SEARCH

SEARCH BY CITATION

Keywords:

  • colorectal cancer;
  • screening;
  • fecal occult blood test;
  • colonoscopy;
  • endoscopy;
  • socioeconomic factors;
  • race;
  • ethnicity

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

BACKGROUND:

National surveys have reported declines in rates of home-based fecal occult blood test (FOBT) screening for colorectal cancer (CRC) in the last decade. However, socioeconomic status (SES) and racial/ethnic differences in FOBT trends and their changes relative to endoscopic CRC screening have not been evaluated.

METHODS:

Data on adults ages 50 to 64 years from the 2000, 2005, and 2008 National Health Interview Surveys were used. Weighted analyses and multivariate logistic regression were used to study trends in the use of FOBT and endoscopic CRC screening during this period.

RESULTS:

Between 2000 and 2008, significant declines in FOBT prevalence occurred in higher SES groups, but not in lower SES groups (uninsured and publicly insured, those without a usual source of care, lower educated, lower income, and immigrants to the United States) or Hispanics. Endoscopic CRC screening during the period studied consistently increased in all higher SES subgroups. In contrast, few lower SES subgroups (publicly insured, lower educated, near poor individuals, long-term immigrants) and Hispanics experienced increases in CRC endoscopic screening, and these increases were smaller than those observed in higher SES subgroups.

CONCLUSIONS:

Socially and economically disadvantaged groups experienced little or no change in FOBT prevalence, and few of these groups experienced contemporaneous increases in CRC endoscopic screening. These trends suggest the continued availability and acceptance of FOBT in these groups. If national CRC screening goals are to be achieved in populations with lower access to colonoscopy, then annual high-sensitivity FOBT should be promoted as an immediately accessible and viable alternative. Cancer 2012. © 2012 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Colorectal cancer (CRC) is the third leading cause of cancer mortality in US men and women, with nearly 50,000 deaths estimated in 2011.1 Screening for CRC in asymptomatic, average-risk populations reduces CRC incidence and mortality.2, 3 Current CRC screening guidelines vary slightly between organizations but generally recommend annual screening with high-sensitivity, home-based fecal occult blood tests (FOBTs) or endoscopic tests at longer intervals, including flexible sigmoidoscopy every 5 years or colonoscopy every 10 years.2, 3 Although colonoscopy procedures underpin any CRC screening program that aims to reduce CRC mortality, annual FOBT testing is a lower cost, minimal-risk modality that has demonstrated the ability to reduce CRC incidence and mortality in randomized controlled trials.2, 3 It has been estimated that screening with high-sensitivity stool tests, including guaiac-based or immunochemical tests, when adhered to annually, confer similar gains in life-years compared with colonoscopy every 10 years.4

During the past decade, the prevalence of guideline-consistent CRC screening (either FOBT in the past year or endoscopy in the past 10 years) has increased significantly in average-risk individuals aged ≥50 years.5 However, trends by screening modality indicate that home-based FOBT screening has declined significantly, whereas the use of colonoscopy has increased.5-8 Studies have established distinct socioeconomic (SES) and race/ethnic differences in guideline-consistent CRC screening trends in the elderly and nonelderly populations.9-11 However, few studies have investigated SES differences in trends in FOBT use among populations for which regular screening is recommended, and most of those studies focused on adults aged ≥65 years. Because most elderly adults are Medicare insured, those studies identified factors associated with disparities in FOBT trends that are independent of health insurance access, such as educational or income level, the presence of a usual source of care, and race and ethnicity.9, 12-16 To our knowledge, no study has investigated trends in FOBT use alone and compared with endoscopic screening trends in the non-Medicare-insured and uninsured population ages 50 to 64 years. Identifying these patterns and trends can inform policy and program interventions to increase CRC screening rates in particular SES and race/ethnic subgroups. To address this issue, we used nationally representative data on CRC screening rates for nonelderly adults (ages 50-64 years) between the years 2000 and 2008.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Data Source and Analytic Sample

Data for this study for 2000 and 2008 were obtained from the sample adult and person-level file in the National Health Interview Survey (NHIS), an annual, nationally representative, cross-sectional, household-based, multistage sample survey of noninstitutionalized individuals conducted by the National Center for Health Statistics (NCHS).17 Data for the year 2005 were obtained from the Cancer Control Module administered to sampled adults in the NHIS. Our analytic sample consisted of adults ages 50 to 64 years with no history of a CRC diagnosis (n = 6008 for the year 2000, n = 6687 for the year 2005, and n = 4970 for the year 2008). Survey response rates declined from 72.1% in 2000 to 62.6% in 2008. Adults aged <50 years were excluded, because current screening guidelines for average-risk adults do not recommend testing in this age group.

