Abstract
Purpose
To understand patient experiences and persistent barriers to colorectal cancer (CRC) screening amid centralized outreach at urban family medicine practices.
Approach
Following a pragmatic trial assessing mailed fecal immunochemical test (FIT) outreach, we invited a subset of participants to complete a semi-structured qualitative interview and structured questionnaire.
Setting
Single urban academic healthcare system.
Participants
Sixty patients who were eligible and overdue for CRC screening at the time of trial enrollment.
Method
Using Andersen’s Behavioral Model, we developed an interview guide to systematically assess factors shaping screening decisions and FIT uptake. Close-ended responses were analyzed using descriptive statistics. Qualitative data were analyzed using the constant comparative method.
Results
Most participants (82%) self-reported that they had ever completed any modality of CRC screening, and nearly half (43%) completed the mailed FIT during the trial. Most patients (60%) preferred FIT to colonoscopy due to its private, convenient, and non-invasive nature; however, persistent barriers related to screening beliefs including fear of test results and cancer treatment still prevented some patients from completing any form of CRC screening.
Conclusions
Mailed FIT can overcome many structural barriers to CRC screening, yet clear communication and follow-up amid centralized outreach are essential. For some patients, tailored outreach or navigation to address screening-related fears or other screening beliefs may be needed to ensure timely completion of CRC screening.
Keywords
Purpose
Colorectal cancer (CRC) remains the second most common cause of cancer death in the United States. 1 The United States Preventive Services Task Force (USPSTF) recommends CRC screening for adults aged 50 or older using various modalities including colonoscopy and stool-based tests. 2 Timely CRC screening and follow-up on abnormal results can reduce the risk of CRC mortality through removal of precancerous polyps and early detection of more curable lesions. 3 However, CRC screening rates across the population remain suboptimal. In 2018, only 68.8% of all adults recommended for CRC screening were up to date, which falls below the national target of 80% screening coverage. 4 Additionally, the recent pandemic caused by the SARS-CoV-2 virus and related coronavirus disease (COVID-19) drastically reduced cancer screening across the country in 2020 as patients delayed preventive care encounters. 5 This necessitates implementing effective strategies to improve screening for all populations, but particularly for those at greatest risk of CRC mortality including Black patients and those with lower socioeconomic status. 6
Several interventions have effectively increased CRC screening rates, including patient outreach strategies (e.g., patient reminders and educational outreach), provider-directed interventions (e.g., electronic health record (EHR) alerts or audit and feedback), and population health approaches (e.g., centralized outreach and tracking) that reduce the burden on individual providers.7-10 In particular, programs that incorporate direct mailing of fecal immunochemical tests (FIT) to patients have been shown to be effective at increasing screening among hard-to-reach populations including those with limited access to healthcare or who may face other individual or system-level barriers.7,8 Building upon this evidence, part of our team conducted a pragmatic trial to improve CRC screening among eligible patients receiving care at urban family medicine clinics by testing the effects of financial incentives on mailed FIT completion. 11 Despite implementing organized approaches, including a centralized team to mail and track FIT kits to all patients and monitor results and follow-up within the EHR, overall completion rates were limited (23.5% at 2 months and 28.9% at 6 months), and financial incentives had no significant effect. 11 Following this pragmatic trial, we sought to understand patient perspectives on mailed FIT and other population health approaches and identify any remaining barriers to CRC screening that our outreach may not have addressed. Although previous studies have examined barriers and facilitators to CRC screening completion,12-15 our study is unique in that it focused on an urban, predominantly Black population, who are known to experience greater disparities in CRC outcomes. Additionally, this mixed methods study builds upon the results of our pragmatic trial by providing contextual insight into the potential benefits of tailoring outreach to address screening beliefs and other persistent barriers to CRC screening completion. 16
Approach
Study Population and Setting
From August to November 2018, we conducted semi-structured interviews with 60 patients who were enrolled in the pragmatic trial and who continued to receive care at academic family medicine practices affiliated with our health system. These practices serve a high proportion of racially and socioeconomically diverse primary care patients, particularly in and around West Philadelphia. All activities were approved by our Institutional Review Board prior to study commencement.
Following trial completion, we invited participants from the study to complete a one-time, semi-structured interview and structured questionnaire. At the time of trial enrollment (December 2015–July 2017), participants had to be between 50–75 years old and overdue for CRC screening. 11 Given the embedded design of the mixed methods cohort, we stratified sampling by study arm (15 participants within each of the 4 treatment groups) to ensure we reached data saturation within each group. Study arms in the pragmatic trial included the following: (1) no financial incentive; (2) an unconditional $10 incentive included with the mailing; (3) a $10 incentive conditional on FIT completion; and (4) a conditional lottery with a 1-in-10 chance of winning $100 after FIT completion. 11 The full details of the pragmatic trial have been reported in a prior publication. 11 From the 897 patients enrolled in the pragmatic trial, we randomly contacted a total of 369 patients and connected with 128 individuals; 71 agreed to participate, and 60 completed an interview. Eligible interview patients were invited by mailed letter and follow-up telephone call to participate in the study. Participants who completed an interview received a $20 gift card in recognition of their time.
