Abstract
We investigated the associations of perceived risk and cancer worry with colorectal cancer screening by the faecal occult blood test, colonoscopy or both. This study was based on the 2013 Korean National Cancer Screening Survey, including 2154 randomly selected, cancer-free and over 50-year-old adults. Individuals with higher cancer worry were 1.53 times more likely to undergo colorectal cancer screening, influenced by emotional reaction; individuals with greater perceived risk were 1.61 times more, affected by subjective awareness. However, cancer worry was only associated with the faecal occult blood test. Better understanding of cancer worry and perceived risk on screening behaviours may help to increase colorectal cancer screening rates.
Introduction
Colorectal cancer (CRC) now ranks worldwide as the third most common cancer in men (10.0% of all cancer patients) and the second in women (9.2%) (Ferlay et al., 2012). Although the incidence of CRC has declined or stabilised in many countries, nearly 55 per cent of all CRC cases occur in developed countries (Center et al., 2009; Jemal et al., 2011); it is a major problem in the developed countries of Asia (Center et al., 2009). In Korea, CRC incidence rates are rapidly increasing, with annual percent changes of 5.7 per cent in men and 4.3 per cent in women between 1999 and 2012 (Jung et al., 2015).
CRC screening in average-risk asymptomatic populations has been shown to reduce both incidence and mortality and is therefore widely recommended. The US Preventive Services Task Force recommends screening for CRC using the faecal occult blood test (FOBT), sigmoidoscopy or colonoscopy in adults, beginning at age 50 years and continuing until age 75 years (Lau and Kirby, 2009). The FOBT is simple, safe, inexpensive and known to reduce CRC mortality, although it has also been shown to be less accurate than colonoscopy (Mandel et al., 2000). Meanwhile, colonoscopy allows for direct inspection of the entire bowel and definitive treatment by polypectomy. However, rare complications there from may elicit serious side effects (Levin et al., 2008).
In Korea, the National Health Insurance Service (NHIS) manages mandatory social health insurance for the entire population. Under the NHIS, the National Cancer Screening program (NCSP) was introduced in 1999, and as part of the NCSP, a CRC screening programme was launched in 2004 (Shim et al., 2010). Individuals older than 50 years are eligible for CRC screening annually via the FOBT. Only upon detection of an abnormal finding colonoscopy is provided under the NCSP. Through the screening programme, the Medical Aids Program (MAP) recipients and the NHIS beneficiaries in the lowest 50 per cent income stratum are eligible for CRC screening free of charge. The remaining NHIS beneficiaries are also eligible for screening with a co-payment of 10 per cent of the cost of the procedure. CRC screening is often conducted at outpatient clinics or private health-assessment centres for opportunistic screening, especially through colonoscopy.
Although screening has been shown to reduce the incidence and mortality of CRC, uptake of CRC screening has remained low in comparison to rates for other cancers (Lemon et al., 2001; Vernon et al., 1997). In Korea, the overall CRC screening rate for both organised and opportunistic screening was only 55.6 per cent in 2013, the lowest among other cancers. As well, the FOBT screening rate (27.6%) was lower than that of colonoscopy screening (35.2%) (Suh et al., 2015).
Many studies have shown social factors, such as age, socioeconomic status and education, to be associated with screening behaviours. However, little is known about the effects of perceived risk and cancer worry on CRC screening uptake. In the literature, cancer worry is defined as an emotional reaction to the threat of cancer (Bowen et al., 2003). Although some studies have reported contradictory results, many suggest that higher levels of cancer worry are associated with increased odds of undergoing cancer screening (Choi et al., 2015). Perceived risk, meanwhile, is explained as one’s subjective belief about their own risk of developing a particular disease (Glanz et al., 2008; Hay et al., 2005; Katapodi et al., 2004) and has been found to play an important role in promoting healthy behaviours (McQueen et al., 2008). Therein, greater perceived risk stimulated feelings of vulnerability, thereby promoting preventive health behaviour to reduce the risk of getting a disease (Robb et al., 2004). Both cancer worry and perceived risk of getting cancer are considered as facilitators of screening uptake. However, few studies have compared perceived risk (cognitive conception) to worry (affective conception), and there is some debate as to their influence on screening behaviours across cancer types. McCaul et al. (1996b) found that both risk and worry judgements were independent predictors of mammography screening. Also, Moser et al. (2007) reported that risk and worry (but not their interaction) were associated with regular mammography and having had a sigmoidoscopy or colonoscopy, but not with the FOBT or prostate-specific antigen screening. Meanwhile, however, Cameron and Diefenbach (2001) suggested that worry, but not perceived risk, predicted interest in undergoing genetic screening for cancer. Most recently, a study reported that, although neither absolute risk perceptions nor comparative risk perceptions exerted a direct impact on women’s compliance with recommendations to undergo cervical cancer screening, both types of risk perception had an indirect effect on cervical cancer screening that was mediated by cancer worry (Zhao and Nan, 2016).
