and the probability of selection. A small percentage (3.6%) of persons did not provide valid responses to at least 50% of the Cancer Control Module questions and were excluded from analysis. 11 The 2000 Cancer Control Module The 2000 NHIS Cancer Control Module included questions concerning diet and nutrition; physical activity; tobacco; screening for breast, cervical, skin, prostate, and colorectal cancer; genetic testing; and family history of cancer. Information regarding cigarette smoking, alcohol intake, and leisure time activity was collected in the core questionnaire. Questions regarding three types of endoscopic tests (proctoscopy, sigmoidoscopy, and colonoscopy) were asked separately for the first time in the 2000 NHIS. CRC Screening Questions In the 2000 NHIS, 11,734 men and women age 50 years responded to questions regarding the use of FOBT, and 11,816 responded to questions regarding the use of sigmoidoscopy, colonoscopy, or proctoscopy. Respondents were asked whether they had ever undergone an FOBT, the timing of their most recent test, and the reason the test was performed. A definition of FOBT was provided if requested. Response categories for reasons for receiving the examination included (verbatim): 1) part of a routine physical examination/screening test; 2) because of a specific problem; 3) follow-up test of an earlier test or screening examination; 4) family history; and 5) other. Respondents were asked to identify only the main reason for receiving a test and multiple reasons were not accepted. Categories 1 and 4 were combined into screening categories and Categories 2, 3, and 5 were combined into nonscreening categories. Respondents who had never had undergone a FOBT or had not undergone one recently (within the last year) were asked whether a physician or health care professional had recommended the test in the past year. Physician recommendation was not asked of those who reported that their doctor didn t order it or they don t have doctor. Respondents also were asked if they had ever undergone a sigmoidoscopy, colonoscopy, or proctoscopy, with definitions read if requested. Those who answered affirmatively were asked which of the 3 tests they had received most recently (within the last 10 years). The same sequence of questions used for FOBT with regard to timing of the most recent examination, the reason for undergoing the examination, and the reason for not undergoing the examination was asked concerning sigmoidoscopy/colonoscopy/proctoscopy use. For the majority of the analyses performed, we evaluated the use of CRC tests performed for any indication. Persons with a history of CRC were excluded from all analyses (n 179). We analyzed only home FOBT because only home tests have been demonstrated to be effective in clinical trials, 1,2,5 7 and their use is recommended in national guidelines We defined an FOBT within the past year as having been performed within the recommended time interval, which is consistent with national guidelines We chose the time interval of 10 years as a measure of appropriate endoscopic screening to capture the use of all endoscopic procedures. National guidelines have not been consistent concerning the recommended timing and type of screening endoscopy. Because the wording of the survey questions captured the type and timing of only the most recent of the three possible types of endoscopies, we grouped sigmoidoscopy, colonoscopy, and proctoscopy into the single category of endoscopy. Data were analyzed for endoscopies performed within the previous 5 years (data not shown), and the results were similar to those reported herein. Other Variable Definitions With regard to race, only white, black, and other were reported because of small sample sizes for other races. Other includes American Indian/Alaska Native, Asian, Native Hawaiian, and other Pacific Islander. Race was imputed by the NCHS for the 49 persons with missing race data or who reported multiple race with no main race. 11 Although CRC test use questions were asked of respondents age 40 years, we analyzed only data from respondents age 50 years because CRC screening guidelines target average-risk persons age 50 years We also grouped persons by age 65 years or age 65 years to assess the association between Medicare coverage and CRC test use.

