Survey Respondents Planning to Have Screening Colonoscopy Report Unique Barriers
Winter 2011 - Volume 15 Number 1
Tailoring colorectal cancer screening interventions to address the needs of individuals for whom screening is recommended requires accurate identification of the barriers experienced by each targeted group. The primary purpose of this survey study was to test differences in the barriers to undergoing screening colonoscopy reported by men and women. In addition, we were interested in differences in barriers reported by 1) 50-year-olds versus those age 51 to 80 years, 2) persons reporting readiness for colonoscopy versus those not reporting readiness, and 3) persons who had had a primary care encounter in the preceding 12 months versus those who had not. Four thousand members of a health maintenance organization (Scott & White Health Plan) were surveyed. Response rate overall was 30.85%. No differences in barriers to screening colonoscopy were identified for men versus women. We did identify differences in barriers reported by persons reporting readiness versus those not reporting readiness. Findings suggest that interventions to increase rates of screening colonoscopy require addressing different sets of barriers depending on whether persons report readiness to have a colonoscopy within 6 months.
Rates of adherence to recommended colorectal cancer screening remain suboptimal in the US. Targeted and tailored interventions have been used effectively to improve rates of cancer screening rates by motivating persons to seek and obtain recommended tests1–3 and may be one way to help address disparities in adherence to recommended screening that have been documented for gender, race, and ethnicity.4 The survey study we report was conducted as part of a quality-improvement effort undertaken to address lower rates of screening colonoscopy among men than among women in Texas5 and in our local system, especially younger men (ie, closer to age 50 years, when screening for average risk is first recommended). This trend is opposite of that in the US population overall, in which women are less likely to be screened or report being screened less often than men.6,7 The effort focused on screening colonoscopy because it is the preferred approach for colorectal cancer screening at our institution and per national guidelines,8 in part because of the demonstrated net benefit compared with other screening tests9,10 and because it is the only colon cancer screening tool that allows detection and removal of precancerous polyps. The purpose of the survey was to document barriers to screening colonoscopy for men and women age 50 to 80 years.
In addition to testing the hypothesis that men and women would report different barriers to colonoscopy, we were interested in testing for differences in barriers among other subgroups that may improve our ability to tailor future interventions in ways that address barriers most relevant to those groups. We hypothesized that patients who are not in regular contact with a primary care physician (PCP) would report different barriers to screening than those who had regular encounters with primary care because lack of recommendation from a physician is a barrier that is commonly documented in the literature.11,12 We also hypothesized that persons reporting more and less readiness for colonoscopy would report different barriers, according to the Transtheoretical Stages of Change model.13–15 This model is the most commonly employed model on which informed decision-making cancer-screening interventions are based.15 The model proposes that likelihood of a certain behavior, such as undergoing colonoscopy, increases as one's stage of readiness to change progresses, and interventions to move persons toward increased readiness should therefore increase the desired behavior.15 Finally, we were interested in whether barriers to colonoscopy are different for older patients than for those at about age 50 years, the age at which screening is first recommended for persons at average risk for developing colorectal cancer.
Participants were members of the Scott & White Health Plan, a health maintenance organization (HMO). They were identified as having been between 50 and 80 years of age (inclusive) between 2003 and 2008. Those in whom inflammatory bowel disease or colon cancer had been diagnosed and those with a family history of colon cancer were excluded because we were interested in barriers reported by those for whom the average risk guidelines for screening colonoscopy were most applicable.
Four samples of eligible members were selected: 1) the first 1000 men to turn age 50 years during the years 2003 to 2008, 2) the first 1000 women to turn age 50 years between 2003 and 2008, 3) a random sample of men ages 51 to 80 years between 2003 and 2008, and 4) a random sample of women ages 51 to 80 years between 2003 and 2008. No member was eligible for two lists; that is, members surveyed in the sample of 50-year-olds were not eligible to be randomly sampled in the samples of 51- to 80-year-olds. By oversampling 50-year-olds who were just reaching the age of first recommended screening, we sought to identify any unique barriers for persons who are just becoming eligible for recommended screening.
Two-page surveys (see Appendix: Colonoscopy Survey at: www.thepermanentejournal.org/issues/2011/winter/375-colonoscopy-survey-appendix.html) requested information about whether and when respondents had undergone colonoscopy, whether they intended to undergo colonoscopy within six months (readiness), gender, age, race and ethnicity, distance to a system clinic, and months since last primary care appointment. To keep the survey concise, we collapsed across stages of change proposed in the Transtheoretical Model and requested to know only whether respondents intended to be screened within six months ("yes" = "ready"). This six-month time frame captures whether respondents are in the "contemplation" or "preparation" stages of readiness, which precede the "action" stage in the model and indicate that one is likely open to health-promotion programs such as one that encourages colonoscopy.13
Respondents who had not had a colonoscopy in the preceding ten years were asked in the survey to choose from several possible barriers that have kept them from undergoing a colonoscopy. An "other" category was available for optional written comments.
Surveys returned between January and July 2009 were included in the analysis. Respondent characteristics were tabulated or described by median and range, as appropriate. Barriers to change were considered only among respondents who reported that they had not had a previous colonoscopy or who responded that their previous colonoscopy was more than ten years earlier. The primary comparison of interest was between men and women. We also compared respondents who indicated readiness for a colonoscopy versus those who did not, and respondents who had visited their PCP within the preceding year versus those who had not. Groups were compared using the Χ2 test or Fisher's exact test, as appropriate. Barriers to colonoscopy were considered for all age groups and for 50-year-olds only. Analyses were performed with SAS 9 (SAS Institute, Cary, NC). All statistical tests were two-sided. P values of <0.05 were considered statistically significant.
