SEARCH

SEARCH BY CITATION

Keywords:

  • colorectal neoplasms;
  • prevention and control;
  • mass screening;
  • ethnic groups;
  • statistics and numeric data;
  • endoscopy;
  • utilization;
  • occult blood;
  • California

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

BACKGROUND

Recent research has supported the use of colorectal cancer (CRC) tests to reduce disease incidence, morbidity, and mortality. A new health survey has provided an opportunity to examine the use of these tests in California's ethnically diverse population. The authors used the 2001 California Health Interview Survey (CHIS 2001) to evaluate 1) rates of CRC test use, 2) predictors of the receipt of tests, and 3) reasons for nonuse of CRC tests.

METHODS

The CHIS 2001 is a random-digit dial telephone survey that was conducted in California. Responses were analyzed from 22,343 adults age ≥ 50 years. CRC test use was defined as receipt of a fecal occult blood test in the past year and/or receipt of an endoscopic examination in the past 5 years.

RESULTS

Nearly 54% of California adults reported receipt of a recent CRC test. Insurance coverage and having a usual source of care were the most important predictors of CRC testing. Latinos age < 65 years were less likely to be tested than whites (relative risk [RR], 0.84; 95% confidence interval [95% CI], 0.77–0.92). Men were more likely to be tested than women, an effect that was greater among individuals age 50–64 years (RR, 1.28; 95% CI, 1.23–1.32) than among individuals age ≥ 65 years (RR, 1.19; 95% CI, 1.15–1.23). Women were more likely than men to say that their physician did not inform them the test was needed and that CRC tests were painful or embarrassing.

CONCLUSIONS

Results of the current study indicate a need for physicians to recommend CRC testing to their patients. Assuring that all individuals have both health insurance and a usual source of care would help address gaps in the receipt of CRC tests. Cancer 2004. © 2004 American Cancer Society.

Colorectal cancer (CRC) is the third most common cancer both in California and nationally. In California, there were approximately 7400 new diagnoses of CRC and 2600 deaths from CRC in 2001.1 Patient outcomes are highly dependent on the stage of disease at the time of diagnosis and surgery. When CRC is detected early, there is a reported 90% 5-year survival rate; among patients with advanced disease, the 5-year survival rate is reduced to 9%.2

CRC screening is recommended by the American Cancer Society, the Centers for Disease Control and Prevention, the National Cancer Institute, and the U.S. Preventive Services Task Force. The specific guidelines proposed by each organization differ, but all recommend regular fecal occult blood testing (FOBT) and flexible sigmoidoscopy to reduce the disease burden of CRC for individuals age ≥ 50 years.3–6

Despite the demonstrated benefits of CRC screening, the proportion of the population that receives CRC examinations remains quite low. In the 2000 National Health Interview Survey (NHIS), only 41% of men and 38% of women were screened recently.7 Past studies suggest that there are significant barriers to the widespread uptake of CRC screening and that these barriers may be more pronounced among minorities, females, individuals with low income, older adults, and rural residents.8–14 Using data from the NHIS, researchers consistently have found CRC test use differences according to race, education, insurance status, and whether an individual has a usual source of care (USOC).3, 7, 11

California is the most populous state in the U.S., and the 2001 California Health Interview Survey (CHIS 2001) is the largest state population-based health survey in the nation. The 1999 Behavior Risk Factor Surveillance Survey had approximately 1500 responses from California, and the 2000 NHIS surveyed 39,000 households nationwide.7, 15 The CHIS 2001, which is much larger, includes data from > 55,000 households, with a sample design that has national significance. It informs policy at the national level, because it provides the only available data regarding several racial and ethnic groups with insufficient samples in national surveys, such as American Indians/Alaska Natives, Pacific Islanders, and Asians. In addition, the CHIS 2001 was administered in several other languages, resulting in the participation of populations that have been excluded historically from English-only administered population-based surveys.16 Inclusion of these racial, ethnic, and linguistic minorities is particularly important in accurately evaluating the nation's progress in cancer screening. For the current study, we examined individual-level predictors of CRC test use and analyzed the reasons identified by different gender, racial, and ethnic groups for not undergoing CRC tests. Together, these analyses contribute to the CRC screening literature by identifying the groups most at risk for not receiving CRC screening tests and by providing insights into the reasons for varying screening rates.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

CHIS 2001

The CHIS 2001 sample and questionnaire were designed to represent California's ethnically diverse population: 55,428 households were selected randomly from within the state for a random-digit dial telephone survey. From each participating household, one randomly selected adult was interviewed. Respondents age ≥ 50 years were asked about their use of CRC tests. Data from the survey were weighted to the 2000 Census at the county and state levels. The completed interview participation rate was 63.7%. More detailed descriptions about the methods have been published elsewhere.17–20 CHIS was approved by the University of California at Los Angeles Institutional Review Board and by the California State Committee for the Protection of Human Subjects.