Outcome Variables

In the 3 study years, respondents were asked similar questions regarding lifetime receipt and recency of an FOBT with a home kit and endoscopic testing (colonoscopy, sigmoidoscopy, or proctoscopy). In 2005 and 2008, the skip pattern of the questions changed so that complete responses could be obtained on the questions measuring the time since cancer screening. To maintain consistency of the screening measures over time, we used a common method to estimate the time since cancer screening as recommended by the NCHS.18 Dichotomous variables (yes/no) were created for an FOBT in the past year and for an endoscopic test (proctoscopy, sigmoidoscopy or colonoscopy) in the past 10 years that were largely consistent with screening guideline recommendations.2, 3

Other Variables

We considered the following socioeconomic variables in our analyses: insurance status (private or military, public, uninsured, and other), usual source of care (yes/no), race and ethnicity (Hispanic, non-Hispanic white, non-Hispanic black, and other), poverty level (with poor defined as individuals at or below 99% of the federal poverty level, near-poor defined as those at or between 100% and 199% of the poverty threshold, and not poor defined as either at or between 200% and 399% of the poverty threshold or at or above 400% of the poverty threshold), education (less than high school, high school or General Educational Development Diploma [GED], some college or Associate degree, and college or higher), immigration status (born in the United States, in the United States for ≥10 years, and in the United States for <10 years). Missing income data that constituted >10% in any given year were imputed using multiple imputation methods. In addition to these variables, all adjusted analyses controlled for demographic variables (age, sex, region) and a measure of health care use (the number of physician office visits).

Statistical Analyses

Estimates of FOBT in the past year and endoscopy in the past 10 years were age-adjusted to the year 2000 standard, and linear contrasts were used to test for significant differences over time between 2000 and 2008 for subgroups of the SES and race and ethnicity variables. Adjusted prevalences (predicted marginals) were calculated using multiple logistic regression analyses, which controlled for the variables listed above. Differences in trends over time between subgroups of the SES variables were tested using interaction terms between year of survey and SES variable and linear contrasts of the adjusted prevalences. SAS-callable-SUDAAN (SAS Institute Inc., Cary, NC) was used to calculate weighted estimates and standard errors that take into account the complex survey design of the NHIS.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Sample Characteristics and Changes Between 2000 and 2008

In all 3 years, the majority of our analytic sample had private or military insurance, had a usual source of care, was non-Hispanic white, had a high school education or more years of education, was not poor, and was born in the United States (Table 1). Between 2000 and 2008, the proportion with private and military coverage declined, whereas public coverage increased; and the proportion of those with a high school degree or less than a high school degree declined, whereas those with some college or with a college degree or more increased. There was a decline in the proportion of non-Hispanic whites and the proportion of those born in the United States but an increase in the proportion living in the West and those living in the United States for ≥10 years.

Table 1. Characteristics of Adults Ages 50 to 64 Years Without a History of Colorectal Cancer: United States, 2000, 2005, and 2008
Characteristic200020052008
No.% (95% CI)No% (95% CI)No.% (95% CI)
  1. Abbreviations: CI, confidence interval; GED, General Educational Development Diploma.

