Abstract
Conflicting information currently exists about the role played by health-related quality of life (HRQOL) in influencing colorectal cancer screening. The current study aimed to determine the relationship between HRQOL and colorectal cancer screening, using nationally representative public data from the Behavioral Risk Factor Surveillance System (BRFSS). The 2010 BRFSS data were used for this study. Individuals younger than age 50 years were excluded from the study. Missing data were imputed using the multiple imputation technique. Multiple multivariate logistic regression models were fitted to the data to determine the association between different components of HRQOL (physical HRQOL, mental HRQOL, activity limitation caused by poor mental or physical HRQOL, and general health status) and receipt of colorectal cancer screening tests (fecal occult blood testing [FOBT] in the past year, sigmoidoscopy in the past 5 years, and colonoscopy in the past 10 years). The study sample comprised 301,488 individuals. Approximately 12% of the respondents had received FOBT in the past year, 62.6% had received sigmoidoscopy in the past 5 years, or colonoscopy in the past 10 years, and 65.4% had received either of the screening tests within appropriate time frames. After controlling for demographic and health-related covariates, an inverse relationship was observed between HRQOL and colorectal cancer screening with the exception of mental HRQOL and FOBT. The relationship between mental HRQOL and FOBT was found to be nonsignificant. Policy makers should consider including HRQOL as an important parameter when designing interventions aimed at improving colorectal cancer screening rates. (Population Health Management 2013;16:178–189)
Introduction
Screening is an effective way to alleviate the morbidity and mortality associated with colorectal cancer. The aim of screening is to detect the disease at an early stage, when it is asymptomatic, preventable, and curable, thereby improving the probability of better health outcomes by providing proper and timely treatment. 6 Several studies have shown that early screening reduces mortality associated with colorectal cancer 7 –12 ; as many as 90% of deaths associated with colorectal cancer could be prevented through early screening. 13 Given the benefits of early screening, it is not surprising that studies have found colorectal cancer screening techniques to be cost effective. 13,14 Colorectal cancer screening has been recommended by all the leading cancer organizations. 15 The United States Preventive Services Task Force and the American Cancer Society (ACS) recommend 3 types of colorectal cancer screening tests starting at age 50 years: (1) high-sensitivity fecal occult blood testing (FOBT) every year, (2) sigmoidoscopy every 5 years, and (3) colonoscopy every 10 years. 16
Over the past decade, the national colorectal cancer screening rate among individuals aged 50 years or older has increased from 53% in 2002 to 64% in 2008. 17 However, despite these improvements, the current screening rate falls short of the 75% goal set by the ACS for the year 2015. 18 In order to achieve this goal, the critical first step is to ascertain the factors that influence screening. Several studies have shown that race, age, income, educational status, body mass index (BMI), health insurance coverage, screening awareness, smoking, and alcohol use affect colorectal cancer screening compliance. 15,19 –24 Though these studies provide useful information about factors influencing colorectal cancer screening, limited information is currently available about the role played by health-related quality of life (HRQOL). HRQOL is a multidimensional construct pertaining to the physical, emotional, mental, social, and behavioral components of well-being and function as perceived by patients and/or observers. 25 Self-reported HRQOL has been found to be a more powerful predictor of mortality and morbidity than objective measures of health. 26,27 A few studies have assessed the role played by HRQOL in influencing colorectal cancer screening; however, the results of these studies have been inconsistent. Although some studies have found a positive association between HRQOL and colorectal cancer screening, 15,19,27 others have reported a negative association 24,29 –33 ; some studies have found no association between HRQOL and colorectal cancer screening. 34,35 One plausible reason for the lack of consistent findings among these studies could be that most of them measured HRQOL using a single item—“perceived general health status”—and did not consider the core aspects of HRQOL (ie, the physical and mental components of HRQOL).