Inclusion Criteria

Individuals age ≥ 50 years without a personal history of CRC were included in this study. Respondents for whom receipt of a colorectal test could not be determined as a result of having responded “refused” or “don't know” to questions concerning testing were excluded from all analyses.

Dependent Variable: Receipt of Testing

Survey respondents were considered tested if an FOBT was performed in the 12 months prior to the interview or if either a flexible sigmoidoscopy or colonoscopy was performed within 5 years prior to the interview. These frequency criteria parallel those used in analyses of NHIS data.3–6

For the purposes of this analysis, we made no distinction regarding the reason an individual underwent testing. Classically, screening refers to the use of a test in an individual without symptoms. To avoid potential misclassification, the receipt of colorectal examinations for any reason was used as the outcome of interest and is reported as “testing” instead of “screening.”

Variables

We examined the extent to which demographic factors, socioeconomic status (SES), health insurance coverage, access to care, health status, and acculturation factors predicted CRC test use in a population of adults age ≥ 50 years. Many studies have shown that demographic characteristics (such as age, gender, and race/ethnicity) and indicators of SES are associated with the use of health care services.21–24 Measures of access to care, health insurance coverage, and the presence of a USOC also are especially strong predictors of the receipt of preventive health care, such as cancer screening.3, 7, 24, 25 Finally, because of California's large immigrant population, we examined factors of acculturation. Studies have suggested that, among immigrants, English proficiency and recency of immigration to the U.S. are important factors in CRC screening.7, 26

Demographic variables

Age was modeled as a continuous variable. Race and ethnicity were tabulated into six mutually exclusive categories (white, Latino, Asian, African American, other/multiracial, and American Indian/Alaskan Native). Marital status was analyzed as currently married, never married, or other.

SES and access-to-care indicators

The socioeconomic variables examined included education and income. Education was grouped into four levels based on the highest level of education attained by the respondent (less than high school graduation, high school graduation, some college, and college degree or greater). Income was categorized based on the federal poverty level (FPL). The categories were < FPL, 100–199% of FPL, 200–299% of FPL, and ≥ 300% of FPL.

Access to care was constructed as a composite variable combining insurance status with the presence (or absence) of a USOC. Different variables were constructed for individuals age < 65 years versus individuals age > 65 years. For respondents age < 65 years, insurance categories included employer-based, privately purchased, Medicaid/other public, or uninsured. For respondents age ≥ 65 years, insurance categories included Medicare and other, Medicare and Medicaid, Medicare only, no insurance (uninsured), or other only. Respondents were considered to have a USOC if they had a regular provider of care that was not an emergency department or an urgent care clinic.

Health status and utilization

Self-rated health status was included as a dichotomous variable (fair/poor vs. good/very good/excellent). The number of physician visits in the 12 months prior to the interview was categorized as 0 visits, 1–2 visits, 3–4 visits, 5–9 visits, or > 10 visits.

Acculturation

Two variables were employed as indicators of acculturation. English proficiency was evaluated as a dichotomous variable (proficient in English vs. limited English proficiency). The percent of life years lived in the U.S. was constructed as a categoric variable based on the number of years lived in the U.S. and the respondent age.26 Respondents were grouped into 3 categories (100% of life lived in the U.S., 0–50% of life lived in the U.S., and 51–99% of life lived in the U.S.).

Reasons for Nonparticipation in Testing

Respondents who reported that they did not undergo testing were asked the main reason for nonparticipation. One reason was obtained for each test modality. Frequencies of each reason for nonparticipation were compared for differences between gender and racial and ethnic groups using chi-square analysis. Racial/ethnic differences were based on comparisons between each group and the reference group (white). Reasons for nonparticipation in FOBT and endoscopic testing were compared separately, because respondents could indicate a different reason for each modality.

Statistical Analysis

Weighted multivariate logistic regression was used to analyze each respondent's likelihood of undergoing CRC testing. Two regression models were estimated, 1 for respondents ages 50–64 years and 1 for respondents age > 65 years, to account for age-related differences in health insurance coverage. For example, most adults age < 65 years are covered by employer-based health insurance, and most adults age > 65 years are covered by Medicare.

Data analyses were performed with the Survey Data Analysis statistical package and SAS statistical software (version 8.01; SAS Institute, Cary, NC). Because the outcome was relatively common (>10%), estimated relative risks (RR) and 95% confidence intervals (95% CI) were calculated using the adjusted odds ratios from the logistic regression models.27

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Table 1 presents the characteristics of the 22,343 respondents who met the inclusion criteria. Of these, 12,211 respondents (55%) were ages 50–64 years, and 10,132 respondents (45%) were age > 65 years. Women comprised 59.9% of the sample. The majority of respondents were white (80%); and the remainder of the sample consisted of Latinos (6%), Asians (5%), African Americans (5%), other/multiracial (3%), and American Indians/Alaska Natives (2%). Greater than half of respondents (55%) reported an income ≥ 300% of the FPL, 33% of respondents had attained at least a college degree, and 3% of respondents reported limited English proficiency.