3 CRC Test Use in Adults/Seeff et al Persons who reported the emergency department as their only usual source of care were included among those with no usual source of care. With regard to mammography and Papanicolau (Pap) testing, appropriate intervals were defined according to U.S. Preventive Services Task Force guidelines, including mammography within 2 years after age 40 years and Pap smear testing within 3 years after age 18 years. 12,13 The following formula was used to calculate metabolic equivalents (METS) from the number of minutes per week of moderate and vigorous exercise: 4.5 moderate minutes per week 7.0 vigorous minutes per week. 14 Physical activity was defined as no activity (METS 0), some activity (METS 1 675), or meets/exceeds recommendations (METS 675) according to the Surgeon General s recommendation of 30 minutes of moderate activity per day at least 5 days a week. 15 Consistent with national guidelines, 16 fruit and vegetable intake was dichotomized as fewer than five or five or more servings per day. An algorithm developed at the National Cancer Institute 17 was used to categorize the fruit and vegetable information into categories consistent with these amounts. Statistical Analysis Screening rates with 95% confidence intervals (95% CIs) were calculated using SAS (version 8.2) 18 and SUDAAN (version 8.0) 19 software to account for the complex survey design and to allow for weighting. Screening rates were age-standardized to the 2000 U.S. standard million population using the direct method. 20 Respondents who refused or did not know the answer to a question were excluded from analysis of the specific question. The total number of respondent refusals or unknowns (for respondents age 50 years with no history of CRC) was 7.4%, 8.0%, 6.7%, and 7.1%, respectively, for questions regarding FOBT ever, FOBT recently, endoscopy ever, and endoscopy recently. Multivariate logistic regression was used to determine the independent variables associated with each of the three endpoints: FOBT within the previous year, endoscopy (sigmoidoscopy/colonoscopy/proctoscopy) within the past 10 years, and either FOBT within the previous year and/or endoscopy within the past 10 years. Additional models were designed to determine whether either mammography or Pap smear tests were associated with CRC test use among women. Two additional models compared screening test versus no test and nonscreening test versus no test. Variables were included in the multivariate models if they have been previously associated with CRC or CRC screening, if they appeared to be associated with test use based on descriptive tables, or if they varied according to test indication. All modeling results are presented as odds ratios (ORs) with 95% CIs. RESULTS In 2000, the overall age-adjusted percentages of respondents who reported ever having undergone CRC tests for any reason were 36.7% (95% CI, %) for home FOBT, 38.1% (95% CI, %) for endoscopy, and 54.2% (95% CI, %) for either 1 or both tests. With regard to tests used for screening versus nonscreening purposes, 88.3% (95%CI, %) of FOBT tests performed within the past year were performed for screening and 11.7% (95% CI, %) were performed for nonscreening purposes. With regard to endoscopy performed within the past 10 years, 60.8% (95%CI, %) were performed for screening and 39.2% (95%CI, %) were performed for nonscreening purposes. For CRC tests performed for all reasons combined (screening and nonscreening), 17.1% of respondents reported undergoing a home FOBT within the past year, 33.9% reported undergoing an endoscopy within the previous 10 years, and 42.5% reported undergoing 1 or the other test within the recommended time intervals (Table 1). With regard to FOBT, test rates increased with increasing age until ages years and then decreased. Persons age 65 years were more likely to report having undergone an FOBT compared with persons ages years. White, non-hispanic, and married persons were more likely to report having undergone a CRC test than black, Hispanic, or unmarried persons. Those who had private health insurance, Medicare or Medi-GAP, or a combination of private insurance and Medicare had higher rates of FOBT use than those with other public insurance or no insurance. Having a usual source of care and the frequency with which the respondent reported seeing a physician were both associated with higher reported test rates. Patterns of associations for endoscopy were similar to those for FOBT use, with the exception that being male was associated with higher rates of endoscopy. Those respondents with a personal or family history of cancers other than CRC or a family history of CRC were more likely to report having undergone endoscopy than those without such a history. Women who underwent a mammogram or Pap smear test within recommended intervals were more likely to report having undergone CRC tests compared with those who did not. Persons who exercised regularly