In sum, screening rates for CRC in Korea are lower than those for other types of cancers, and studies have yet to outline factors, especially psychosocial factors, influencing adherence to CRC screening and individual screening modalities. Therefore, better understanding of the influence of cancer worry and perceived risk of getting cancer on screening behaviours, especially for specific screening tests, may help with developing patient-centred interventions with which to efficiently increase CRC screening rates. Thus, in this study, we aimed to identify moderators of cancer worry and perceived risk for developing CRC among several known risk factors. Then, we set out to examine the influence of cancer worry and perceived risk on CRC screening uptake and to investigate associations between these two indicators and CRC screening via the FOBT and colonoscopy.
Methods
Participants and procedure
Data were acquired from subjects included in the 2013 Korean National Cancer Screening Survey (KNCSS). The KNCSS is an annual cross-sectional survey performed to investigate screening rates among Koreans for five common cancers (gastric, liver, colorectal, breast and cervix) through nationally representative random sampling. Men aged ⩾40 years and women ⩾30 years were selected based on the Resident of Registration Population data for July 2013 compiled by Statistics Korea, using multistage random sampling according to sex, age groups and sizes of the populations in each district. Sampling error was allowed up to ±2.2 per cent (95% confidence interval (CI)). The survey was conducted from 26 September to 18 October 2013, at which time investigators from a professional research agency went door-to-door to recruit residents. At least three attempts were made to contact individual residents. The 2013 KNCSS included 4100 participants. The response rate after making contact was 69.3 per cent. Of the respondents, 2154 cancer-free adults aged ⩾50 years were finally included in this study, since the NCSP only provides CRC screening for individuals aged ⩾50 years. Informed consent was obtained from all study participants, and this study was approved by the Institutional Review Board of the National Cancer Center, Korea.
Measures
To assess utilisation rates of CRC screening according to screening modality (FOBT or endoscopy), we administered a brief survey. Questions were prefaced with a description of two to three sentences in length about the screening tests to help respondents differentiate between the two tests. Descriptions of the following distinguishing characteristics of the tests were provided: (1) the FOBT is a test used to discover blood in a stool sample. This test can be administered at home using a special kit or stool samples stored in a container can be sent to a medical centre; (2) endoscopy (e.g. sigmoidoscopy and colonoscopy) is performed with a flexible tube passed through the anus. We then assessed the respondent’s previous experiences with CRC screening by asking them (1) whether they had ever been screened for CRC, (2) when they underwent their most recent screening and (3) which tests they underwent during their most recent CRC screening (FOBT or endoscopy). A respondent was considered to be ‘up-to-date’ if he or she had undergone an FOBT test within the past 1 year or an endoscopy test within the past 10 years.
To measure perceived risk for developing CRC, each participant was asked the following question: ‘Compared with the average person of your age, what do you think your chances are of developing colorectal cancer?’ The question was derived from an article on developing instruments to measure psychological factors affecting CRC screening (Vernon et al., 1997). Participants were asked to report their perceived risk on a five-point scale, with higher scores reflecting greater perceived risk. Although examining perceived risk with a total scale of five points is better for detecting effects in analysis, the number of responses for the lowest or the highest level of cancer worry were too small to be directly used for analysis. Scores were then divided into three levels of perceived risk: low (score of 1–2), middle (score of 3) and high (score of 4–5).
Cancer worry was measured via a modified Lerman’s Cancer Worry Scale designed to reflect cancer worry in Korea (Lerman et al., 1991a, 1991b). Overall, cancer worry was assessed via two questions: (1) ‘How often have you worried about your own chances of developing cancer?’ and (2) ‘How often have you thought about your own chances of developing cancer?’ The worry items used in this study refer to frequency (i.e. ‘How often do you worry?’) rather than magnitude (i.e. ‘How much do you worry?’). For each question, participants were asked to respond via the five following scaled response items: (1) never, (2) rarely, (3) sometimes, (4) often and (5) always. Cancer worry was calculated as the total sum of the scores for both questions divided by two, with possible scores ranging from 1 to 5. Scores were then divided into three levels of cancer worry: low (score of 1–2), middle (score of 3) and high (score of 4–5).