The current study aimed to determine the relationship between HRQOL and colorectal cancer screening using nationally representative public data from the Behavioral Risk Factor Surveillance System (BRFSS). The primary objective of the study was to determine the association between different components of HRQOL (physical HRQOL, mental HRQOL, activity limitation caused by poor mental or physical HRQOL, and general health status) and the receipt of colorectal cancer screening tests (FOBT, sigmoidoscopy, or colonoscopy) within the recommended time frames.
Methods
Study design and data
This study utilized a cross-sectional study design. The 2010 BRFSS data were used for the purpose of the study. The BRFSS is a federally funded annual telephonic survey conducted in each of the 50 states and Washington, DC by the respective state health departments in association with the Centers for Disease Control and Prevention. 36 The survey asks questions about health care access, risky health behaviors, and preventive health practices related to chronic diseases and injuries. 36 –38 Information is collected in each state from a random probability sample of noninstitutionalized adults aged 18 years or older, using computer assisted telephonic interviewing techniques. For the study analyses, the authors included only those respondents who were aged ≥50 years because colorectal cancer screening has been recommended for people of this age group. 16 Each observation in the BRFSS survey is weighted to enable generalization of the findings to the national population. Further details about the survey design and methodology have been provided elsewhere. 39 Because BRFSS is a publically available de-identified data set, this study was exempt from review by the institutional review board at the University of Mississippi.
Study variables
The BRFSS survey contains 5 items related to colorectal cancer screening. The first item is “A blood stool test is a test that may use a special kit at home to determine whether the stool contains blood. Have you ever had this test using a home kit?” Those who respond “yes” to this item are then asked “How long has it been since you had your last blood stool test using a home kit?” The response to this question ranges from within the past year to more than 5 years ago. The third item asks respondents whether they have ever had sigmoidoscopy or colonoscopy. Those who respond positively to this item are then asked about the time since the last screening test, and whether the screening test was sigmoidoscopy or colonoscopy. The main dependent variables in the study were receipt of FOBT within the past year (FOBT), receipt of sigmoidoscopy within the past 5 years/colonoscopy within the past 10 years (endoscopy), and receipt of FOBT, sigmoidoscopy, or colonoscopy within the appropriate time frame (FOBT/endoscopy). These variables were coded dichotomously for analysis.
The 4 HRQOL items in the BRFSS survey are: (1) “Would you say that in general your health is excellent, very good, good, fair, or poor?” (general health status); (2) “Now thinking about your physical health, which includes physical illness and injury, for how many days during the past 30 days was your physical health not good?” (physical health); (3) “Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good?” (mental health); and (4) “During the past 30 days, for about how many days did poor physical or mental health keep you from doing your usual activities, such as self-care, work, or recreation?” (activity limitation). These HRQOL measures used in BRFSS have been derived from the original version of the Medical Outcomes Study 36-item Short Form Survey (SF-36) instrument, which is widely considered to be the gold standard for HRQOL measurement. 40 These measures also have been validated in both healthy and disabled populations with acceptable criterion validity and reliability on par with multiple-item SF-36 subscales. 41 –45 As is common among the studies using the 4 BRFSS HRQOL items, the response options were modified in accordance with the current study needs for these items. General health status was dichotomized into good health (containing responses of excellent, very good, and good health) and poor health (containing responses of fair and poor health). 29,43 Physical health, mental health, and activity limitation were dichotomized into less than 14 days and 14 days or more. The 14-day cutoff point has been used regularly as a clinical indicator in medical research. 43,46,47 These modified HRQOL items were the main independent variables in this study.
The covariates included in the study were age, race, sex, income, BMI, marital status, health insurance status, education, smoking status, alcohol use, physical activity, geographic region, employment status, and interval since last routine checkup. Age was categorized as 50 to 59 years, 60 to 69 years, 70 to 79 years, and 80 and older. Race was divided into 5 categories: non-Hispanic whites, non-Hispanic blacks, Hispanic, non-Hispanic multiracial, and non-Hispanic others. Sex was classified as male and female. Income consisted of 5 categories: <$10,000, $10,000–$19,999, $20,000–$34,999, $35,000–$74,999, and ≥$75,000. BMI included 3 categories: neither overweight nor obese (BMI<25), overweight (25≤BMI<30), and obese (BMI≥30). Health insurance status was coded as “yes” or “no” depending on whether the respondent had health insurance. Marital status was divided into 4 categories: married/unmarried couple, divorced or separated, widowed, and never married. Education level included 4 categories: less than high school, graduated high school, attended college or some technical school, and college graduate. Smoking status had 4 categories: never smoked, former smoker, current smoker and smokes sometimes, and current smoker and smokes every day. Drinking status was coded as “yes” or “no” depending on whether the respondent had drinks within the past 30 days. Similarly, physical activity was coded as “yes” or “no” depending on whether the respondent indulged into some physical activity within the past 30 days. Geographic region was divided into 4 categories: northeast, midwest, south, and west. Employment status included 5 categories: employed or self-employed, out of work, student or homemaker, retired, and unable to work. Interval since the last routine checkup included 3 categories: less than 1 year, more than 1 year, and never had any checkup. For each study variable, responses labeled as “don't know/refused” in the original survey were considered to be missing.
Statistical analysis
Bivariate analyses were conducted using chi-square tests. To determine the association between HRQOL and colorectal cancer screening, multiple multivariate logistic regression models were fitted to the data. Considering that the 4 modified HRQOL items were the main independent variables of interest and there were 3 main dependent variables (FOBT, endoscopy, FOBT/endoscopy), a total of 12 logistic regression models were fitted in the study. This approach is consistent with that used by Masterson et al in their study of association between mental health status and recent mammography screening. 48 The covariates listed were adjusted for in each model. All analyses were conducted using SAS version 9.2 (SAS Institute Inc., Cary, NC). Logistic regression procedures conducted in SAS follow a list-wise deletion pattern, wherein observations with missing values on the dependent variable or any of the independent variables are excluded from the analysis. 49 If the excluded observations are systematically different than the retained observations, the results may be biased. 50 In order to avoid bias, the missing values were imputed using the Multiple Imputation (MI) technique, which has been used widely to deal with the problem of missing data in survey research. 51,52 The variables physically unhealthy days, mentally unhealthy days, days with activity limitation caused by poor physical and mental health, weight, and height were found to have skewed distributions, and hence were fifth-root transformed to near normality before they were included in the MI procedure. Considering the complex sampling design of BRFSS data, PROC SURVEY procedures in SAS were used for analysis. PROC SURVEYMEANS was used to compare frequencies and percentages for the categorical variables. PROC SURVEYLOGISTIC was used to conduct logistic regressions. PROC MI and PROC MIANALYZE were used for missing data adjustment using the MI technique. All results reported are weighted unless noted otherwise.
Results
The BRFSS data were available for 401,375 individuals in year 2010. Of these, 301,488 (75.1%) were 50 years of age or older and were considered for the analysis. Table 1 describes the sociodemographic characteristics of the sample and results from bivariate analysis. Roughly 12% of the respondents had received FOBT in the past year, 62.6% had received endoscopy (sigmoidoscopy/colonoscopy), and 65.4% had received FOBT/endoscopy within the appropriate time frames.
Receipt of FOBT within the past year.
Receipt of sigmoidoscopy within the past 5 years or colonoscopy within the past 10 years.
Receipt of FOBT within the past year, sigmoidoscopy within the past 5 years, or colonoscopy within the past 10 years.
Unweighted results.
Others include Asians, native Hawaiians, other Pacific Islanders, American Indians, or Alaska natives.
FOBT, fecal occult blood testing; Wt, weighted; BMI, body mass index.
