A better FIT: Risk stratification of screening intervals and FIT thresholds in bowel cancer screening

Abstract

The risk of developing bowel cancer is associated with numerous lifestyle and genetic factors, meaning that individuals with a higher risk gain substantially more from screening than those at low risk. Risk stratification has therefore been suggested to improve the balance of screening harms and benefits, and to accommodate resource limitations. Within the bowel cancer screening pathway, risk stratification could be incorporated at the point of first invitation (eligibility), or at the point of the screening test via stratified screening intervals and/or referral thresholds for colonoscopy. The focus of my thesis is risk-stratified faecal immunochemical test (FIT) thresholds and screening intervals. In order for risk stratification of bowel cancer screening to be successful, the public must find it acceptable, understandable, and be willing to participate in a risk-stratified screening programme. This thesis aims to understand public acceptability and communication needs associated with risk stratification of FIT thresholds and screening intervals within population-based bowel cancer screening, and to investigate the potential impact of risk stratification on bowel screening uptake. My first two studies are mixed methods systematic reviews of the acceptability of risk-stratified cancer screening from the perspective of healthcare professionals (HCPs) and the general public respectively, using the Joanna Briggs Institute convergent integrated approach. I included seven papers about HCP attitudes and found that risk stratification was broadly acceptable, providing additional support and training is delivered prior to implementation. There was a paucity of research in relation to cancers other than breast and greater evidence is needed to understand acceptability of de-intensified screening for low-risk groups. I identified 22 papers considering public attitudes and found that acceptability was generally high. I then developed ten priorities for implementation across four key areas: (I) addressing public information needs; (II) understanding communication preferences for risk estimates; (III) mitigating barriers to acceptability; (IV) the role of HCPs in supporting reduced screening among low-risk groups. To explore public acceptability of risk stratification in greater depth, specifically in relation to risk-stratified screening intervals and FIT thresholds for bowel cancer, I conducted two community juries. I found that risk stratification in this context was acceptable to informed members of the public, and the use of data within the current system in the form of age, sex, and FIT result, was considered preferable to existing age-based screening. The use of additional lifestyle and genetic data was acceptable but required greater consideration of practical aspects before implementation. Although participants acknowledged concerns for low-risk individuals they felt this could be mitigated by carefully considered public communication. One means of public communication is via written screening resources such as information leaflets. My fourth study, a series of think aloud interviews paired with a user testing survey, aimed to develop and test public understanding of a bowel cancer screening leaflet relating to risk-stratified screening intervals. I conducted 13 think aloud interviews in which participants identified 42 suggested changes to the screening leaflet, including clearer language, simplification of concepts related to risk stratification and additional information to evidence the safety of extending screening intervals for those at low risk. The refined leaflet was then tested using a true or false style survey (n=20). A minimum of 90% of participants answered each user testing statement correctly, above the pre-established threshold of understanding suggesting the final leaflet was fit for purpose and further rounds of revision and user testing were not required. Overall, I found that the public are capable of understanding information about risk-stratified bowel cancer screening presented as part of population-level screening programme information. The information text developed in the think aloud study formed the basis for the final study in my thesis. The text was used as part of an online survey which aimed to quantify public acceptability and anticipated uptake of potential approaches to risk stratification at three points on the bowel cancer screening pathway. Participants received information about four possible screening programmes: (I) screening as usual; (II) risk-stratified eligibility; (III) risk-stratified FIT threshold; (IV) risk-stratified screening intervals. Overall acceptability of screening programmes employing risk-stratified eligibility criteria and FIT thresholds was significantly greater than for screening as usual (p<0.001). This high level of acceptability was maintained when participants were asked about screening practices for high-risk individuals (p<0.0001), indicating strong acceptability for escalation of screening. The inverse was true of de-escalating screening for low-risk groups, as low-risk practices were significantly less acceptable than an age-based approach and up to 16.8% fewer participants would take up FIT screening in this scenario compared to screening as usual (p<0.0001). Despite these differences, anticipated uptake for low-risk scenarios was over 75%. A similar pattern was observed when participants were asked about the impact of risk on their decision to attend colonoscopy whereby the higher the risk level, the greater the likelihood of attending. Female participants were less likely to accept one or more risk-based strategy despite reporting screening as usual to be acceptable (odds ratio [OR] 1.5, [95% confidence interval [CI] 1.1-2.0]). Finally, a combination of data derived from primary care records, lifestyle questionnaire and genetic data was most acceptable within each strategy, indicating a preference for a more comprehensive approach to data collection. This body of research demonstrates that risk stratification is likely to be an acceptable adjustment to the UK bowel cancer screening pathway, specifically at the point of FIT. I have identified several requirements that must be fulfilled in order to facilitate successful implementation of such an approach, including training and support for HCPs, changes to screening infrastructure and increased evidence needs. Additionally, the importance of communicating decreased screening for low-risk groups cannot be overlooked, as this represents a pivotal opportunity with potential to determine the ultimate success of risk-stratified bowel cancer screening. Future research should explore acceptability of risk stratification from an experiential perspective, seek to measure uptake of this approach in reality and ascertain the most viable strategy for grouping the population according to risk.