The overall response rate was 30.85% (1234 of 4000); it was 24.9% (498 of 2000) for 50-year-olds and 36.8% (736 of 2000) for 51- to 80-year-olds. Table 1 summarizes the demographic characteristics of all survey respondents. Seven respondents were included in the survey mailed to 50-year-olds, yet they stated that they were not 50 years of age and were excluded. Some respondents were 49 years of age yet had a birthday that year and so were included. Sixty-six percent of respondents had a previous colonoscopy, and most were within the preceding ten years. Only 25% were planning a colonoscopy within six months.
Owing to the skip pattern of the survey, data on barriers to colonoscopy were collected only among participants who had not had a colonoscopy in the preceding ten years. Therefore, the following analyses are limited to this group. Although 66% of respondents indicated that they had had a previous colonoscopy, many did not indicate a time frame, so this group was excluded.
Four hundred thirty-eight respondents answered that they either had not had a previous colonoscopy or that they had one more than ten years earlier. Table 2 summarizes the demographics and barriers to screening for this group overall and by gender. The most commonly reported barrier to colonoscopy was "My regular doctor has not told me I need a colonoscopy" (42%), followed by "I do not think I need the test" (21%). Compared with women, men were more likely to report that they did not think they needed the test (26% of men vs 18% of women; p = 0.042), that they could not take the time off from work (7.3% of men vs 2.7% of women; p = 0.025), and that they had had a bad experience with colonoscopy (2.2% of men vs 0% of women; p = 0.027).
Table 3 shows the other comparisons of interest among all respondents who had never had a colonoscopy or who had their last test more than ten years earlier. Most barriers were significantly different between those planning a colonoscopy within six months versus those who were not. Respondents planning a colonoscopy within six months were less likely to respond that they did not think they needed the test (p < 0.0001), that they were afraid it would hurt (p = 0.010), that they would be embarrassed to take the test (p = 0.018), that they did not want to do the test preparation (p = 0.001), that they were not at risk for colon cancer (p = 0.003), that the test cost too much (p = 0.020), that the distance to the test site was too far to travel (p = 0.037), or that they did not want to have the test (p < 0.0001). People who had not visited their PCP in the last year reported that it was too far to travel to get the test more frequently than did people who had visited their PCP in the preceding year (6% no visit vs 1.4% visit; p = 0.036).
Table 4 presents data only for the group of 50-year-olds, which included 319 respondents who reported that they had never had a colonoscopy or had had their previous colonoscopy more than ten years earlier. There was no statistically significant difference in barriers to colonoscopy between men and women in this subgroup. Table 5 shows the other comparisons of interest among the 50-year-old group.
Contrary to our hypothesis, men and women of all ages reported similar barriers to colonoscopy. Those just reaching the age at which screening is recommended reported similar barriers to those reported by respondents ages 51 to 80 years. However, both overall and in the subgroup of respondents 50 years of age, barriers reported tended to differ between those planning a colonoscopy within six months (conceptualized as "readiness") versus those who were not planning a colonoscopy within six months. The most commonly identified barriers to screening colonoscopy identified across groups was lack of physician recommendation.
Our findings point to the importance of tailoring colorectal cancer screening interventions, at least those designed to increase rates of screening colonoscopy, to individuals' readiness for the procedure. In addition, it points to the opportunity to address rates of screening colonoscopy in part through improved patient–physician communication about colonoscopy. However, any intervention that targets the patient–physician interaction must consider constraints on physician time for preventive care16,17; more effective rather than longer discussions are likely important.6,7
Our study was limited in several ways. First, we asked only about screening colonoscopy to the exclusion of other colorectal cancer screening tests. However, only 6.85% of those reporting barriers to colonoscopy said they had undergone another type of screening. Second, we surveyed members of an HMO receiving care within the same large integrated health care system. Less than 10% of those who had not undergone colonoscopy in the preceding ten years indicated that cost was a barrier. Nearly 85% had seen a PCP in the preceding year. The group was not representative of persons without adequate health insurance and did not include adequate representation of racial or ethnic minorities to test for differences based on these characteristics. In addition, the survey was sent once without incentives or follow-up contact. We chose this method in part to help ensure that survey recipients did not misunderstand follow-up contacts as being tied directly to their health information. We believed that it was necessary to reinforce the accurate perception that responses could not be tied to clinical information or information held by respondents' health insurer. The trade-off was a survey response rate of less than 40%. However, we were satisfied that the diversity of respondents' characteristics was reasonable, given the demographics of the counties in which HMO members lived. Patients were given contact information if they wished to obtain more information about colon cancer screening.
Continued development of interventions to encourage colorectal cancer screening in general and colonoscopy specifically should consider similarities among barriers for men and women, similarities for 50-year-olds and those who are older than age 50 years, and differences in barriers for those not demonstrating readiness for screening within 6 months.
This study was supported by a grant from the Scott and White Research Grants Program.
Katharine O'Moore-Klopf, ELS, of KOK Edit provided editorial assistance.
1. Bastani R, Maxwell AE, Bradford C, Das IP, Yan KX. Tailored risk notification for women with a family history of breast cancer. Prev Med 1999 Nov;29(5):355–64.