4 2096 CANCER May 15, 2004 / Volume 100 / Number 10 TABLE 1 Percentage of Respondents Age > 50 Years Who Reported Undergoing CRC cancer Tests within Recommended Time Intervals by Sociodemographic, Healthcare Access, and Health Behavior Characteristics, NHIS, 2000 a FOBT within previous year Endoscopy b within previous 10 years Either test within recommended time interval No. c % (95% CI) No. % (95% CI) No. % (95% CI) Total 11, ( ) 11, ( ) 11, ( ) Gender Male ( ) ( ) ( ) Female ( ) ( ) ( ) Race d White ( ) ( ) ( ) Black ( ) ( ) ( ) Other ( ) ( ) ( ) Hispanic or Latino No 10, ( ) 10, ( ) 10, ( ) Yes ( ) ( ) ( ) Age (group) (yrs) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) Age by Medicare enrollment status (yrs) ( ) ( ) ( ) ( ) ( ) ( ) Education 2 yrs ( ) ( ) ( ) High school graduate ( ) ( ) ( ) Some college ( ) ( ) ( ) College graduate ( ) ( ) ( ) Annual household income $20, ( ) ( ) ( ) $20,000 34, ( ) ( ) ( ) $35,000 54, ( ) ( ) ( ) $55,000 74, ( ) ( ) ( ) $75, ( ) ( ) ( ) Marital status Married e ( ) ( ) ( ) Unmarried f ( ) ( ) ( ) Health care coverage Private only g ( ) ( ) ( ) Medicare/Medi-GAP ( ) ( ) ( ) Private Medicare/Medi-GAP ( ) ( ) ( ) Other/multiple carriers h ( ) ( ) ( ) None ( ) ( ) ( ) Usual source of care No ( ) ( ) ( ) Yes 10, ( ) 10, ( ) 10, ( ) No. of physician visits in last year None ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) General health status Excellent/good ( ) ( ) ( ) Fair/poor ( ) ( ) ( ) BMI (kg/m 2 ) Normal ( 25) ( ) ( ) ( ) Overweight (25 29) ( ) ( ) ( ) Obese ( 30) ( ) ( ) ( ) (continued)

5 CRC Test Use in Adults/Seeff et al TABLE 1 (continued) FOBT within previous year Endoscopy b within previous 10 years Either test within recommended time interval No. c % (95% CI) No. % (95% CI) No. % (95% CI) Personal history of cancer d No 10, ( ) 10, ( ) 10, ( ) Yes ( ) ( ) ( ) Family history of CRC No ( ) ( ) ( ) Yes ( ) ( ) ( ) Family history of Non CRC No ( ) ( ) ( ) Yes ( ) ( ) ( ) Mammogram within past 2 yrs No ( ) ( ) ( ) Yes ( ) ( ) ( ) Pap smear test within past 3 yrs No ( ) ( ) ( ) Yes ( ) ( ) ( ) Physical activity None ( ) ( ) ( ) Some ( ) ( ) ( ) Meet/exceed Recommendations ( ) ( ) ( ) Fruit/vegetables 5 servings/day ( ) ( ) ( ) 5 servings/day ( ) ( ) ( ) Cigarette smoking Never ( ) ( ) ( ) Former ( ) ( ) ( ) Current ( ) ( ) ( ) Alcohol use None ( ) ( ) ( ) 1 14 drinks/week ( ) ( ) ( ) 14 drinks/week ( ) ( ) ( ) CRC: colorectal cancer; NHIS: National Health Interview Survey; FOBT: fecal occult blood test; 95% CI: 95% confidence interval; BMI: body mass index; Pap: Papanicolaou. a 2000 estimates age-adjusted to 2000 U.S. Bureau of the Census Decennial Census. b Sigmoidoscopy, colonoscopy, or proctoscopy. c Sample size for each question. d Persons of multiple race who selected a main race are coded according to the main race; Other racial category includes American Indian/Alaska Native, Asian, Native Hawaiian and other Pacific Islander. e Includes married and living with partner. f Includes divorced/separated, widowed, never married. g Includes private health insurance from employer, private health insurance purchased directly, and Civilian Health and Medical Programs of Uniformed Services (CHAMPUS). h Includes Medicaid; Indian Health Service, state-sponsored health plan, other government program, military health care/veterans Administration, any combination of private or Medicare. i Excludes persons with a history of colorectal cancer. reported higher rates of test use for all CRC tests evaluated. Former cigarette smokers reported higher test rates compared with never-smokers or current smokers. Adjusted patterns of association were similar to unadjusted rates, with some exceptions (Table 2). Women were more likely to have undergone an FOBT compared with men. Race was no longer found to be a predictor of FOBT use. The strongest association was noted with the number of times the respondent had seen a physician. Compared with respondents who reported no physician visits in the past year, the odds of having undergone a FOBT were nearly sevenfold higher for respondents who reported greater than six physician visits in the preceding year. Having a usual source of care and having undergone a recent mammogram or Pap test also were found to be strongly associated with FOBT use (OR 2.05, 3.06, and 2.24, respectively). There was no significant difference noted with regard to endoscopy use between blacks and whites, but persons of other race were less likely to have