We also assessed several socioeconomic (age, marital status, level of education and monthly household income) and health-related factors (private cancer insurance, family history of any cancer, alcohol consumption, cigarette smoking and subjective overall health) to evaluate their influence on screening behaviour. Individuals were asked whether they had joined private cancer insurance: private cancer insurance in Korea refers to supplementary health insurance that provides a fixed amount of benefits for patients diagnosed with cancer. We also questioned, ‘Have you or any member of your family ever had genetic testing to assess yours or their hereditary risk of cancer? If yes, then indicate the family relationship’. Subjective overall health status was measured by asking ‘Would you say that, in general, your health is …’ Respondents were to answer as very good, good, fair, poor and very poor. Furthermore, the amount of lifetime smoking, as well as the frequency of alcohol drinking, was recorded (Centers for Disease Control and Prevention (CDC), 2009).
Statistical analysis
Descriptive statistics were assessed to characterise the study population across CRC screening experiences. In the analysis, we indicated subjects who underwent CRC screening with the FOBT within 1 year or with colonoscopy within the past 10 years as the ‘screened group’. The ‘unscreened group’ included respondents who had never been screened or who had been screened before the ‘up-to-date’ criteria. We did not give any weight to the data because the samples were designed to be representative of the entire Korean population demographically. Chi-square tests were used to compare the distributions of risk factors according to the levels of cancer worry and perceived risk for developing CRC. Multivariate logistic regression and multinomial logistic regression analysis were conducted to identify associations between cancer worry and perceived risk for developing CRC, as well as up-to-date screening for CRC (screened vs not screened) by screening method. All variables with a p-value of <0.05 in univariate analysis were included in the logistic regression models as potential predictors. These variables included gender, age, private cancer insurance, subjective overall health, cancer worry and perceived risk of CRC. In the multinomial logistic analyses, we assessed the adjusted odds ratios (aORs) of uptake of the FOBT or colonoscopy, as well as uptake of both methods (FOBT and colonoscopy), in comparison to the not up-to-date screened group. Additionally, we conducted trend analysis for the ordinal categorical variables to highlight the significance of the trend. Finally, we applied a heterogeneity test of odds ratios for each independent variable across screening modalities. Statistical analyses were performed using STATA software (version 13; StataCorp LP, College Station, TX), and all p-values <0.05 were considered statistically significant.
Results
Descriptive data for the subjects according to their experiences with screening are presented in Table 1. Of the 2154 respondents, 67.0 per cent underwent CRC screening: 29.7 per cent were tested by FOBT, 18.5 per cent were tested by colonoscopy and 18.8 per cent were tested by both FOBT and colonoscopy. The CRC screening rates differed significantly according to cancer private insurance, subjective health status, cancer worry and perceived risk of developing CRC (p < 0.01). Those who had cancer private insurance, those who often worried about getting cancer and those who had a higher perceived risk for getting CRC were more likely to have undergone CRC screening in the past. Those who considered their subjective overall health to be good were less likely to have undergone CRC screening. Additionally, significant differences were noted in CRC screening rates for each screening method according to age, subjective health status, cancer worry and perceived risk of developing CRC (Table 1).
Demographic and health-related characteristics according to up-to-date colorectal cancer screening uptake.
FOBT: faecal occult blood test.
The difference between non-screened and screened individuals for each categorical variable.
The difference between the non-screened, FOBT only, colonoscopy only and FOBT + colonoscopy screened groups for each categorical variable.
Table 2 shows the levels of cancer worry and perceived risk for developing CRC according to known risk factors for CRC. Among the respondents, 29.7 per cent were often or always worried about developing cancer, 53.4 per cent were sometimes and 16.9 per cent were rarely or never worried about developing cancer. With statistical significance, those who had private cancer insurance and those who had family history of cancer were more frequently worried about developing cancer, while those who considered their subjective overall health to be good were rarely worried about cancer. Regarding perceived risk, 17.8 per cent of all participants thought their chances of developing CRC were higher than those for other men and women of the same age, 51.7 per cent thought their risk was the same and 30.5 per cent thought their risk was lower. Male individuals who more frequently consumed alcohol and who smoked more than five packs in their lifetime reported a higher perceived risk for developing CRC.