Table 2 presents the results of the multivariate logistic regression analyses conducted to determine the association between the receipt of FOBT screening and HRQOL. Individuals who reported less than 14 physically unhealthy days were 13% less likely to receive FOBT as compared to individuals who reported 14 or more physically unhealthy days (Model 1). The association between mental HRQOL and receipt of FOBT was found to be statistically insignificant (Model 2). Individuals who reported less than 14 days of activity limitation due to poor physical or mental health were 10% less likely to receive FOBT as compared to individuals who reported 14 or more days of activity limitation (Model 3). Individuals with good general health status were 5% less likely to receive FOBT as compared to individuals with poor general health status (Model 4).
Others include Asians, native Hawaiians, other Pacific Islanders, American Indians, or Alaska natives.
BMI, body mass index.
Table 3 presents the results of the multivariate logistic regression analyses conducted to determine the association between endoscopy (sigmoidoscopy/colonoscopy) and HRQOL. Individuals who reported less than 14 physically unhealthy days were 16% less likely to receive an endoscopy as compared to individuals who reported 14 or more physically unhealthy days (Model 5). The estimated odds of receiving an endoscopy were 12% lower in individuals who reported less than 14 mentally unhealthy days as compared to individuals who reported 14 or more mentally unhealthy days (Model 6). Individuals who reported less than 14 days of activity limitation caused by poor physical or mental health were 17% less likely to receive an endoscopy as compared to individuals who reported 14 or more days of activity limitation (Model 7). Individuals with good general health status were 11% less likely to receive an endoscopy as compared to individuals with poor general health status (Model 8).
Receipt of endoscopy is defined as receipt of sigmoidoscopy within the past 5 years or colonoscopy within the past 10 years.
Others include Asians, native Hawaiians, other Pacific Islanders, American Indians, or Alaska natives.
BMI, body mass index.
Table 4 presents the results of the multivariate logistic regression analyses conducted to determine the association between FOBT/endoscopy and HRQOL. Individuals who reported less than 14 physically unhealthy days were 16% less likely to receive FOBT/endoscopy as compared to individuals who reported 14 or more physically unhealthy days (Model 9). Individuals who reported less than 14 mentally unhealthy days were 11% less likely to receive FOBT/endoscopy as compared to individuals who reported 14 or more mentally unhealthy days (Model 10). The estimated odds of receiving FOBT/endoscopy were 17% lower in individuals who reported less than 14 days with activity limitation as compared to individuals who reported 14 or more days with activity limitation (Model 11). Individuals with good general health status were 10% less likely to receive FOBT/endoscopy as compared to individuals with poor general health status (Model 12).
Receipt of fecal occult blood testing (FOBT)/endoscopy is defined as receipt of FOBT in the past year, sigmoidoscopy in the past 5 years, or colonoscopy in the past 10 years.
Others include Asians, native Hawaiians, other Pacific Islanders, American Indians, or Alaska natives.
BMI, body mass index.
Discussion
Screening for colorectal cancer has been found to alleviate the morbidity and mortality associated with the disease. Despite these benefits, almost two fifths of individuals aged 50 years or older do not undergo colorectal cancer screening as recommended. Understanding factors associated with compliance with colorectal cancer screening recommendations is the critical first step in designing strategies aimed at increasing the screening rates. The current study contributes to the literature concerning the factors predicting colorectal cancer screening by evaluating the role of HRQOL in this regard. A nationally representative database was used to determine the association between the core components of HRQOL and colorectal cancer screening.
A majority of earlier studies in this area have used a single-item HRQOL measure (general health status) to study the association with colorectal cancer screening. However, in this study, a complete measure of HRQOL (physical, mental, activity limitation caused by poor physical and mental health, and general health status) was used to evaluate the relationship with colorectal cancer screening. In a seminal article, Hennessy et al discussed the rationale for using the 4 separate domains of HRQOL in public health surveillance. 42 Physical and mental aspects of HRQOL need to be evaluated separately because the general health status component could confound the distinct effect of physical and mental HRQOL. 42,53 Inclusion of general health status as the only parameter to evaluate HRQOL may not work effectively when an individual experiences excellent physical health but poor mental health or vice versa. In addition, several individuals interpret “general health” as physical health, which could further confound results in the absence of distinct dimensions of HRQOL. 42 Similarly, separate evaluation for “activity limitation due to poor physical or mental HRQOL” is necessary to communicate information that is not conveyed by the single general health status item. When analyzing the 1993 BRFSS data from 6 states, Hennessy et al found that 45% of the respondents who reported “excellent” or “very good” health reported 1 or more activity limitation days, whereas 24% of the respondents who reported fair or poor health had listed no activity limitation days. 42 To the best of the authors' knowledge, the current study is the first to examine the relationship between colorectal cancer screening and the 4 distinct dimensions of HRQOL. Given the unique information captured by each of the 4 HRQOL measures used in BRFSS, it is highly recommended that future studies include all 4 aspects when examining the relationship with cancer screening.
Significant results emerged from this examination of the relationship between colorectal cancer screening and HRQOL dimensions in the presence of study confounders. With the exception of FOBT and mental HRQOL, an inverse relationship was observed between colorectal cancer screening and HRQOL. The relationship between FOBT and mental HRQOL was found to be nonsignificant. In their study of the relationship between colorectal cancer screening and self-reported health status among patients enrolled in the Veteran Affairs system, Sultan et al found higher screening rates among older individuals with poor health status. 54 However, in a similar study, Wong et al found no relationship between colorectal cancer screening and physical and mental HRQOL among Asian Americans and non-Latino whites in the state of California. 35 It is difficult to compare the results of this study to previous studies in this area because of differences in patient population, study methodology, and differential use of HRQOL measure. The authors feel that by using a nationally representative sample and holistic HRQOL measure, this study provides a better perspective on the relationship between colorectal cancer screening and HRQOL as compared to previous studies. Regular contact with the health care provider and greater health concern may explain the higher screening rate observed among individuals with poor HRQOL. 24,55,56 Further, it is likely that individuals with poor HRQOL are subjected to intrusive exams on a regular basis, which may make them less averse to undergoing screening tests such as sigmoidoscopy and colonoscopy. It will be interesting to see if these results hold true across other cancer screening settings.
The findings of this study have important implications for health care policy makers. To increase the national colorectal cancer screening rate, it is necessary to target interventions toward individuals aged 50 or older with high HRQOL. It is likely that individuals with high HRQOL perceive themselves to be less susceptible to colorectal cancer and are less aware of the benefits of colorectal cancer screening. Hence, public health awareness campaigns or personalized interventions such as mailed brochures, which emphasize the susceptibility of these individuals to contract colorectal cancer, the benefits of colorectal cancer screening, and the serious consequences of contracting colorectal cancer, are needed for these individuals. Future studies should evaluate the effectiveness of public health programs in improving colorectal cancer screening among individuals with high HRQOL.
This study has a few limitations. Given the cross-sectional design of the study, it is not possible to infer a causal relationship between colorectal cancer screening and HRQOL. The BRFSS survey data are based on self-report, which makes them vulnerable to recall bias. Coding errors could have occurred during data entry and processing. The authors were not able to identify whether the endoscopic procedures conducted were for screening or diagnostic purposes. The results of this study must be interpreted in light of these limitations.
The current study provides useful insights into the relationship between colorectal cancer screening and HRQOL using nationally representative data. The results from regression analyses showed higher screening rates among individuals with lower HRQOL. This relationship held true across different colorectal cancer screening techniques and HRQOL dimensions. When designing interventions aimed at increasing colorectal cancer screening rates, policy makers also should account for factors related to an individual's HRQOL.
Footnotes
Author Disclosure Statement
Mr. Mahabaleshwarkar and Drs. Khanna, West-Strum, and Yang received no funding and declare no conflicts of interest with respect to the research, authorship, and/or publication of this article.