Table 1. Sample Characteristics, California 2001ab
CharacteristicUnweighted no. (unweighted %)
  • FPL: Federal poverty level; USOC: usual source of care.

  • a

    Data were obtained from the 2001 California Health Interview Survey.

  • b

    The sample included respondents age ≥ 50 years with no personal history of colorectal cancer.

Age 
 50–64 yrs12,211 (54.7)
 ≥ 65 yrs10,132 (45.3)
Gender 
 Female13,394 (59.9)
 Male8949 (40.1)
Race 
 White17,971 (80.4)
 Latino1420 (6.4)
 Asian1005 (4.5)
 African American1042 (4.7)
 Other/multiracial572 (2.6)
 American Indians/Alaska Natives333 (1.5)
Marital status 
 Married11,525 (51.6)
 Never married1234 (5.5)
 Other (divorced, separated, widowed)9584 (42.9)
Income 
 ≥300% FPL12,194 (54.6)
 200–299% FPL3493 (15.6)
 100–199% FPL4558 (20.4)
 0–99% FPL2098 (9.4)
Education 
 ≥ College degree7411 (33.2)
 Some college6623 (29.6)
 High school5787 (25.9)
 < High school2522 (11.3)
Self-reported health status 
 Good/very good/excellent17,308 (77.5)
 Fair/poor5035 (22.5)
Insurance status 
 Respondents ages 50–64 yrs 
  Without USOC 
   Uninsured414 (3.4)
   Medicaid or other public insurance76 (0.6)
   Privately purchased100 (0.8)
   Employer-based367 (3.0)
  With USOC 
   Uninsured837 (6.9)
   Medicaid or other public insurance1473 (12.1)
   Privately purchased1071 (8.8)
   Employer-based7873 (64.5)
 Respondents age ≥ 65 yrs 
  Without USOC 
   Uninsured13 (0.1)
   Other15 (0.2)
   Medicare78 (0.8)
   Medicare and Medicaid73 (0.7)
   Medicare and other185 (1.8)
  With USOC 
   Uninsured23 (0.2)
   Other363 (3.6)
   Medicare637 (6.3)
   Medicare and Medicaid1604 (15.8)
   Medicare and other7141 (70.5)
No. of visits to physician in last 12 mos 
 None2171 (9.7)
 1–27796 (34.9)
 3–45489 (24.6)
 5–93901 (17.4)
 ≥ 102986 (13.4)
Percent of life lived in the U.S. 
 100%19,527 (87.4)
 51–99%1604 (7.2)
 0–50%1212 (5.4)
English proficiency 
 Limited English proficiency767 (3.4)
 No stated limits21,576 (96.6)

Unadjusted Rates of Test Use

Nearly 54% of respondents in the sample reported receipt of a recent test for CRC. Table 2 presents unadjusted testing rates according to our independent variables for adult respondents age 50–64 years and age > 65 years. Older adults were more likely to receive testing than younger adults; 62% of respondents age ≥ 65 years had a recent test, compared with 48% of respondents ages 50–64 years. In both age groups, men were more likely to have received a test than women.

Table 2. Unadjusted Rates of Recent Colorectal Cancer Testingab
CharacteristicPercentage of patients
Ages 50–64 yrsAge ≥ 65 yrs
  • FPL: Federal poverty level; USOC: usual source of care.

  • a

    Data were obtained from the 2001 California Health Interview Survey.

  • b

    The sample included respondents age ≥ 50 years with no personal history of colorectal cancer.

  • c

    The estimate was not reliable statistically, because there were too few observations.

Overall48.061.8
Gender  
 Female43.056.4
 Male53.469.2
Race  
 White50.963.0
 Latino33.051.9
 Asian41.755.1
 African American50.162.1
 Other/multiracial47.561.0
 American Indians/Alaska Natives42.463.5
Marital status  
 Married51.166.6
 Never married37.355.4
 Other (divorced, separated, widowed)42.355.8
Income  
 ≥ 300% FPL52.266.4
 200–299% FPL43.964.1
 100–199% FPL39.956.9
 0–99% FPL31.251.8
Education  
 ≥ College degree54.368.6
 Some college46.663.6
 High school44.157.7
 < High school34.053.7
Self-reported health status  
 Good/very good/excellent49.163.3
 Fair/poor43.457.7
Insurance status  
 Respondents ages 50–64 yrs  
  Without USOC  
   Uninsured8.2 
   Medicaid or other public insurancec 
   Privately purchased22.5 
   Employer-based23.0 
  With USOC  
   Uninsured25.7 
   Medicaid or other public insurance49.3 
   Privately purchased51.1 
   Employer-based53.4 
 Respondents age ≥ 65 yrs  
  Without USOC  
   Uninsured c
   Other c
   Medicare c
   Medicare and Medicaid c
   Medicare and other 30.6
  With USOC  
   Uninsured c
   Other 60.5
   Medicare 54.1
   Medicare and Medicaid 54.1
   Medicare and other 66.1
No. of visits to physician in last 12 mos  
 None19.525.1
 1–248.862.6
 3–454.764.3
 5–9 visits55.965.1
 ≥ 10 visits54.265.5
Percent of life lived in the U.S.  
 100%50.263.0
 51–99%45.459.5
 0–50%33.849.1
English proficiency  
 Limited26.645.5
 No stated limits49.562.6