Univariate analyses of perceived risk and cancer worry for colorectal cancer (n = 2154; Korea, 2013).
Table 3 lists the adjusted association between the individual characteristics of the study population and the uptake of CRC screening. Those who had private cancer insurance were 1.35 times more likely to be screened than those who did not. Also, individuals who often or sometimes worried about getting cancer were, respectively, 1.53 times and 1.36 times more likely to undergo screening than those who rarely worried about getting cancer (significance of testing linear trend < 0.001). Subjects at greater perceived risk for developing CRC were also significantly more likely to be screened than those with lower perceived risk (significance of testing linear trend < 0.001).
Logistic regression for up-to-date colorectal cancer screening in Korea, 2013.
aOR: adjusted odds ratio; CI: confidence interval.
Testing for linear trend (p-value).
The factors most associated with CRC screening uptake were explored using a multinomial logistic regression model (Table 4). Analysed via multinomial logistic regression using non-screened individuals as the reference category, males with higher levels of cancer worry and perceived risk of developing CRC were more likely to be screened by the FOBT. Individuals who often or sometimes worried about cancer were, respectively, 1.63 and 1.44 times more likely to be screened by the FOBT than those who rarely worried about cancer. Furthermore, respondents at greater perceived risk for developing CRC were also significantly more likely to be screened by the FOBT than those who had lower perceived risk (aOR = 1.43; 95% CI = 1.02–2.00). Regarding colonoscopy test, individuals who reported greater perceived risk for developing CRC were also significantly more likely to undergo colonoscopy than those who had lower perceived risk (aOR = 2.01; 95% CI = 1.38–2.92). Finally, subjects of ages 60–69 years who had private cancer insurance and who had higher levels of cancer worry and perceived risk were more likely to undergo screening by both the FOBT and colonoscopy; meanwhile, those of good subjective health status were less likely to undergo both the FOBT and colonoscopy screening. Linear trend analysis for four ordered categorical independent variables were estimated. In the trend analysis, subjective overall health and perceived risk of CRC showed linear trends for CRC screening regardless of screening method: those with poor health status and a higher level of perceived risk were more likely to undergo CRC screening. CRC screening rates for each screening method only differed with statistical significance according to age: individuals of ages 60–69 years were significantly more likely to undergo both the FOBT and colonoscopy screening than the FOBT only screening (p for heterogeneity = 0.010).
Multinomial logistic analysis of factors associated with up-to-date colorectal cancer screening uptake by screening method in Korea, 2013.
FOBT: faecal occult blood test; aOR: adjusted odds ratio; CI: confidence interval.
Testing for linear trend (p-value).
Discussion
The purpose of this study was to investigate the associations between cancer worry, perceived risk for developing CRC and screening for CRC in the general Korean population. Our results demonstrated that Koreans are indeed concerned with getting cancer, as 30 per cent responded that they always or often worry about getting cancer, and 18 per cent of the subjects thought that they had higher chances of developing CRC than other men or women of the same age. Across the levels of cancer worry and perceived risk for CRC, several risk factors were shown to be influential. Individuals with a family history of cancer and those who considered their subjective overall health to be poor were more likely to frequently worry about getting CRC than their counterparts. Meanwhile, males who consumed alcohol more than twice a week and smoked reported significantly greater perceived risks of developing CRC; previous studies have also reported that men feel a significantly greater susceptibility to developing CRC than women (Wong et al., 2013).
Interestingly, we found that the factors associated with perceived risk and cancer worry differed in this study. Overall, cancer worry seemed to be influenced by one’s emotional reaction to the threat of cancer, such as family history of cancer or poor subjective health status. However, perceived risk seemed to be affected by one’s subjective awareness about their likely risk of developing disease, such as male sex, alcohol drinking or smoking behaviour. Consistent with this study, previous studies have also reported that cancer worry is affected by emotional reactions to the disease and that perceived risk is affected by subjective awareness of one’s health and the presence of risk factors. One previous study reported that feelings of emotion in response to one’s own or another’s experiences with cancer can strongly influence concern for getting cancer (Peipins et al., 2015). The authors further explained that the emotional experience not only heightens cancer worry but also perceived risk. Driven not only by emotional factors, perceived risk is also affected by the presence of specific risk factors, such as smoking and the number of relatives with cancer. In this study, we discovered that cancer worry and perceived risk for developing CRC were independently associated with CRC screening uptake. Previous studies on the relationship between cancer worry and perceived risk have shown that the two variables are clearly distinct, although significantly correlated, with a moderate correlation coefficient of 0.3–0.4 (Cameron and Diefenbach, 2001; Kent et al., 2000; Lipkus et al., 2000). Studies also report independent associations for cancer worry and perceived risk with cancer screening (Moser et al., 2007).