6 2098 CANCER May 15, 2004 / Volume 100 / Number 10 TABLE 2 Association between Selected Characteristics and FOBT and Sigmoidoscopy/Colonoscopy/Proctoscopy, NHIS, 2000 a FOBT within past year (n 10,370) Endoscopy within past 10 years (n 10,438) Either test within recommended time interval (n 10,362) Characteristic OR (95% CI) OR (95% CI) OR (95% CI) Gender b Male 1.0 (Reference) 1.0 (Reference) 1.0 (Reference) Female 1.15 ( ) 0.77 ( ) 0.89 ( ) Race b White Black 1.06 ( ) 1.09 ( ) 1.08 ( ) Other 0.80 ( ) 0.62 ( ) 0.67 ( ) Hispanic or Latino b Yes 0.69 ( ) 1.03 ( ) 0.92 ( ) Age (group) (yrs) b ( ) 1.56 ( ) 1.45 ( ) ( ) 1.71 ( ) 1.69 ( ) ( ) 1.42 ( ) 1.25 ( ) Education b 12 yrs High school graduate 1.23 ( ) 1.29 ( ) 1.27 ( ) Some college 1.46 ( ) 1.55 ( ) 1.53 ( ) College graduate 1.64 ( ) 1.93 ( ) 1.83 ( ) Marital status b Married Unmarried 0.87 ( ) 0.86 ( ) 0.80 ( ) Health care coverage b Private only 1.39 ( ) 1.71 ( ) 1.66 ( ) Medicare/Medi-GAP 1.52 ( ) 1.72 ( ) 1.82 ( ) Private Medicare/Medi-GAP 1.43 ( ) 1.98 ( ) 1.93 ( ) Other 1.34 ( ) 1.68 ( ) 1.69 ( ) None Usual source of care b Yes 2.05 ( ) 1.52 ( ) 1.65 ( ) No. of physician visits in last year b None ( ) 1.92 ( ) 2.40 ( ) ( ) 2.56 ( ) 3.30 ( ) ( ) 3.56 ( ) 4.68 ( ) General health status b Excellent/good Fair/poor 1.04 ( ) 1.09 ( ) 1.07 ( ) BMI (kg/m 2 ) b Normal ( 25) Overweight (25 29) 0.99 ( ) 1.10 ( ) 1.07 ( ) Obese ( 30) 1.07 ( ) 1.09 ( ) 1.11 ( ) Personal history of cancer b Yes 1.09 ( ) 1.52 ( ) 1.42 ( ) Family history of CRC b Yes 1.10 ( ) 2.32 ( ) 2.04 ( ) Family History of Non-CRC b Yes 1.01 ( ) 1.27 ( ) 1.24 ( ) (continued)

7 CRC Test Use in Adults/Seeff et al TABLE 2 (continued) FOBT within past year (n 10,370) Endoscopy within past 10 years (n 10,438) Either test within recommended time interval (n 10,362) Characteristic OR (95% CI) OR (95% CI) OR (95% CI) Mammogram within 2 yrs c Yes 3.06 ( ) 2.55 ( ) 2.96 ( ) Pap test within 3 yrs d Yes 2.24 ( ) 2.06 ( ) 2.41 ( ) Physical activity b None Moderate 1.25 ( ) 1.25 ( ) 1.25 ( ) Meet/exceed Recommendations 1.63 ( ) 1.55 ( ) 1.64 ( ) Recommendations Cigarette smoking b Never Former 1.16 ( ) 1.15 ( ) 1.17 ( ) Current 0.93 ( ) 0.82 ( ) 0.82 ( ) Alcohol use b None drinks/week 1.13 ( ) 1.09 ( ) 1.14 ( ) 14 drinks/week 1.21 ( ) 0.98 ( ) 0.97 ( ) FOBT: fecal occult blood test; NHIS: National Health Interview Survey; OR: odds ratio; 95% CI: 95% confidence interval; BMI: body mass index; CRC: colorectal cancer; Pap: Papanicolaou. a 2000 estimates age-adjusted to 2000 US Bureau of the Census Decennial Census. b Odds ratios adjusted for all other variables except mammography and Papanicolaou test use. c Odds ratio for mammography use adjusted for all variables except gender and Papanicolaou test use; n 5904 (fecal occult blood test); n 5964 (endoscopy); and n 5906 (either test). d Odds ratio for Papanicolaou test use adjusted for hysterectomy status as well as adjusted for all other variables except gender and mammography use; n 5858 (fecal occult blood test); n 5911 (endoscopy); and n 5859 (either test). undergone an endoscopy or either test. The association between ethnicity and endoscopy use or the use of either test disappeared when controlling for other factors. Again, the strongest associations were observed with the number of times the respondent had seen a physician; compared with respondents who reported no physician visits in the past year, the odds of having undergone an endoscopy was nearly fourfold higher for respondents who reported greater than six physician visits and nearly fivefold higher for those reporting undergoing either test. For persons with a family history of CRC, the odds of having undergone an endoscopy within the past 10 years were almost 2.5 times higher than for those without such a history. The most common reason cited for not having undergone FOBT or endoscopy was lack of awareness of the need for screening ( never thought about it ), followed by lack of recommendation by the physician ( doctor didn t order the test) (Table 3). Other commonly cited reasons also suggested a lack of awareness of the need for the test ( didn t need it, haven t had any problems, put it off ). Few respondents cited cost, pain/unpleasantness/embarrassment, or having undergone another type of colorectal examination as their reasons for not undergoing CRC tests. Cost and pain/unpleasantness/embarrassment were more likely to be associated with not having undergone endoscopy than with not having undergone FOBT, whereas having undergone another type of colorectal examination was more commonly a barrier to FOBT than endoscopy. Between 4 7% of respondents who did not undergo CRC testing had been advised by a physician to do so. Respondents ages years were more likely to report that a physician recommended the test, more likely to report that they had not had any problems, and more likely to put the test off compared with those age 65 years. In the current study, we only present differences in test use by indication (screening vs. nonscreening) for endoscopy because FOBT is used primarily for CRC screening and not as a diagnostic test. Although women were less likely to undergo endoscopy for screening than men, they were as likely to have undergone the test for nonscreening (Table 4). Black persons were slightly more likely to have undergone screening endoscopies but less likely to receive non-

8 2100 CANCER May 15, 2004 / Volume 100 / Number 10 TABLE 3 Reasons for Never Undergoing Colorectal Examinations or for Undergoing Tests beyond the Recommended Time Interval (NHIS) by Ages Years and > 65 Years, 2000 a FOBT Endoscopy b Age yrs Age > 65 yrs Aged yrs Aged > 65 yrs No. c % (95% CI) No. % (95% CI) No. c % (95% CI) No. % (95% CI) Reason for not undergoing CRC tests Never thought about it ( ) ( ) ( ) ( ) Doctor didn t order it ( ) ( ) ( ) ( ) Didn t need it ( ) ( ) ( ) ( ) Have not had any problems ( ) ( ) ( ) ( ) Put it off ( ) ( ) ( ) ( ) Too expensive/no insurance ( ) ( ) ( ) ( ) Too painful, unpleasant, embarrassing ( ) ( ) ( ) ( ) Had another type of colorectal examination ( ) ( ) ( ) ( ) Don t have a doctor ( ) ( ) ( ) ( ) Other ( ) ( ) ( ) ( ) Doctor recommendation in past year d Yes ( ) ( ) ( ) ( ) No ( ) ( ) ( ) ( ) NHIS: National Health Interview Survey; FOBT: fecal occult blood test; 95% CI: 95% confidence interval; CRC: colorectal cancer. a Sigmoidoscopy, colonoscopy, or proctoscopy. b Estimates not adjusted for age. c Sample size for each question. d Only asked of respondents who did not undergo the tests. Those who responded don t have doctor and those with no physician visits within the past 12 months were excluded. screening endoscopies compared with white persons. The association between educational attainment and CRC test use was more pronounced for screening than nonscreening. With increasing frequency of physician visits, the odds of having undergone a nonscreening endoscopy increased more dramatically than the odds of having undergone a screening endoscopy. Those persons in fair/poor health were more likely to have undergone a test for nonscreening purposes whereas those in excellent/good health were more likely to have undergone a test for screening purposes. The association between a family history of CRC and having undergone endoscopy was stronger with screening tests than with nonscreening tests. The same pattern was noted for women who reported a recent mammogram or Pap smear test. The proportion of persons who used CRC tests was lower than the proportion of the appropriate population who used mammography and Pap smear tests (data not shown). DISCUSSION The results of the current study demonstrate that less than half of the U.S. population age 50 years reported undergoing home FOBT and endoscopy (sigmoidoscopy, colonoscopy, or proctoscopy) within recommended time intervals. Consistent with previously published literature regarding CRC screening, we have shown that test use varies by gender, race, ethnicity, age, education, income, health care coverage, and having a usual source of care With the exception of age, the screening behavior patterns observed in the current study data are consistent with the literature with regard to mammography and Pap smear test screening behaviors For CRC, test use continues to increase through the eighth decade of life, whereas mammography test use is highest among women in their 60s and then decreases with increasing age. 28 We found that practicing other healthy behaviors (including having other cancer screening tests, exercising, and eating larger amounts of fruits and vegetables) was associated with higher rates of test use, which again is consistent with previously published data regarding CRC, breast cancer, and cervical cancer screening test behaviors. 21,27 29 The variable found to be most strongly associated with CRC test use was having seen a physician at least once in the preceding year, a finding that is consistent with earlier studies. 30,31 We reported an increased association with increasing numbers of physician visits during the preceding year. Nonscreening tests were more strongly associated with an increasing frequency