For both age groups, white and African-American individuals were more likely to be tested than individuals from other racial and ethnic groups. Among younger adults, 51% of whites and 50% of African Americans reported recent CRC testing; among older adults, the rates were 63% for whites and 62% for African Americans. Asians and Latinos had the lowest screening rates. Only 33% of younger Latinos and 52% of older Latinos had undergone a recent examination. Among Asians, only 42% of the younger group and 55% of the older group had undergone a recent test.

Adults ages 50–64 years consistently were more likely to have been tested regardless of their health insurance status if they reported having a USOC. Among respondents with employer-based health insurance coverage, 53% of individuals with a USOC underwent testing, compared with 23% of individuals with no USOC. Lack of insurance combined with no USOC led to dramatically poorer rates of test use. Only 8% of uninsured individuals with no USOC reported a recent test.

Most older adults age ≥ 65 years were covered by Medicare plus supplemental insurance (71%). Within this insurance group, having a USOC was associated with higher rates of screening (31% for respondents with no USOC compared with 66% for respondents with a USOC).

Multivariate Results

Table 3 presents the results of our multivariate models for predicting receipt of CRC screening for adults ages 50–64 years and age ≥ 65 years. Men in both age groups were more likely to be tested than women. This effect was greater in respondents ages 50–64 years (RR, 1.28; 95% CI, 1.23–1.32) than in respondents age ≥ 65 years (RR, 1.19; 95% CI, 1.15–1.23). Among respondents age < 65 years, increasing age was associated positively with screening (RR for 5-year interval, 1.49; 95% CI, 1.43–1.57). In respondents age ≥ 65 years, this effect reversed direction—older respondents were less likely to be screened (RR for 5-year interval, 0.91; 95% CI, 0.88–0.94).

Table 3. Predictors of Recent Colorectal Cancer Testingab
CharacteristicRR (95% CI)
Ages 50–64 yrsAge ≥ 65 yrs
  • RR: risk ratio; 95% CI: 95% confidence interval; FPL: Federal poverty level; USOC: usual source of care.

  • a

    Data were obtained from the 2001 California Health Interview Survey.

  • b

    The sample included respondents age ≥ 50 years with no personal history of colorectal cancer. The results are listed as adjusted RRs with 95% confidence intervals.

  • c The age effect is the risk ratio for a 5-year increment (i.e., individuals age 60 years vs. individuals age 55 years).

Agec1.49 (1.43–1.57)0.91 (0.88–0.94)
Gender  
 Female1.001.00
 Male1.28 (1.23–1.32)1.19 (1.15–1.23)
Race  
 White1.001.00
 Latino0.84 (0.77–0.92)0.95 (0.85–1.05)
 Asian0.92 (0.84–1.01)0.98 (0.89–1.05)
 African American1.03 (0.94–1.11)1.04 (0.97–1.11)
 Other/multiracial0.96 (0.84–1.08)1.02 (0.91–1.13)
 American Indian/Alaska Native0.93 (0.65–1.22)1.06 (0.76–1.30)
Marital status  
 Married1.001.00
 Never married0.83 (0.75–0.92)0.88 (0.78–0.97)
 Other (divorced, separated, widowed)0.91 (0.86–0.95)0.93 (0.89–0.96)
Income  
 ≥ 300% FPL1.001.00
 200–299% FPL0.89 (0.83–0.96)1.04 (1.00–1.09)
 100–199% FPL0.90 (0.83–0.96)1.00 (0.95–1.04)
 0–99% FPL0.81 (0.72–0.91)1.00 (0.94–1.06)
Education  
 ≥ College degree1.001.00
 Some college0.87 (0.83–0.92)0.97 (0.93–1.02)
 High school0.85 (0.79–0.90)0.89 (0.85–0.94)
 > High school0.87 (0.78–0.97)0.91 (0.85–0.97)
Self-reported health status  
 Good/very good/excellent1.001.00
 Fair/poor0.96 (0.91–1.02)0.92 (0.88–0.96)
Insurance status  
 Respondents ages 50–64 yrs  
  Without USOC  
   Uninsured0.32 (0.23–0.43) 
   Medicaid or other public insurance0.54 (0.33–0.82) 
   Privately purchased0.52 (0.34–0.75) 
   Employer-based0.53 (0.43–0.63) 
  With USOC  
   Uninsured0.61 (0.53–0.69) 
   Medicaid or other public insurance1.02 (0.94–1.10) 
   Privately purchased0.98 (0.91–1.06) 
   Employer-based1.00 
 Respondents age ≥ 65 yrs  
  Without USOC  
   Uninsured 0.08 (0.00–1.21)
   Other 0.51 (0.13–1.14)
   Medicare 0.41 (0.24–0.64)
   Medicare and Medicaid 0.59 (0.40–0.81)
   Medicare and other 0.58 (0.46–0.72)
  With USOC  
   Uninsured 0.62 (0.37–0.92)
   Other 0.90 (0.81–0.98)
   Medicare 0.86 (0.79–0.93)
   Medicare and Medicaid 0.88 (0.83–0.93)
   Medicare and other 1.00
No. of visits to physician in last 12 mos  
 None1.001.00
 1–21.55 (1.49–1.61)1.41 (1.37–1.45)
 3–41.66 (1.60–1.73)1.47 (1.42–1.50)
 5–91.71 (1.64–1.77)1.46 (1.42–1.50)
 ≥ 101.71 (1.64–1.78)1.48 (1.43–1.51)
Percent of life lived in the U.S.  
 100%1.001.00
 51–99%1.07 (1.00–1.15)0.96 (0.90–1.02)
 0–50%0.96 (0.87–1.05)0.86 (0.76–0.97)
English proficiency  
 Limited0.94 (0.82–1.07)1.00 (0.88–1.11)
 No stated limits1.001.00

Among adults age < 65 years, Latinos were the only racial/ethnic group that was significantly less likely than whites to have received recent testing (RR, 0.84; 95% CI, 0.77–0.92). For older individuals, race/ethnicity was not statistically significant.

Income was associated with rates of testing in younger individuals, but not in older individuals. Among adults ages 50–64 years, there was a poverty gradient associated with test use. Respondents living below the FPL were significantly less likely to be tested than the highest income group (RR, 0.81; 95% CI, 0.72–0.91). Respondents in intermediate income groups (100–299% of FPL) had rates of testing between the lowest and the highest income groups. Among older individuals, income level was not found to be predictive of testing. In both age groups, higher levels of education were correlated with a greater likelihood of testing.

Health insurance status was a significant predictor of likelihood of testing in both age groups. Table 1 shows that the majority of individuals ages 50–64 years had employer-based insurance with a USOC (65%). For individuals ages 50–64 years with insurance and a USOC, the type of insurance—privately purchased, Medicaid, or other public—was not predictive of testing compared with individuals who had employment-based insurance and a USOC. Uninsured individuals with a USOC, however, were much less likely to have received testing than individuals who had employer-based insurance with a USOC (RR, 0.61; 95% CI, 0.53–0.69).

Individuals ages 50–64 years with no USOC were less likely to be screened than individuals with a USOC and employment-based insurance. The magnitude of this effect was consistent across all other types of insurance (RR range, 0.52–0.54). Individuals who had no insurance and no USOC were far less likely to receive CRC testing (RR, 0.32; 95% CI, 0.23–0.43).

Among respondents age ≥ 65 years, the majority (71%) had a USOC and Medicare plus supplemental insurance coverage (other than Medicaid) (Table 1). Relative to these individuals, adults with other types of insurance—Medicare only or Medicare plus Medicaid—were less likely to be tested (RR range, 0.86–0.90) regardless of whether they had a USOC (Table 3).

Among younger adults, health status was not predictive of receipt of CRC testing. However, among adults age ≥ 65 years, individuals with fair or poor health status were less likely to be tested than individuals with good, very good, or excellent health status (RR, 0.92; 95% CI, 0.88–0.96). Individuals with a greater frequency of visits to a physician in the prior 12 months had a greater likelihood of being screened. This effect was greater among adults ages 50–64 years than among adults age ≥ 65 years.

Individuals age ≥ 65 years who were recent immigrants to the U.S. (0–50% of lifetime in the U.S.) were less likely to be tested than lifetime U.S. residents (RR, 0.86; 95% CI, 0.76–0.97). This effect was not present in the younger age group. Limited English proficiency was not a significant predictor of testing for either age group.

Reasons for Non-Use of CRC Testing

Respondents with no recent FOBT (in the past year) and respondents with no recent endoscopy (in the past 10 years) were asked the main reason why they did not have these tests. Among adults age ≥ 50 years who had no personal history of CRC, 17,695 responded to this question regarding FOBT, and 11,945 responded to this question regarding endoscopy.

Endoscopic testing

The most commonly reported reason respondents did not undergo endoscopic CRC testing was that a physician did not say an examination was needed (Table 4). Other commonly reported reasons were “no reason/never thought about it” or “haven't had any problems.” Two-thirds of respondents reported those three reasons.

Table 4. Reasons for Not Undergoing Colorectal Cancer Testing by Gender and by Race/Ethnicityab
Response/type examinationRow totalMaleFemaleWhiteLatinoAsianAfrican AmericanOther/multiracialAIAN
  • AIAN: American Indian/Alaska Native; FOBT: fecal occult blood test; Endoscopic: endoscopic examination.

  • c, a

    Data were obtained from the 2001 California Health Interview Survey.

  • b

    For gender, males and females were compared; for race/ethnicity, comparisons were made with the white reference group.

  • c

    P values < 0.05.

  • d

    The estimate was not reliable statistically because there were too few observations.

The physician did not tell me I needed it         
 Endoscopic26.121.629.1c28.022.1c21.1c24.524.823.8
 FOBT28.825.731.3c29.328.225.1c32.424.724.2
No reason         
 Endoscopic21.725.519.2c20.922.222.627.9c22.824.2
 FOBT21.823.720.3c21.423.122.422.323.920.6
Have not had problems         
 Endoscopic19.620.519.016.128.7c31.4c15.720.514.0
 FOBT14.515.313.912.118.2c27.1c14.014.617.4
Did not know a test was needed         
 Endoscopic7.77.38.07.66.89.68.17.48.1
 FOBT13.513.913.113.116.3c13.612.013.214.2
Painful/embarrassing         
 Endoscopic6.43.78.2c7.82.7c3.3c4.9cdd
 FOBT1.00.71.3c1.1ddddd
Put it off/laziness         
 Endoscopic4.14.63.84.73.8dddd
 FOBT3.52.84.1c4.11.4c1.6c4.24.9d
Expensive/no insurance         
 Endoscopic2.93.62.4c2.73.83.3ddd
 FOBT0.90.90.90.71.8cdddd
Had another type of examination         
 Endoscopic1.41.91.0c1.6ddddd
 FOBT4.95.44.4c5.91.7c2.6c3.4cdd
Do not have a physician         
 Endoscopic0.81.30.5c0.8ddddd
 FOBT0.60.60.60.6ddddd
Other         
 Endoscopic9.310.08.910.08.06.7c8.612.212.7
 FOBT10.611.010.311.87.6c5.4c10.012.714.5

Reported reasons for not having endoscopic testing differed by race and ethnicity. Asians and Latinos were significantly more likely than whites to report that they were not tested because of an absence of symptoms or perceived health problems (Asians, 31%; Latinos, 29%; whites, 16%; P < 0.001). The same groups were less likely than whites to report that they were not tested because the endoscopic examination was painful or embarrassing (Asians, 3%; Latinos, 3%; whites, 8%; P < 0.001) or because their physician did not tell them the test was needed (Asians, 21%; Latinos, 22%; whites, 28%; P < 0.001).

The reasons reported for not undergoing endoscopic examination also varied by gender. Women were more likely to say that their physician had not informed them they needed the examination (29% vs. 22%; P < 0.001). Women also were more than twice as likely as men to perceive endoscopic examinations as painful/embarrassing (8% vs. 4%; P < 0. 001), although this was not a commonly mentioned reason for either gender.

FOBT testing

The reasons identified for not having an FOBT were similar to the reasons for not undergoing endoscopic screening tests, with lack of physician recommendation the most common reason reported by all groups (Table 4). Asians (27%) and Latinos (18%) were significantly more likely than whites (12%; P < 0.001) to state an absence of symptoms as the main reason for not having an FOBT. The largest difference between men and women was in the percentage of respondents who reported that a physician did not tell them a test was needed, with women more likely to report this reason than men (31% and 26%, respectively; P < 0.001).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The rates of testing for colorectal cancer in the CHIS 2001 were higher compared with the rates seen in recent national health surveys. California, with 54% of its population age ≥ 50 years reporting a CRC test in 2001, appears to be ahead of national rates, which were 41% for men and 38% for women in 2000.7 Although the results of this analysis suggest relatively widespread use of CRC testing in California, two areas require further consideration. First, our results from CHIS were consistent with NHIS data in finding that women were less likely to undergo tests for CRC. One interpretation of this lower use is that women and their health care providers may perceive CRC as less important compared with men and their providers. Women also were more likely than men to say they had not had a recent CRC test, because their physician did not inform them the test was needed or because the tests were painful or embarrassing. The finding that women report using CRC tests less than men in California is somewhat surprising, because cervical and breast cancer screening tests historically have been used widely by women and recommended by their physicians.3, 28 One possible explanation for this finding is that some women use obstetrician/gynecologists as their primary care provider, and it has been shown that physicians in this specialty are much more likely than other primary care physicians to conduct FOBT by digital rectal examination.29 Digital rectal examination-based FOBT is a nonstandard approach that is discouraged in major guidelines and was not recognized as appropriate testing in the CHIS.

The widespread use of other cancer tests suggests that most women already employ cancer screening and readily may accept their physician's recommendation for CRC testing. Professional guidelines and quality-of-care measures are needed to specify appropriate CRC test use. Inclusion of CRC testing as a new Health Plan Employer Data and Information Set measure beginning in 2004 may encourage more providers to recommend CRC testing for their female patients. Further work is needed to understand better the reasons for lower CRC test use among women.

A second area of concern is the lower use of CRC tests by Latinos, and especially Latinas. Even after controlling for a broad range of demographic and socioeconomic factors, Latinos ages 50–64 years were less likely to be tested than whites. Disparities for Latinos did not appear to be the result of a language barrier: Respondents with limited proficiency in English did not have worse rates of testing in a comprehensive multivariate model. The reported reasons for not receiving recent CRC screening tests point to possible cultural differences in the perception of how early detection programs work. Latinos and Asians were twice as likely as whites to have stated an absence of symptoms or health problems as the reason for not being tested. With Latinos and Asians, outreach efforts should emphasize the fact that CRC screening is designed for individuals with no symptoms.

Our results need to be interpreted in the context of the limitations of the CHIS 2001. The survey was administered by telephone and, thus, may not be as representative of individuals in the lowest socioeconomic strata, who are less likely to have telephone services. However, the CHIS 2001 was weighted to minimize the effects of this characteristic of telephone surveys.30 Despite its limitations, the CHIS 2001 provides the first adequate sample sizes to develop estimates for Asians and American Indian/Alaska Natives as distinct populations. With future iterations of the CHIS, it will be possible to provide additional clarity in describing population-based uptake of CRC testing in California.

What barriers need to be overcome to achieve even greater acceptance for CRC examinations? Having both insurance coverage and a USOC was the most powerful predictor of whether an individual received a CRC test in our analysis. Among insured adults age 50–64 years, Medicaid beneficiaries with a USOC were nearly twice as likely to be tested as Medicaid beneficiaries without a USOC. This finding suggests two policy consequences for Medicaid: First, state policy makers should be aware that ensuring a regular source of care among Medicaid beneficiaries is a policy goal. The ability to maintain a regular source of care for Medicaid beneficiaries is driven largely by the continuity of Medicaid coverage.31, 32 California, as in most other states in times of fiscal distress, exercises budgetary controls by tightening eligibility rules for Medicaid.33 Requiring beneficiaries to undergo frequent recertification procedures to maintain program eligibility may lead to disruptions in coverage and loss of continuity of care.

Second, our results confirmed previous literature in finding insignificant differences in rates of test use between public and private coverage for younger individuals.34, 35 Policies that target increased screening among adults age < 65 years should focus on establishing continuous coverage to ensure a regular source of health care, with less attention to uncovering screening differentials between public and private coverage. Underinsurance, however, was a more important issue among the elderly with respect to CRC test use. Low-income elderly without supplemental insurance were less likely to be tested than individuals who were covered by Medicare with supplemental insurance.

The last decade has seen a remarkable increase both in the amount of evidence supporting the efficacy of CRC testing and in the acceptance of these examinations. At the time of CHIS 2001, over half of California residents age ≥ 50 years had undergone recent testing for CRC. The ongoing challenge remains to provide effective early detection programs to the greatest proportion of the population. Because CRC is preventable, California should be applauded for its achievements and encouraged to continue its efforts.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The authors greatly appreciate the support of Marion Standish, Program Director at The California Endowment; Dr. Ralph Coates, Associate Director for Science, Division of Cancer Prevention and Control, Centers for Disease Control and Prevention; and Dr. Rachel Ballard-Barbash, Associate Director, Applied Research Program, National Cancer Institute.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  • 1
    American Cancer Society and Public Health Institute, California Cancer Registry. California cancer facts and figures, 2001. Oakland: American Cancer Society, California Division, 2000.
  • 2
    Ries L, Eisner M, Kosary C, et al. SEER cancer statistics review, 1975– 2000. Available at: http://seer.cancer.gov/csr/1975_2000.
  • 3
    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: 17041713.
  • 4
    American Cancer Society. Cancer facts and figures, 2003. Atlanta: American Cancer Society, 2003.
  • 5
    U.S. Preventive Services Task Force. Screening for colorectal cancer: recommendation and rationale. Ann Intern Med. 2002; 137: 129131.
  • 6
    Centers for Disease Control and Prevention. Screening for colorectal cancer—United States, 1992–1993, and new guidelines. MMWR Morb Mortal Wkly Rep. 1996; 45: 107110.
  • 7
    Swan J, Breen N, Coates RJ, Rimer BK, Lee NC. Progress in cancer screening practices in the United States: results from the 2000 National Health Interview Survey. Cancer. 2003; 97: 15281540.
  • 8
    O'Malley A, Beaton E, Yabroff K, Abramson R, Mandelblatt J. Patient and provider barriers to colorectal cancer screening in the primary care safety-net. Prev Med. 2004; 39: 5663.
  • 9
    Beeker C, Kraft J, Southwell B, Jorgensen C. Colorectal cancer screening in older men and women: qualitative research findings and implications for intervention. J Commun Health. 2000; 25: 263278.
  • 10
    Holmes-Rovner M, Williams G, Hoppough S, Quillan L, Butler R, Given C. Colorectal cancer screening barriers in persons with low income. Cancer Pract. 2002; 10: 240247.
  • 11
    Seeff L, Nadel M, Klabunde C, et al. Patterns and predictors of colorectal cancer test use in the adult U.S. population. Cancer. 2004; 100: 20932103.
  • 12
    Greiner K, Engelman K, Hall M, Ellerbeck E. Barriers to colorectal cancer screening in rural primary care. Prev Med. 2004; 38: 269275.
  • 13
    Mandelson M, Curry S, Anderson L, et al. Colorectal cancer screening participation by older women. Am J Prev Med. 2000; 19: 149154.
  • 14
    Adams E, Thorpe K, Becker E, Joski P, Flome J. Colorectal cancer screening, 1997–1999: role of income, insurance and policy. Prev Med. 2004; 38: 551557.
  • 15
    Centers for Disease Control and Prevention. Trends in screening for colorectal cancer—United States, 1997 and 1999. MMWR Morb Mortal Wkly Rep. 2001; 50: 162166.
  • 16
    Ponce NA, Lavarreda SA, Yen W, Brown ER, DiSogra C, Satter D. Translation methodology in the California Health Interview Survey: challenges in language translation for a major survey. Public Health Rep. 2004; 119: 118.
  • 17
    California Health Interview Survey. CHIS 2001 methodology series: Report 1—sample design. Los Angeles: UCLA Center for Health Policy Research, 2002.
  • 18
    California Health Interview Survey. CHIS 2001 methodology series: Report 2—data collection methods. Los Angeles: UCLA Center for Health Policy Research, 2002.
  • 19
    California Health Interview Survey. CHIS 2001 methodology series: Report 3—data processing procedures. Los Angeles: UCLA Center for Health Policy Research, 2002.
  • 20
    California Health Interview Survey. CHIS 2001 methodology series: Report 4—response rates. Los Angeles: UCLA Center for Health Policy Research, 2002.
  • 21
    Andersen RM, Yu H, Wyn R, Davidson PL, Brown ER, Teleki S. Access to medical care for low-income persons: how do communities make a difference? Med Care Res Rev. 2002; 59: 384411.
  • 22
    Seeff LC, Shapiro JA, Nadel MR. Are we doing enough to screen for colorectal cancer? Findings from the 1999 Behavioral Risk Factor Surveillance System. J Fam Pract. 2002; 51: 761766.
  • 23
    Shi R, Berkel H, Sartor O. Comparison of utilization of preventive health care services between two racial populations. Ann Epidemiol. 2000; 10: 454.
  • 24
    Corbie-Smith G, Flagg EW, Doyle JP, O'Brien MA. Influence of usual source of care on differences by race/ethnicity in receipt of preventive services. J Gen Intern Med. 2002; 17: 458464.
  • 25
    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: 257270.
  • 26
    Maxwell AE, Bastani R, Warda US. Demographic predictors of cancer screening among Filipino and Korean immigrants in the United States. Am J Prev Med. 2000; 18: 6268.
  • 27
    Zhang J, Yu KF. What's the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA. 1998; 280: 16901691.
  • 28
    Augustson EM, Vadaparampil ST, Paltoo DN, Kidd LR, O'Malley AS. Association between CBE, FOBT, and Pap smear adherence and mammography adherence among older low-income women. Prev Med. 2003; 36: 734739.
  • 29
    Klabunde CN, Frame PS, Meadow A, Jones E, Nadel M, Vernon SW. A national survey of primary care physicians' colorectal cancer screening recommendations and practices. Prev Med. 2003; 36: 352362.
  • 30
    Centers for Disease Control and Prevention. Colorectal cancer prevention and control initiatives. Available from URL: http://www.cdc.gov/cancer/colorctl/ [accessed October 16, 2003].
  • 31
    California Health Interview Survey. CHIS 2001 methodology series: Report 5—weighting and variance estimation. Los Angeles: UCLA Center for Health Policy Research, 2002.
  • 32
    Newacheck PW, Pearl M, Hughes DC, Halfon N. The role of Medicaid in ensuring children's access to care. JAMA. 1998; 280: 17891793.
  • 33
    Kasper JD, Giovannini TA, Hoffman C. Gaining and losing health insurance: strengthening the evidence for effects on access to care and health outcomes. Med Care Res Rev. 2000; 57: 298318; discussion, 319–225.
  • 34
    Kaiser Family Foundation. News on Medicaid and state budgets: July and August 2003 update. Available from URL: http://www.kff.org/medicaid/kcmu4140report.cfm [accessed August 16, 2003].
  • 35
    Chimento L, Forbes M, Menges J, Theisen A, Pande N. Simplifying Medi-Cal enrollment: opportunities and challenges in tight fiscal times. Medi-Cal Policy Inst Issue Brief. 2003; 6: 112.