Method
Data Collection
Participant characteristics overall and by colorectal cancer (CRC) screening history*.
Abbreviation: CRC, colorectal cancer.
*Table describes participant characteristics following the domains within Andersen’s Behavioral Model (ABM). 17
Significant differences by CRC screening history are bolded (P < .05 using Fisher’s Exact Test). Numbers may not sum to group totals or percentages to 100% due to missingness.
†Higher belef indicates agreement with the statement unless otherwise indicated. Each statement assesses one of five constructs related to CRC screening: C = Coherence; E = Perceived Efficacy; N = Social Norms; S = Perceived susceptibility; or W = Worry. 19
Data Analysis
Data analysis was conducted from January to August 2020. We used the constant comparative method, guided by modified grounded theory, to analyze the qualitative data. 21 Consistent with study objectives and Andersen’s Behavioral Model, we deductively identified a priori domains of interest and inductively explored emergent themes within and across participants. 16 This process included several stages of analysis. First, directed by an experienced qualitative researcher (KR), 2 trained coders from the MMRL independently read through a subset of the transcripts to identify themes within each a priori domain. Second, we used this list to develop an initial coding dictionary and apply it to a subset of the interview transcripts. Third, we iteratively refined the codebook using inter-rater reliability to document and improve coding consistency. Last, once high inter-rater reliability was achieved (kappa >.7), we applied the full coding dictionary to the transcripts using NVivo software (Version 12, QRS International) and produced thematic reports summarizing our findings. We conducted descriptive and bivariate analysis to analyze structured questionnaire data using Stata (Version 15.1, StataCorp) and used concurrent mixed methods to compare and triangulate quantitative data with qualitative patterns. 16
Results
All interview participants (n = 60) completed the structured questionnaire (Table 1). Most respondents were non-Hispanic Black (68%), female (68%), and under age 65 (75%). Nearly all had at least a high school diploma (97%) and health insurance (93%). At the time of the interview, the majority of participants (82%) self-reported that they had ever completed CRC screening (any modality). Nearly half (43%) completed the mailed FIT during the trial, which is higher than overall trial rates (29%). 11 Most participants (60%) reported that they would prefer the FIT kit if offered a choice of screening test. With regard to barriers and facilitators shaping CRC screening completion, two main themes emerged from the interview and questionnaire responses: (1) Mailed FIT overcomes many structural barriers to CRC screening but requires consistent follow-up and (2) screening beliefs continue to present as persistent barriers.
Theme I: Mailed FIT Overcomes Many Structural Barriers but Requires Follow-up
A common barrier to CRC screening was discomfort with the preparation and procedure itself, highlighted by one participant’s experience: “The colonoscopy is what was offered to me to do and I wasn’t thrilled about going in the hospital and having the procedure” (Black female, Age 62, Never Screened).
Yet, many participants described how mailed FIT provided an option to overcome barriers to colonoscopy and therefore complete screening that they had previously delayed or avoided. For example, one participant commented: “I waited several years before I even tested or went for a test. Because of the fact that I got this kit in the mail—with this option of not doing the invasive colorectal screening—I opted to do that. Otherwise, I don’t think I would have been screened at all” (White female, Age 54, Screened).
Some participants also preferred the additional privacy FIT provides, as one stated: “[I] like the fact that I can do it in the privacy of my own home” (White female, Age 53, Screened).
In addition to added privacy and avoiding invasive procedures, patients commented on the convenience of FIT. This was particularly noted by patients who needed to arrange childcare, transportation, or time off of work for appointments. As one participant described: “It was so easy. Everything was there…It’s not the horrible prep. Then it’s not taking time off from work, having to have someone pick you up. So that all is a factor into the …convenience and the ease [of the FIT kit]” (White female, Age 64, Screened).
For some patients, the ability to complete screening at home meant not having to choose between financial resources and health, as another participant noted: “The home kit was okay. That was more convenient on my days off. But as far as any other [test]…I get paid by the hour and I have bills. I don’t want to be naive about my health at the same time. If I can’t pay my bills, I won’t even have a doctor” (Black male, Age 57, Screened).
These experiences show how structural barriers to CRC screening, particularly among patients with limited resources, can be mitigated by direct outreach and mailed FIT.