9 CRC Test Use in Adults/Seeff et al TABLE 4 Association between Selected Characteristics and Endoscopy by Test Indication, NHIS, 2000 a Characteristic Screening b (n 9575) Endoscopy Nonscreening b (n 8951) Gender c Male (Reference) (Reference) Female 0.58 ( ) 0.94 ( ) Race c White (Reference) (Reference) Black 1.22 ( ) 0.65 ( ) Other 0.62 ( ) 0.52 ( ) Hispanic or Latino c Yes 1.13 ( ) 0.89 ( ) Age (group) (yrs) b (Reference) (Reference) ( ) 1.23 ( ) ( ) 1.29 ( ) ( ) 0.80 ( ) Education c 12 yrs (Reference) (Reference) High school graduate 1.37 ( ) 1.32 ( ) Some college 1.87 ( ) 1.49 ( ) College graduate 2.57 ( ) 1.80 ( ) Health care coverage c Private only 2.13 ( ) 1.63 ( ) Medicare/Medi-GAP 1.90 ( ) 1.67 ( ) Private Medicare/Medi-GAP 2.08 ( ) 2.37 ( ) Other 1.52 ( ) 1.84 ( ) None (Reference) (Reference) Usual source of care c Yes 1.77 ( ) 1.50 ( ) No. of physician visits in last year c None (Reference) (Reference) ( ) 1.52 ( ) ( ) 3.19 ( ) ( ) 5.20 ( ) General health status c Excellent/good (Reference) (Reference) Fair/poor 0.81 ( ) 1.29 ( ) Personal history of cancer c Yes 1.45 ( ) 1.71 ( ) Family history of CRC c Yes 2.78 ( ) 1.69 ( ) Mammogram within 2 yrs d Yes 3.65 ( ) 2.11 ( ) Pap smear test within 3 yrs e Yes 2.75 ( ) 1.56 ( ) NHIS: National Health Interview Survey; CRC: colorectal cancer; Pap: Papanicolaou. a 2000 estimates age-adjusted to 2000 U.S. Bureau of the Census Decennial Census. b Versus no test. c Odds ratios adjusted for all other variables except mammography and Papanicolaou smear use. d Odds ratio for mammography use adjusted for all variables except gender and Papanicolaou smear use; n 5573 (screening) and n 5345 (nonscreening). e Odds ratio for Papanicolaou test use adjusted for hysterectomy status as well as adjusted for all other variables except gender and mammography; n 5516 (screening) and n 5287 (non-screening). of physician visits compared with screening tests, and it may be that the repeated visits were associated with diagnostic testing and were unrelated to CRC screening. However, regardless of test indication, any versus no physician visits increased the likelihood of test use. Three other variables (doctor recommendation, doctor didn t order it, and usual source of care) also illustrated an association between contact with a physician or health care system and CRC test use. We cannot determine causality between physician visits and screening, since the physician visit needs to occur for the CRC test to be ordered or performed. We reported some associations that to our knowledge were not previously described in national data, including the association between a personal or family history of cancer, a family history of CRC, and marital status. Those with a family history of CRC reported having undergone endoscopy at twice the rate of those persons without such a history. Overall, more tests were reported for screening than for nonscreening purposes, a finding that is consistent with previous reports. 25 The associations between family history of CRC, mammogram or Pap smear use, and educational attainment were found to be stronger for screening test use than nonscreening test use. Black persons underwent screening endoscopy at rates equal to white persons, but appeared to undergo nonscreening endoscopy at lower rates. Women were as likely to report undergoing nonscreening endoscopy but less likely to report undergoing screening endoscopy. These findings suggest an important differential in medical care related to gastroenterologic disease across race and gender. The 2000 NHIS dataset does not allow us to explore whether this differential is explained by patient choice or by health care system factors. However, observations regarding test use according to indication should be treated with caution because we do not know with certainty how familiar the survey respondents were with the reason they received the tests. To our knowledge, this is the first report from nationally representative data to examine reasons for not undergoing CRC tests. A lack of awareness by the respondent of the need for the test and a lack of recommendation by a physician for the test to be performed were found to be the most commonly reported barriers to undergoing the test. Lack of a physician recommendation clearly was an important barrier; among persons who reported undergoing no CRC testing or none recently, only 5% reported that a physician had recommended CRC testing. These barriers will be addressed in more detail in comparison with barriers from the physician perspective in a separate report (unpublished data).