Cancer worry and perceived risk are considered to be facilitators of screening behaviour (Halabi et al., 2000; Kwak et al., 2009; Moser et al., 2007). Indeed, recent study of cancer risk perception and screening reported relatively low rates of cancer screening among Asian women, which might be the result of low levels of perceived risk of cancer (Haas et al., 2005; Kim et al., 2014). However, some studies suggest that cancer fear and worry instead act as barriers to colonoscopy screening (Bynum et al., 2012; Green et al., 2008). In the case of having low numeracy level, respondents were more likely to be defensive against CRC screening and less likely to participate in it (Smith et al., 2016). Moreover, van Dooren et al. (2004) indicated that perceived risk of cancer is negatively associated with quality of life and psychological well-being. In this study, we found that cancer worry and perceived risk play a critical role in motivating CRC screening behaviour. Herein, perceived risk was predictive of the FOBT and colonoscopy screening; cancer worry specifically exhibited a strong association with the FOBT uptake, but not colonoscopy screening. There may be several reasons for these findings: while the FOBT is simple, safe and the most inexpensive, it is limited by poorer sensitivity, mainly for premalignant lesions. Meanwhile, although colonoscopy is a more accurate technique than the FOBT, it is invasive, carries a risk of bleeding and perforation, requires preparation and premedication, and involves much higher costs. Therefore, because colonoscopy is a more burdensome test than the FOBT, only high-risk populations tend to favour colonoscopy. Thus, a high level of cancer worry would not necessarily increase the likelihood of undergoing a colonoscopy, whereas a greater sense of higher perceived risk would. Previously reported by other studies, perceived risk, which is expected to motivate individuals to engage in cancer screening by heightened feelings of vulnerability, provoked screening uptake in our study (McCaul et al., 1996a). In a similar context, respondents who thought their subjective health status was good were less likely to receive colonoscopy and combined testing (FOBT and colonoscopy) (significance of testing linear trend = 0.007).
Our study has a number of limitations that warrant consideration. First, the cross-sectional design of this study hindered our ability to implicate any causal relationship for the observed associations. Moreover, reports of experiences with screening concerned behaviour in the past, whereas perceived risk and worry were measured as present thoughts and feelings. Thus, future studies with a longitudinal design should be conducted to track patterns in CRC screening behaviour. Second, we only measured overall cancer worry, not CRC-specific worry, as we were concerned that there might be somewhat different trends between overall cancer worry and CRC-specific worry. Thus, studies that distinguish between worry for individual types of cancer are needed.
Despite these limitations, this study is important in that we showed cancer worry and perceived risk for developing CRC to be associated with CRC screening uptake in a nationally representative sample of adult men and women. Our results further demonstrated that cancer worry and perceived risk for developing CRC are independently associated with CRC screening uptake: more specifically, cancer worry with the FOBT uptake and perceived risk with colonoscopy. These findings suggest that risk and worry are both important to understanding screening behaviour. As well, this study provides insights that may be of use in guiding the development of intervention strategies for improving compliance with CRC screening recommendations. We suggest that providing people with tailored consultation for cancer screening and education on their CRC cancer risk would help disseminate accurate perceptions of their risk thereof. In similar concept, several studies illustrated the potential application of the tailored consultation using a screening navigation programme, or identifying patterns of people’s defensive information profess. From those studies, screening navigation programme did play an important role in operating health belief, resultantly reducing social inequalities in perceiving efficacy of CRC screening (Vallet et al., 2016). According to McQueen et al.’s (2014) recent study, defensive information processing was closely related to individual difference and situational variables, such as health-care providers. So, modifying the style of communication to a particular screening test being offered may be beneficial to improving adherence with the test. Thereby, interventions to address cancer worry and perceived risk may play an important role in increasing participation and equity in CRC screening.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a Grant-in-Aid for Cancer Research and Control from the National Cancer Center of Korea (#1310232).