Although most patients preferred FIT over colonoscopy, they also reported some disadvantages of the at-home test including discomfort with handling stool and struggling to remember to complete a test received through the mail. As stated by this patient: “If you walk out of the doctor’s office with [the kit] after the discussion, that’s probably better than having something be sent to you a month later, when you don’t even remember what it was you discussed… kind of point of sale approach would increase compliance. Here’s your kit, take it home and tomorrow do this. Put it in the mailer and send it straight back” (White female, Age 53, Screened).
Other reported barriers included inconsistent communication about why the patient had received the FIT kit and challenges ensuring all kits actually reached patients. For example, when asked for the main reason this patient had not been screened, she responded, “I kept forgetting to get one… I was supposed to get a kit sent to be at home but I didn’t get the kit” (Black female, Age 52, Never Screened).
For those that did return the mailed FIT, provider communication was essential: “Listening to what the doctor says and the reason why they said I should have it is the reason why I’ve done it. If the doctor probably never suggested it, I probably wouldn’t have had it done” (Black female, Age 57, Screened).
These comments highlight the continued importance of clear communication and follow-up amid centralized outreach.
The participants’ responses regarding mailed FIT in the structured questionnaire reflected their interview responses. As shown in Table 1, 60% of respondents indicated that they would prefer the FIT kit if offered a choice of screening test. Additionally, out of the 49 participants who had completed any form of colorectal cancer screening, 86% had ever used the mailed FIT alone or in combination with colonoscopy. These results further support the participants’ preference to use the less invasive, stool-based test.
Theme II: Screening Beliefs Are a Persistent Barrier
Among our patients, fear and other screening beliefs emerged as a persistent barrier to screening in both the qualitative interviews and structured questionnaire. Experiences reported in the interviews, particularly among those who had never been screened, described these beliefs as powerful factors that influenced their decisions to not engage in screening. For example, when asked which factors make screening more difficult, one participant responded, “Fear of having the results being positive” (Asian female, Age 56, Never Screened).
Another participant noted, even when recognized, fear remained a barrier to screening: “Fear allows you to manufacture stories or justifications in your mind to avoid doing something that you don’t really… want to do” (Black female, Age 65, Never Screened).
For these patients, fear of positive test results and downstream effects emerged as a persistent and powerful barrier to CRC screening.
Some patients elaborated on their hesitation to be screened because of the subsequent care that might follow a positive test result: “I just don’t need another health issue and so I think I’ve been stupidly avoiding [screening]” (Black female, Age 65, Never Screened).
When asked why he had not been screened, another patient responded: “To go there and find out…something else [is] going on with you and then they give a treatment plan that might not always be 100% of curing what’s going on with you…You take something and they give you some type of medication that’s supposed to cure what’s going on with you, but it affects the other organs in your body. I take my chances with God and the remedies at home” (Black male, Age 60, Never Screened).
Regardless of mailed FIT or direct outreach, these quotes highlight how uncertainty, distrust, and fear about medical care continue to serve as barriers to screening for some patients.
The responses to the structured survey support our qualitative findings, particularly with regard to the impact of fear among the never screened patients. In the structured survey (Table 1), we found clear differences in screening beliefs between screened and never screened patients, some of which reached statistically significant difference. A majority (73%) of those never screened agreed with the statement, “I am afraid of having an abnormal colon cancer screening test result” as opposed to 43% who had been screened. Additionally, among those who had never been screened, when compared with those who had, we observed higher rates of perceived cancer risk (18% vs 8% agree with the statement: “The chance that I might develop colon cancer is high”) and significantly lower perceived benefits of CRC screening (18% vs 61% disagree with the statement: “I will be just as healthy if I avoid having colon cancer screening;” P = .017). By comparing the perspectives of groups with different screening histories, we observed that fear and other preconceived screening beliefs have served as real barriers that prevented some patients from engaging in screening, even in the context of reduced structural barriers.
Discussion
In this study, we sought to understand persistent barriers and facilitators of CRC screening in underserved populations and how these factors may have shaped the success or failure of our centralized approaches. Similar to previous studies, we found that the option to complete CRC screening using mailed FIT alleviated many barriers associated with colonoscopy.22,23 Our findings build upon prior research by offering contextual insight into the specific factors that make at-home testing a favorable option for many patients, including convenience, privacy, and less invasive implementation. Additionally, our data further support the importance of using direct outreach to increase CRC screening rates. These findings are particularly important for practitioners working to recover screening rates in response to the COVID-19 pandemic. An analysis from one large test site in the US that continued to mail FIT kits to patients during the pandemic showed that it was able to maintain testing rates. 5 Our data expand upon existing evidence by providing contextual insight into the benefits of a remote screening option that does not require in-person contact for the first round of testing.