10 2102 CANCER May 15, 2004 / Volume 100 / Number 10 Our findings regarding the reported use of CRC tests performed for screening and diagnostic purposes combined are consistent with the rates of use of FOBT and sigmoidoscopy or colonoscopy (lower endoscopy) reported in the 2001 Behavioral Risk Factor Surveillance System (BRFSS), a state-based health behavior telephone survey of the U.S. population. 32 BRFSS does not ask respondents to distinguish between tests used for screening and diagnostic purposes. In 2001, an estimated 23.5% of BRFSS respondents reported undergoing FOBT within the past 12 months, 43.4% reported undergoing a lower endoscopy within the previous 10 years, and 53.1% reported undergoing 1 or both tests within the recommended time intervals. The slightly higher reported rates of test use in the BRFSS compared with the NHIS most likely reflect differences in sampling, response rates, and survey administration (in-person vs. telephone). The current analysis has some limitations. NHIS survey data are based on self-report and are not validated through medical record review. Furthermore, because of the design of the survey instrument and probable confusion by respondents between the various endoscopic procedures, leading to possible misclassification, we grouped the use of any endoscopy together. We likely overestimated compliance with guidelines for screening by sigmoidoscopy (recommended every 5 years) by choosing 10 years as the appropriate screening interval for endoscopy. However, choosing 5 years as a measure of appropriate screening would have underestimated compliance with screening guidelines for colonoscopy, and more people reported undergoing colonoscopy than sigmoidoscopy. The primary challenge for the public health and medical communities regarding the continued low rates of CRC test use is to continue to educate both the public and physicians regarding the importance of screening for this disease. We could not evaluate the influence that physician time pressure and/or the use of reminder systems may have had on the use of CRC tests. At the time of the 2000 NHIS, there was no Health Plan Employer Data Information Set (HEDIS) measure for CRC screening; HEDIS measures and reminder systems for mammography and Pap smear testing have been reported to be effective in helping to increase test utilization. 33,34 A HEDIS measure for CRC has just been accepted, and will likely have an impact on these low rates of test use. Identifying physician incentives to recommend screening, addressing health care coverage of screening tests, and making changes to health systems such as monitoring CRC screening use and using provider and client reminder systems, also should aid in the effort to increase CRC screening. Efforts to reduce disparities in test use across gender, race, ethnicity, income, education, and health care coverage also must be heightened. A previous report using 2000 NHIS data also emphasized disparities according to place of birth. 35 Because contact with a physician and having a usual source of care increases the likelihood that a person will obtain a CRC test, ensuring that all persons have access to routine primary medical care would create regular opportunities to recommend screening. REFERENCES 1. Mandel JS, Church TR, Bond JH, et al. The effect of fecal occult-blood screening on the incidence of colorectal cancer. N Engl J Med. 2000;343: 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: Selby JV, Friedman GD, Quesenberry CP Jr., Weiss NS. A case-control study of screening sigmoidoscopy and mortality from colorectal cancer. N Engl J Med. 1992;326: Newcomb PA, Norfleet RG, Storer BE, Surawicz TS, Marcus PM. Screening sigmoidoscopy and colorectal cancer mortality. J Natl Cancer Inst. 1992;84: Hardcastle JD, Chamberlain JO, Robinson MH, et al. Randomised controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet. 1996;348: Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sondergaard O. Randomised study of screening for colorectal cancer with faecal-occult-blood test. Lancet. 1996;348: Mandel JS, Church TR, Ederer F, Bond JH. Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood. J Natl Cancer Inst. 1999;91: Agency for Healthcare Research and Quality. Colorectal cancer screening. In: Guide to clinical preventative services, 3 rd edition: periodic updates. July Available from URL: [accessed 9 April 2004]. 9. Smith RA, von Eschenbach AC, Wender R, et al. American Cancer Society guidelines for the early detection of cancer: update of early detection guidelines for prostate, colorectal, and endometrial cancers. CA Cancer J Clin. 2001;51: Winawer SJ, Fletcher RH, Miller L, et al. Colorectal cancer screening: clinical guidelines and rationale. Gastroenterology. 1997;112: National Health Interview Survey survey description. Washington, DC: National Center for Health Statistics, March Available from URL: ftp://ftp.cdc.gov/pub/healthstatistics/nchs/dataset_documentation/nhis/2000/srvydesc. pdf [accessed 1 Nov 2003]. 12. Agency for Healthcare Research and Quality. Breast cancer screening. In: Guide to clinical preventative services, 3 rd edition: periodic updates. February Available from URL: [accessed 9 April 2004]. 13. Agency for Healthcare Research and Quality. Cervical cancer screening. In: Guide to clinical preventative services, 3 rd edition: periodic updates. January Available from URL: [accessed 9 April 2004].

11 CRC Test Use in Adults/Seeff et al Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. 2000;32:S465 S U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta: Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Stables GJ, Subar AF, Patterson BH, et al. Changes in vegetable and fruit consumption and awareness among US adults: results of the 1991 and A-Day for Better Health Program surveys. J Am Diet Assoc.2002;102: National Cancer Institute. Fruit and vegetable screeners. Available from URL: fruitveg/scoring/allday.html [accessed 9 April 2004]. 18. SAS statistical analysis software. Cary, NC: SAS Institute Inc., SUDAAN software for the statistical analysis of correlated data. Research Triangle Park, NC: Research Triangle Institute, Breslow NE, Day NE. Statistical methods in cancer research. Volume 2: the design and analysis of cohort studies. IARC Scientific Pub. No. 82. Lyon, France: International Agency for Research on Cancer, 1987: Shapiro JA, Seeff LC, Nadel MR. Colorectal cancer screening and associated health behaviors. Am J Prev Med. 2001;21: Breen N, Wagener DK, Brown ML, Davis WW, Ballard-Barbash R. Progress in cancer screening over a decade: results of cancer screening from the 1987, 1992, and 1998 National Health Interview Surveys. J Natl Cancer Inst. 2001;93: Seeff LC, Shapiro JA, Nadel MR. Are we doing enough to screen for colorectal cancer? Findings from the 1999 Behavorial Risk Factor Surveillance System. J Fam Pract. 2002;51: Nadel MR, Blackman D, Shapiro JA, Seeff LC. Are people being screened for colorectal cancer as recommended? Results from the National Health Interview Survey. Prev Med. 2002;35: Peterson SK, Vernon SW. A review of patient and physician adherence to colorectal cancer screening guidelines. Semin Colon Rectal Surg. 2000;11: Potosky AL, Breen N, Graubard BI, Parsons PE. The association between health care coverage and the use of cancer screening tests. Results from the 1992 National Health Interview Survey. Med Care.1998;36: Coates RJ, Uhler RJ, Brogan DJ, et al. Patterns and predictors of the breast cancer detection methods in women under 45 years of age (United States). Cancer Causes Control. 2001;12: Vernon SW, Laville EA, Jackson GL. Participation in breast screening programs: a review. Soc Sci Med. 1990;30: Norman SA, Talbott EO, Kuller LH, et al. Demographic, psychosocial, and medical correlates of Pap testing: a literature review. Am J Prev Med. 1991;7: Bernstein AB, Thompson GB, Harlan LC. Differences in rates of cancer screening by usual source of medical care. Data from the 1987 National Health Interview Survey. Med Care. 1991;29: Brown ML, Potosky AL, Thompson GB, Kessler LG. The knowledge and use of screening tests for colorectal and prostate cancer: data from the 1987 National Health Interview Survey. Prev Med. 1990;19: Centers for Disease Control and Prevention. Colorectal cancer test use among persons aged 50 years United States, MMWR Morb Mortal Wkly Rep. 2003;52: Lied TR, Sheingold S. HEDIS performance trends in Medicare managed care. Health Care Financ Rev. 2001;23: Kupets R, Covens A. Strategies for the implementation of cervical and breast cancer screening of women by primary care physicians. Gynecol Oncol. 2001;83: Swan J, Breen N, Coates RJ, Ballard-Barbash R. Progress in cancer screening practices in the U.S.: results from the 2000 National Health Interview Survey. Cancer. 2003;97: