Summary
When consulting with patients with high-risk symptoms, only just over one-third of GPs across England would prefer to use a FIT to rule out CRC over referral to the 2WW pathway for a diagnostic colonoscopy; they highlighted the relative importance of test-specific information for the early adoption of a FIT in primary care. GPs’ likelihood of preferring a FIT increased if they believed the accuracy of the test to be high, were confident discussing the benefits with the patients, and strongly agreed that the patients would benefit from a FIT compared with having a colonoscopy. GP responders were less likely to prefer the FIT if they had had >10 2WW referrals for CRC in the previous 12 months and thought the consultations would be longer if FIT was implemented. The fact that the vast majority of GPs had not asked their patients to complete a FIT before means that their GP responder concerns about longer consultation times were not based in actual experience. These concerns present an anticipated barrier that should be considered by commissioners, who will provide pathways to FIT implementation in primary care. GPs who agreed that it is part of their role to support the implementation of a FIT, and that there is a strong need for a FIT, were no more likely to prefer the test than those who did not. Furthermore, there was no association between FIT preference and awareness of the use of FIT in primary care for patients classified as ‘low-risk but not no risk’.
Strengths and limitations
As a relatively large, nationally representative sample, this survey study was able to highlight the barriers and facilitators of implementing the FIT in primary care for patients who may present with high-risk CRC symptoms. However, there were some notable limitations. Despite the large number of responders, the actual survey response rate was <10%, thereby restricting the generalisability of the findings; however, comparable response rates have been observed in previous studies based on primary care using similar methodologies.24,26
It is also worth noting that GPs only expressed their preference in the absence of clinical recommendations. The awareness of different uses of a FIT in symptomatic context prior to the survey was low; as such, the study was designed with caution to avoid confusion that might arise from the different thresholds and recommendations that exist for the different uses of a FIT.15,18,21
Most GPs in this survey prioritised test characteristics — and, specifically, test accuracy — when considering whether to use a FIT in patients with high-risk symptoms. However, it must be noted that, even among GPs who perceived the FIT to be accurate, only 35.6% would prefer to use it over a 2WW referral. This highlights that, even if current research projects showed that a FIT could rule out cancer with 100% accuracy,17,18 more work would need to be done to overcome other anticipated barriers.
Perceptions of FIT might be further confounded by a tendency to overemphasise test sensitivity (as revealed by the relative importance that GPs attributed to true positives), despite the fact that, in the context of a rule-out test, specificity (that is, true negatives and false positives) is equally important, particularly with future sustainability of endoscopy services in mind. One potential way to understand whether accuracy is the key factor in GPs’ decisions to use FIT over 2WW would be using discrete-choice studies, which will investigate test preferences based on FIT statistics (threshold, sensitivity, and specificity) and other test attributes, such as changes in consultation times.
At present, there are very little data on the proportion of test kits that are successfully completed and returned in the diagnostic context. Thus it is important to use insights from studies into barriers to completing similar tests for CRC screening to inform future research into FIT implementation as a diagnostic test. Specifically, it has been shown that non-white ethnic groups, people with low health literacy, non-English speakers, and those who do not engage with the information are less likely to complete the test kit.27–30 As the FIT, if implemented as a rule-out test, would be aimed at patients with high-risk symptoms, it is imperative to address GPs’ concerns about their patients’ ability to complete the test kits and have appropriate safety-netting activities (for example, text-message reminders and follow-up appointments) in place in both primary and secondary care.
Implications for research and practice
As the FIT will be implemented in CRC screening in England,31 there will be potentially three different clinical contexts (screening population, and patients with low- and high-risk symptoms) in which immunochemical testing is used. Each context could vary on a number of aspects; most notably, clinical thresholds used to determine follow-up referrals. Given that GPs’ awareness of using a FIT in both rule-in and rule-out contexts was low among the survey responders, in future, particular attention should be paid to raising GPs’ awareness of different thresholds used in different contexts, as well as patients’ experience of using a FIT and their understanding of the purpose of a FIT in all clinical contexts.
This national survey suggests that a majority of GPs would prefer to use the current default test. Concerns about additional pressures on consultation time should be addressed in anticipation of any changes to NICE’s guidelines. Qualitative interviews with GPs might gain additional insights about potential barriers and identify the most effective ways of supporting implementation of a FIT as a rule-out test for CRC. The current implementation of NICE guidelines for patients with low-risk symptoms will offer further observational evidence.