This study also emphasizes the need to complement direct outreach with messaging that targets negative screening beliefs which may hinder timely completion of screening.24-26 We used Andersen’s Behavioral Model to systematically assess factors that shape screening decisions and found that predisposing characteristics, specifically colorectal cancer screening beliefs, were highly influential. Fear of positive test results and the burden of subsequent cancer treatment were leading factors cited by our patients who had never been screened. Some patients commented that they would rather risk developing cancer than undergo the discomfort and fear of screening and the chance of finding new health problems, especially if they already had other health issues. These particular results are not surprising, as several studies found these factors to be predictive of cancer screening12-15; however, strategies to mitigate negative screening beliefs are not often included in mailed FIT outreach in routine care, and there is mixed evidence on the effectiveness or feasibility of tailored approaches in trials.27-29 Our findings underscore that effective and feasible strategies addressing negative screening beliefs are needed to ensure all patients complete screening including potentially using enhanced outreach by culturally sensitive patient navigators or tailored messaging for patients who do not complete mailed FIT. 30
Additionally, our results highlight the continued importance of enabling factors such as clear communication from care team members to patients throughout the centralized screening process, from notifying patients why they are receiving the mailed FIT through timely communication of results and follow-up care.7,31 Our findings provide unique insight into why centralized outreach like the approach used in our pragmatic trial may not affect screening completion rates for all patients. Our results also highlight the need for further research on the potential benefits of tailoring outreach to address patients’ fears and other screening beliefs, in conjunction with efforts to reduce persistent structural barriers. Finally, with regard to factors of perceived and evaluated need, our study showed notable differences among patients with different screening histories in how they perceive the benefits of screening and their personal cancer risk. These findings further emphasize the importance of tailored communication, particularly among patients who have not been previously screened for colorectal cancer and are less likely to view screening as beneficial.
Our study is strengthened by the use of a theory-based, mixed methods approach to evaluate barriers to CRC screening among a predominantly Black and urban population, who are known to experience greater disparities in CRC outcomes. Additionally, to increase internal validity, we both stratified by trial study arm to enhance capture of experience by intervention exposure and used random sampling within each arm to select eligible participants for the interviews. The limitations of this study include that it was conducted at a single academic healthcare system, so our results may not generalize to other systems or populations. Additionally, we did not collect EHR data following trial completion and therefore relied upon patient-reported screening history to categorize patients as ever or never screened. This limits identification of patients who are overdue for screening, which is likely a large proportion of patients given their inclusion in the trial. Finally, due to the sampling approach, in which we sought data saturation within each study arm to balance participants by intervention exposure within the trial, we did not assess saturation of themes with different racial groups or by sex. Therefore, we cannot systematically speak to differences by these factors within the qualitative analysis. This is an important area of future study to ensure intervention strategies reduce rather than exacerbate persistent disparities in CRC screening.
Conclusion
In conclusion, there are various barriers that prevent eligible patients from engaging in CRC screening, but many of these factors can be addressed by offering at-home tests to patients and reducing underlying fears and concerns. By providing the at-home test option when appropriate and using clear communication and follow-up, practitioners are more likely to increase screening rates. For patients who remain hesitant to screen due to fears or other beliefs, it is important to identify effective strategies to specifically target these barriers. Addressing negative screening beliefs through enhanced outreach may lead to greater increases in CRC screening completion and subsequent decreases in CRC mortality across populations. Colorectal cancer (CRC) screening rates remain suboptimal across the U.S., particularly among underserved populations including those with limited access to healthcare or who may face other individual or system-level barriers. Programs that incorporate direct mailing of fecal immunochemical tests (FIT) to patients have been effective at increasing screening among these populations. A predominantly Black cohort of urban family medicine patients mostly preferred FIT to colonoscopy due to its ability to overcome structural barriers, yet barriers related to screening beliefs including fear of test results and cancer treatment still prevented some patients from completing any form of CRC screening. Offering FIT to patients when appropriate and addressing negative screening beliefs through enhanced outreach may lead to greater increases in CRC screening completion and subsequent decreases in CRC mortality across populations.So What?
What is already known on this topic?
What does this article add?
What are the implications for health promotion practice or research?
Footnotes
Acknowledgments
Contributors: We are grateful for the support and insights provided by all investigators, research staff, and particularly to the study participants for taking the time to speak to us about their experiences. Prior presentation: Aspects of this manuscript were presented at the 2019 Annual Meeting of the American Public Health Association (APHA) in Philadelphia, Pennsylvania.
Author Contributions
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr Mehta has received an honorarium from the American Gastroenterological Association. No other potential conflicts of interest were reported by the authors of this paper.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by a Health Research Formula Grant Award through the Abramson Cancer Center at the University of Pennsylvania from the Commonwealth of Pennsylvania. Dr Mehta’s time was supported by the National Cancer Institute of the National Institutes of Health [K08CA234326]. Shannon Ogden is supported by a training grant from the National Institute on Drug Abuse [T32DA041898].
Ethical Approval
All research activities were approved by the Institutional Review Board at the University of Pennsylvania prior to study commencement [Protocol # 823388].