Age, y      
 50-54260641.9 (40.5, 43.3)281239.9 (38.7, 41.2)192139.2 (37.7, 40.8)
 55-59209932.0 (30.7, 33.3)252333.6 (32.4, 34.8)183033.2 (31.7, 34.7)
 60-64175026.1 (24.9, 27.3)201726.5 (25.3, 27.7)149127.6 (26.1, 29)
Sex      
 Men287448.0 (46.6, 49.4)330148.3 (46.9, 49.6)233548.6 (47, 50.2)
 Women358152.0 (50.6, 53.4)405151.7 (50.4, 53.1)290751.4 (49.8, 53)
Insurance status      
 Private and military475278.5 (77.3, 79.7)533076.9 (75.7, 78)366973.8 (72.3, 75.4)
 Public6377.2 (6.5, 7.9)7828.2 (7.5, 8.9)71511.0 (10, 12)
 Uninsured88311.4 (10.5, 12.4)103812.3 (11.3, 13.2)71212.5 (11.4, 13.7)
 Other combinations1832.9 (2.4, 3.3)2022.7 (2.2, 3.2)1462.6 (2.2, 3.1)
Usual source of care      
 Usual source of care575290.9 (90.1, 91.7)653290.8 (90.1, 91.6)464190.3 (89.3, 91.3)
 No usual source of care6559.1 (8.3, 9.9)7429.2 (8.4, 9.9)5309.7 (8.7, 10.7)
Race/ethnicity      
 Hispanic7917.6 (6.8, 8.3)9058.4 (7.7, 9.1)6229.3 (8.3, 10.2)
 White, non-Hispanic457279.0 (77.8, 80.3)516577.1 (75.9, 78.2)352274.7 (73.2, 76.1)
 Black, non-Hispanic8879.7 (8.9, 10.6)101910.5 (9.7, 11.3)82611.2 (10.2, 12.2)
 Other2053.7 (3, 4.3)2634.1 (3.5, 4.7)2724.9 (4.2, 5.6)
Education level      
 High school130817.3 (16.2, 18.5)111012.9 (11.9, 13.9)74712.8 (11.6, 14)
 High school, GED194931.7 (30.3, 33.1)213130.5 (29.2, 31.9)148428.5 (27, 29.9)
 Some college, associate degree161025.3 (24.1, 26.6)201727.4 (26.2, 28.6)152429.3 (27.8, 30.8)
 ≥College151325.6 (24.3, 26.9)201529.1 (27.8, 30.4)145529.4 (27.9, 30.9)
Poverty level, % of poverty threshold      
 <99%779.28.5 (7.6, 9.3)8558.3 (7.6, 9.1)6449.2 (8.3, 10.1)
 100% to ≤199%975.213.4 (12.2, 14.5)1155.814.1 (13, 15.2)796.213.0 (11.8, 14.1)
 ≥200% to ≤399%1745.626.9 (25.5, 28.3)2029.227.2 (26, 28.5)142126.6 (25.1, 28.1)
 ≥400%295551.2 (49.4, 53)331250.4 (48.8, 52)2380.851.3 (49.4, 53.2)
Region      
 Northeast122119.8 (18.6, 21)131019.6 (18.3, 20.8)90818.2 (16.7, 19.7)
 Midwest147824.6 (23.3, 25.9)174124.6 (23.4, 25.8)120923.9 (22.2, 25.7)
 South239436.6 (35.1, 38.1)266435.1 (33.5, 36.6)190035.7 (33.6, 37.7)
 West136219.0 (17.6, 20.4)163720.8 (19.5, 22.1)122522.2 (20.6, 23.7)
Immigration status      
 Born in the US559589.7 (88.8, 90.6)630588.4 (87.6, 89.3)446787.4 (86.2, 88.5)
 In US ≥10 y6748.8 (8, 9.6)82410.1 (9.3, 10.9)61911.3 (10.2, 12.4)
 In US <10 y1071.5 (1.2, 1.9)1141.5 (1.1, 1.8)731.3 (1, 1.7)
No. of physician office visits      
 None95013.7 (12.7, 14.7)101613.5 (12.5, 14.5)70413.2 (12.1, 14.4)
 198716.2 (15.1, 17.3)111115.6 (14.6, 16.6)76115.4 (14.2, 16.7)
 2-5255241.5 (40.1, 43)300643.0 (41.6, 44.4)221444.2 (42.5, 45.9)
 ≥5186628.6 (27.2, 30)207027.9 (26.7, 29.2)145727.1 (25.5, 28.7)

Trends in Fecal Occult Blood Testing and Endoscopy Screening Between 2000 and 2008

From 2000 to 2008, there was a significant decline of 5.9% points in the rate of FOBT screening in our study population (95% confidence interval, 4.5%-7.3%; P < .001) (Table 2). For SES-related variables, there was no statistically significant change in FOBT prevalence among those without a high school degree (−1.9% points), the poor (−2.4% points), and the near poor (−1.5% points); whereas FOBT prevalence in those with a high school degree or more (high school, −6.5%; some college, −5.9% points; college or higher, −9% points) and the not poor (200% to ≤399% of poverty threshold, −6.9%; ≥400% of the poverty threshold, −7.1% points) declined significantly during this period. FOBT prevalence among respondents with public insurance (−2.7% points), the uninsured (−1.8% points), and those without a usual source of care (−0.9% points) did not change significantly between 2000 and 2008, whereas the privately insured (−6.6% points) and those with a usual source of care (−6.4% points) experienced significant declines. Immigrants to the United States, including those in the United States for ≥10 years (−0.3% points) and those in the United States for <10 years (2.6% points), did not experience significant change in FOBT rates; whereas those born in the United States (−6.6% points) experienced a significant decline. FOBT rates among Hispanics (−0.9% points) did not change significantly during this time, whereas non-Hispanic whites (−6.5% points) and blacks (−5.5% points) experienced significant declines. These results did not change after adjustment for other factors.

Table 2. Fecal Occult Blood Test (FOBT) in the Past Year and Trends in FOBT in the Past Year and in Endoscopy in the Past 10 Years in Adults Ages 50 to 64 Years Without a History of Colorectal Cancer: United States, 2000, 2005, and 2008
CharacteristicFOBT in the Past YearEndoscopy in the Past 10 Yearsa
200020052008Difference Between 2000 and 2008
No.% (95% CI)No.% (95% CI)No.% (95% CI)% Points (95% CI)% Points (95% CI)
  • Abbreviations: CI, confidence interval; GED, General Educational Development Diploma.

  • a

    Endoscopy included proctoscopy, sigmoidoscopy, or colonoscopy within the past 10 years.

  • b

    P < .001.

  • c

    P < .05.

  • d

    P < .01.

Age, y        
 50-54244213.4 (11.8, 15.1)25598.4 (7.2, 9.7)18346.9 (5.7, 8.4)−6.4 (−8.5, −4.4)b13.6 (10.2, 17)b
 55-59194016.4 (14.5, 18.4)229611.7 (10.2, 13.3)17279.4 (8, 10.9)−7.0 (−9.4, −4.6)b21.0 (17.1, 24.8)b
 60-64162616.8 (14.7, 19.1)183214.3 (12.5, 16.2)140913.2 (11.3, 15.5)−3.5 (−6.6, −0.5)c18.9 (14.8, 22.9)b
Sex        
 Men271014.0 (12.5, 15.6)300810.1 (9, 11.4)22259.3 (8, 10.8)−4.6 (−6.7, −2.6)b15.4 (12.2, 18.6)b
 Women329816.4 (15, 18)367911.7 (10.5, 13.1)27459.4 (8.2, 10.7)−7.1 (−9, −5.1)b19.1 (16.4, 21.8)b
Insurance status        
 Private and military444916.6 (15.5, 17.8)484212.1 (11.1, 13.2)350010.0 (8.9, 11.3)−6.6 (−8.3, −5)b20.8 (18.3, 23.3)b
 Public58110.8 (8.3, 13.9)70610.0 (7.6, 13.2)6628.1 (5.9, 10.9)−2.7 (−6.4, 0.9)11.3 (5, 17.6)b
 Uninsured8137.0 (5.2, 9.5)9644.0 (2.9, 5.5)6745.3 (3.5, 8)−1.8 (−4.8, 1.3)4.4 (−0.5, 9.2)
 Other combinations16517.1 (11, 25.7)17515.4 (9.3, 24.5)13414.7 (9, 23.1)−2.4 (−12.5, 7.7)17.7 (2.2, 33.2)c
Usual source of care        
 Usual source of care539116.4 (15.3, 17.5)599711.8 (10.9, 12.7)446710.0 (9, 11)−6.4 (−7.9, −4.9)b19.0 (16.7, 21.3)b
 No usual source of care6144.0 (2.7, 5.8)6872.6 (1.5, 4.5)5033.1 (1.5, 6.1)−0.9 (−3.5, 1.8)2.9 (−2.2, 8)
Race/ethnicity        
 Hispanic7309.0 (6.6, 12.1)8257.4 (5.3, 10.1)5908.3 (5.9, 11.5)−0.7 (−4.5, 3.2)8.4 (2.2, 14.6)d
 White, non-Hispanic428416.0 (14.9, 17.2)472711.6 (10.6, 12.7)33519.6 (8.5, 10.8)−6.5 (−8.1, −4.8)b18.3 (15.8, 20.8)b
 Black, non-Hispanic81414.2 (11.3, 17.8)89610.0 (8, 12.5)7708.7 (6.6, 11.4)−5.5 (−9.6, −1.5)d17.9 (12.2, 23.5)b
 Other18013.2 (8.2, 20.7)2398.0 (4.8, 12.9)2599.3 (6, 14.2)−3.9 (−11.3, 3.4)26.4 (16.1, 36.7)b
Education level        
 <High school12039.3 (7.5, 11.4)10218.6 (6.5, 11.3)7097.4 (5.3, 10.2)−1.9 (−4.9, 1.2)8.8 (4.1, 13.4)b
 High school, GED182213.8 (12.1, 15.6)194010.8 (9.4, 12.4)13977.3 (6, 8.8)−6.5 (−8.7, −4.2)b15.9 (12.1, 19.8)b
 Some college, associate degree152917.0 (15.1, 19.1)184611.4 (9.8, 13.1)145311.1 (9.3, 13.2)−5.9 (−8.7, −3.2)b17.3 (13.1, 21.4)b
 ≥College141619.4 (17, 22)183212.2 (10.7, 13.8)138510.4 (8.6, 12.5)−9.0 (−12.1, −5.8)b19.3 (15, 23.7)b
Poverty level, % of poverty threshold
 <99%7069.6 (7.1, 12.9)749.86.3 (4.6, 8.5)593.87.2 (5.2, 10.1)−2.4 (−6.1, 1.3)5.7 (−0.6, 12)
 100% to ≤199%898.611.0 (8.4, 14.5)1046.410.7 (8.5, 13.4)761.29.5 (7.2, 12.5)−1.5 (−5.6, 2.5)12.0 (6.5, 17.4)b
 ≥200% to ≤399%162315.2 (13.4, 17.3)186811.8 (9.8, 14.1)13588.3 (6.9, 10.1)−6.9 (−9.5, −4.3)b15.9 (11.7, 20.2)b
 ≥400%2780.417.2 (15.6, 18.8)3022.811.4 (10.2, 12.8)225710.1 (8.7, 11.7)−7.1 (−9.2, −4.9)b21.2 (18.2, 24.2)b
Region        
 Northeast112914.0 (12.1, 16.1)115811.3 (9.8, 13.1)8368.1 (6.2, 10.5)−5.9 (−8.9, −3)b22.0 (16.5, 27.4)b
 Midwest139017.3 (15.5, 19.3)160710.6 (8.9, 12.5)11489.0 (7.3, 11.2)−8.3 (−11, −5.5)b17.4 (13, 21.9)b
 South224714.0 (12.4, 15.8)242210.4 (9.1, 11.9)18149.1 (7.5, 10.9)−5.0 (−7.4, −2.5)b17.6 (14.1, 21)b
 West124216.2 (13.7, 19)150012.1 (10.1, 14.4)117211.2 (9.6, 13.1)−4.9 (−8.1, −1.8)d13.0 (7.7, 18.3)b
Immigration status        
 Born in the US523216.1 (15, 17.2)574311.4 (10.5, 12.4)42539.4 (8.5, 10.5)−6.6 (−8.1, −5.1)b18.1 (15.8, 20.4)b
 In US ≥10 y6219.8 (7.2, 13.3)7458.0 (6.1, 10.5)5769.5 (7.3, 12.4)−0.3 (−4.2, 3.7)13.9 (7.5, 20.4)b
 In US <10 y991.4 (0.2, 9.1)1044.7 (1.8, 11.8)664.0 (1.5, 10.3)2.6 (−2.2, 7.3)13.0 (−1.8, 27.8)
No. of physician office visits
 None8821.8 (1, 3.3)9511.6 (0.9, 3)6922.5 (1.3, 4.6)0.7 (−1.2, 2.5)6.5 (2.3, 10.7)d
 191512.5 (10.1, 15.4)10198.8 (6.9, 11.1)7329.1 (6.9, 11.9)−3.4 (−7.1, 0.2)12.7 (7.5, 17.9)b
 2-5240317.0 (15.4, 18.8)277512.5 (11.2, 13.9)21399.5 (8.2, 11.1)−7.5 (−9.7, −5.3)b21.3 (17.9, 24.6)b
 ≥5177920.7 (18.7, 22.9)192514.5 (12.7, 16.5)138712.4 (10.6, 14.5)−8.3 (−11.2, −5.4)b18.9 (14.9, 22.9)b

To explore whether declines in FOBT rates in certain subgroups were coincident with increases in endoscopic procedures in these subgroups, we compared endoscopic screening trends in the past 10 years with FOBT screening trends (Table 2). The higher SES groups that experienced significant declines in FOBT rates consistently experienced significant increases in endoscopic screening rates during this period. The pattern with lower SES groups that experienced no change in FOBT screening rates was more mixed. Significant modest increases in endoscopic screening rates in a few lower SES groups, such as the publicly insured (11.3% points), lower educated (8.8% points), near poor individuals (12% points), long-term immigrants (13.9% points), and Hispanics (8.4% points), offset the stable FOBT prevalence in these groups. However, in most cases, the absolute increase in endoscopic screening rates in these subgroups was lower than the increase in higher SES subgroups within their variable categories.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

The primary finding of this study is that, although FOBT use declined significantly in the past decade, there were differences according to SES; ie, these trends were not uniform between SES and racial/ethnic subgroups. Hispanics and groups with lower SES levels, such as the lower educated, the poor and near-poor, the uninsured and publicly insured, those without a usual source of care, and immigrants to the United States, did not experience significant changes in FOBT prevalence from 2000 to 2008. Conversely, those with higher SES levels and non-Hispanic race groups generally experienced significant declines in FOBT use during the same period; and, at the same time, reports of endoscopic screening increased significantly, indicating migration from FOBT to endoscopic procedures. However, there was no consistent pattern in endoscopic screening trends in lower SES groups. These findings indicate that, although there may have been some shift from FOBT to endoscopic screening procedures in Hispanics and lower SES groups, it was less than the shift observed in higher SES groups.

Evidence for FOBT tests as an effective CRC screening modality comes from randomized controlled trials that have demonstrated reductions in CRC incidence and mortality.19-22 FOBT efficacy is achieved through the detection of occult blood in otherwise asymptomatic individuals, a sign of advanced adenomas or cancer that more commonly will be at an early and curable stage compared with symptomatic disease.23, 24 Although FOBT is regarded as a simple, low-cost alternative to endoscopy, barriers include the need for annual testing, dietary restrictions required by some guaiac-based tests, and the need for colonoscopy when the test is positive. Technology and systems-level limitations include poor sensitivity associated with some tests; poor specimen collection; inadequate processing, test development, and interpretation; and the continued prevalence of single-sample, office-based stool testing. Despite these limitations, the use of effective, lower cost alternatives to endoscopy, such as the high-sensitivity guaiac-based or immunochemical FOBT, may represent the best near term option for some subgroups of adults, including those with lower access to specialty resources because of cost or other factors and those unable or unwilling to undergo endoscopic screening procedures. In this context, the lack of decline in FOBT screening rates in lower SES and Hispanic groups may represent a positive indicator for CRC early detection efforts in the United States, because it suggests the continued acceptance of FOBT screening in these populations. Although these trends are encouraging, it must be noted that FOBT continues to be an underused screening strategy. To better serve the purpose of achieving population CRC screening goals established by national organizations and the federal government, these patterns must be understood in the context of the consistent uptake of colonoscopy during the same period.25-27

Several factors may have influenced national uptake of colonoscopy over FOBT, including provider, patient, and systems level factors. It has been demonstrated that physicians often consider colonoscopy a more optimal test compared with FOBT.23, 24 Conversely, studies of patient test preferences do not indicate a clear inclination for colonoscopy over FOBT or vice versa.28-31 However, studies demonstrate the troubling disparity between patients' stated preference for a specific test and the test that ultimately is administered, with colonoscopy most often prescribed to patients who state a preference for FOBT or other tests.28, 29 Physician preference for and prescription of colonoscopy may be an important reason for increasing colonoscopy rates and low or unchanged FOBT rates. More worrisome, however, is that the discordance between physician recommendations and patient preferences may be contributing to low rates of patient adherence to overall CRC screening recommendations. Because some of these patterns may be influenced by a lack of awareness regarding the relative performance of the 2 screening types, physicians and patients should be made aware that simulations have estimated similar mortality reductions for adults with long-term adherence to high-sensitivity FOBT compared with colonoscopy.4, 32 In addition, the presence of screening modalities other than FOBT and endoscopic tests in the health care marketplace, including flexible sigmoidoscopy and computed tomographic colonography, makes it even more important that there is adequate, high-quality physician-patient discussion about these options, allowing patients to make informed decisions based on their preferences and risks.

Health care and systems level factors also may play a role in the differential uptake of tests. For example, low physician reimbursement rates for FOBT compared with colonoscopy may be a contributing factor.33 Systems and technology infrastructure concerns, such as lack of resources to deliver quality, high-sensitivity FOBT testing, lack of office systems to monitor and follow-up initial FOBT referrals, and potential follow-up examinations, also may arise. To better inform programmatic efforts to improve population FOBT testing, future research must delineate how the complexity and magnitude of these above-mentioned factors influence FOBT versus colonoscopy delivery and use.

Race and ethnic and SES disparities in CRC screening have been documented in the literature.6, 8-11 Irrespective of test type, it is notable that Hispanics are significantly less likely than other non-Hispanic race groups to obtain CRC screening, Reasons for racial/ethnic disparities may include lower SES level, language and cultural barriers associated with recent immigration, and differences in preventive health-related attitudes, beliefs, and health-seeking behaviors.34, 35 Factors like reduced health care access and cost-related barriers contribute to both racial/ethnic and SES disparities in CRC screening.34, 36, 37 Less clear are the reasons for these subgroup differences in trends by test type, ie, FOBT versus colonoscopy trends. It is possible that the factors influencing differences in CRC screening between these groups potentially may influence the choice of 1 modality over the other. In addition, some (but not all) studies have demonstrated SES and race/ethnic differences in patient preference for 1 test modality over another, with Hispanics and lower SES groups demonstrating a preference for FOBT over colonoscopy.29 Some studies also have reported lower rates of physician recommendation for CRC screening among Hispanics and lower SES groups.38, 39 However, the extent to which physicians differentially recommend 1 screening test over the other based on patient characteristics, including SES, demographics, and CRC risk, remains unknown.

Racial/ethnic and SES subgroups also may respond differently to factors in the health care environment that may predict CRC screening type. Studies in the Medicare population have demonstrated that the introduction of colonoscopy screening coverage has led to increases in screening disparities between white and Hispanic populations compared with a period when only FOBT was covered.35, 40 This may be the case because, even among insured populations, the burden of cost-sharing requirements may force lower SES groups to opt for lower cost procedures, such as FOBT, versus more expensive colonoscopy screening.41 In privately insured and Medicaid-insured populations, state legislation ensuring coverage of different CRC screening modalities has the possibility of differentially impacting SES subgroups, with certain higher SES groups accessing endoscopic procedures at higher rates than FOBT.42 It has generally been acknowledged that CRC screening programs that address determinants in the health care system in order to serve the specific needs of racial/ethnic minorities and lower SES groups must recognize their heterogeneous backgrounds that are sensitive to cultural, language, and literacy levels. However, it may be more important that future research efforts effectively delineate the complex interplay of factors at the individual and health care system level that are differentially influencing the uptake of test type between racial/ethnic and SES groups to increase overall CRC screening.

In terms of the limitations of the current study, despite evidence demonstrating that patient recall of receiving preventive services is superior to medical records, self-reported data from the NHIS still may be subject to recall and social desirability biases.43 Small sample sizes precluded us from estimating CRC test use and trends in minority racial and ethnic groups like American Indians/Alaska Natives and Asian Americans.

In conclusion, the current study indicates that FOBT declines in the United States reported in the past decade did not occur among lower SES groups and Hispanics. Annual screening with high-sensitivity FOBT represents an appropriate and low-cost alternative to colonoscopy screening.2, 3 Although, ideally, all adults should have a choice among test options, the potential to achieve high near-term rates of population CRC screening with high-sensitivity stool tests in populations with less access to colonoscopy is a sensible and affordable public health strategy. However, population programs promoting high-sensitivity FOBT testing for CRC screening must fully consider the role of patient and physician attitudes, knowledge, and preferences and systems issues related to cost, referral, adherence, test quality assurance, and adequate follow-up of positive test results if these programs are to be effective alternatives to colonoscopy programs.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

This study was funded by the Intramural Research Program, American Cancer Society.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES