Introduction
Primary care physicians, especially in systems like the UK’s National Health Service, are crucial in the early diagnosis of cancer. General Practitioners (GPs) are often the first point of contact for individuals experiencing symptoms, playing a vital role in initial assessment and referral for specialist care when cancer is suspected. Despite this critical role, cancer survival rates in the UK lag behind many other high-income nations (Cancer Research UK, 2019; Arnold et al., 2019). This disparity may stem from various factors, including delayed patient presentation, GPs overlooking potential cancer indicators during consultations, or delays in the referral process to specialist services (Al-Azri, 2016; Bowen and Rayner, 2002; Koyi et al., 2002). Improving early cancer detection in primary care is therefore a significant public health priority.
Cancer risk assessment tools (RATs) have emerged as a promising strategy to enhance early cancer diagnosis in primary care settings. These tools are designed for use with symptomatic individuals and estimate the likelihood of cancer based on a combination of risk factors and presenting symptoms. By providing a quantitative risk estimate, RATs like QCancer aim to support GPs in making more informed decisions about further investigation and referral, potentially leading to earlier diagnosis and improved outcomes (Hamilton, 2009; Hippisley-Cox and Coupland, 2013). Currently, QCancer (Hippisley-Cox and Coupland, 2013) and the RAT (Hamilton, 2009) are two prominent cancer RATs being promoted for use in UK general practice.
QCancer has undergone independent validation (Collins and Althman, 2012; Collins and Althman, 2013) and demonstrated its accuracy in predicting cancer risk within primary care populations. However, the successful implementation of such tools hinges not only on their accuracy but also on their acceptability and usability within the complex environment of primary care. Crucially, understanding the perspectives of both service users (patients) and primary care practitioners (GPs and practice nurses) is essential to identify potential barriers and facilitators to widespread adoption. This study aimed to address this gap by exploring the perceived barriers and facilitators to implementing QCancer in primary care consultations from the viewpoints of both service users and primary care practitioners.
Methods
This research adopted a qualitative approach, mirroring the methodology of a previous study by Akanuwe et al., (2020) due to both studies originating from the same broader research project and employing a similar design. The study utilized semi-structured individual interviews and focus groups to gather in-depth qualitative data from participants residing in Lincolnshire, located in the East Midlands region of England. Ethical approval for the study was granted by the School of Health and Social Care Ethics Committee at the University of Lincoln.
A convenience sampling method was employed to recruit service users without a cancer diagnosis and primary care practitioners willing to participate. To acknowledge their time commitment, practitioners were offered standard backfill costs for their participation in interviews. Patients with a pre-existing cancer diagnosis were excluded, as cancer RATs are not intended for this population. Additionally, individuals who had been recently referred, were undergoing cancer assessment, or had consulted with symptoms that might raise cancer concerns were also excluded to avoid potential distress associated with discussing cancer risk and diagnosis.
The interview schedule, designed to elicit barriers and facilitators to cancer RAT implementation, was informed by the Consolidated Framework for Implementation Research (CFIR) (Damschroder et al., 2009). The CFIR provides a comprehensive theoretical framework for understanding the multifaceted factors influencing the implementation of innovations, such as cancer RATs, in healthcare settings. Data analysis and interpretation were also guided by relevant CFIR constructs, including:
- Relative Advantage: Practitioners are more likely to adopt cancer RATs if they perceive them as offering advantages over existing methods, such as clinical guidelines alone.
- Patient Needs and Resources: The extent to which cancer RATs effectively address and enhance patient needs and utilize available resources is a crucial factor in their implementation success.
- Compatibility: This refers to how well cancer RATs integrate with existing primary care workflows and systems, and how they align with the values, norms, perceived risks, and needs of both practitioners and patients.
- Knowledge and Beliefs: Practitioners’ understanding of cancer RATs, their skills in using them, and their overall attitudes towards these tools significantly impact their willingness to adopt them.
- Reflecting and Evaluating: Feedback and outcomes derived from the practical application of cancer RATs in primary care play a vital role in ongoing implementation and refinement.
Participants were recruited through various channels, including flyers distributed in public locations (e.g., libraries, community notice boards) and through collaborations with a patient and public involvement group. Individual interviews were deemed more suitable for service users to facilitate open discussion of potentially sensitive information related to personal risk factors. Service users interested in participating contacted the researcher (JA) for further details and to schedule a face-to-face interview, which could take place at their home or at the university, based on their preference.
Practitioners were invited to participate via letters sent to their general practices. Those interested contacted the researcher (JA) to arrange either an individual interview or a focus group, depending on their preference and practice logistics. Individual interviews allowed for in-depth exploration of individual practitioner perspectives, while focus groups facilitated valuable interactions and discussions within practice teams, generating rich and diverse data.
All participants provided written informed consent prior to participation and were assured of their right to withdraw at any point. Audio recordings were made of all interviews and focus groups, with supplementary notes taken to capture contextual details and non-verbal cues. Before eliciting participants’ views, a vignette illustrating the QCancer tool was presented, explained, and demonstrated using either a computer or a paper-based version, ensuring all participants had a basic understanding of the tool’s functionality.
Data analysis followed the framework approach (Ritchie and Spencer, 1994), facilitated by NVivo version 10 qualitative data analysis software. An initial coding framework was developed based on a priori codes derived from the interview guide and the CFIR. Inductive codes emerged iteratively as data analysis progressed, allowing for the capture of emergent themes and nuances within the data. Two researchers (JNA and ANS) independently reviewed transcripts to develop the initial coding framework, which was subsequently discussed and refined by the entire research team. Through iterative interpretation and discussion, initial codes were synthesized into a smaller set of overarching themes representing key barriers and facilitators.
Data collection ceased when thematic saturation was reached, indicating that no new codes or themes were emerging from the data (Hennink et al., 2016). Service user and practitioner data were analyzed separately initially, and then compared to identify areas of convergence and divergence in their perspectives. To enhance the trustworthiness and rigor of the research, the study adhered to the Consolidated Criteria for Reporting Qualitative Studies (COREQ) (Tong et al., 2007), with detailed reporting available in Supplementary Table S1.
Results
A total of 36 participants were included in the study, comprising 19 service users (aged 21-71 years) and 17 practitioners (aged 33-55 years). Data collection took place between September 2014 and September 2015. Two service users had a prior history of cancer, and all other service users reported having family members or friends with a cancer diagnosis, potentially influencing their motivation to participate in the study. Notably, no participants withdrew from the study after providing initial consent. Table 1 provides a detailed breakdown of participant characteristics.
Table 1. Participant Characteristics
Service users | Practitioners |
---|---|
Gender | |
Male | 7 |
Female | 12 |
Age group (years) | |
20–29 | 3 |
30–39 | 4 |
40–49 | 1 |
50–59 | 3 |
60–69 | 5 |
70–79 | 3 |
Ethnicity | |
White British | 19 |
Indian | – |
Pakistani | – |
Asian British | – |
Bangladeshi | – |
Practice patient list size | |
200–2900 | – |
3000–3900 | – |
4000–4900 | – |
5000–5900 | – |
6000–6900 | – |
7000–7900 | – |
8000–8900 | – |
9000–9900 | – |
The analysis identified several overarching themes representing both barriers and facilitators to the implementation of QCancer. Identified barriers included: the anticipated need for additional consultation time; concerns about unnecessary worry related to cancer investigations; the potential for over-referral and consequent burden on services; practitioner skepticism regarding the tool’s value; the need for comprehensive training on tool utilization; and the necessity to demonstrate the tool’s effectiveness compared to current practice. These barriers are elaborated upon below, with further details, illustrative quotes, and links to the CFIR framework provided in Supplementary Table S2.
Barriers to Implementation
Additional Consultation Time
Aligned with the CFIR constructs of readiness for implementation and patient needs and resources (Damschroder et al., 2009), service users and practitioners alike expressed concerns about the already demanding workloads of GPs and nurses and the potential for cancer RATs to further increase consultation times. A service user noted: ‘Practitioners in general practice would need more time to use the tool in consultations‘ (Service User 7: individual interview). Practitioners echoed this sentiment: ‘It’s more a question of more time really, because at the moment we’re in crisis. So, we don’t want more work‘ (Practitioner 11 [GP]: focus group [FG] 2).
Unnecessary Worry Related to Cancer Investigations
Concerns were raised by both service users and practitioners regarding the potential for increased anxiety and unnecessary worry stemming from a rise in cancer investigations triggered by RAT use. This aligns with the CFIR concept of ‘patient needs and resources’ (Damschroder et al., 2009). Participants emphasized the importance of clear communication to patients, ensuring they understood that the tool provides a risk assessment, not a definitive cancer diagnosis.
‘Some people may not understand and they can be too worried especially if they don’t explain that it is just a risk but it is not guaranteed that they will get cancer‘ (Service User 11: individual interview).
‘If you tell the patient they’ve got 1% cancer, which is creating unnecessary anxiety, they will say doctor, you said I have got 1% chance of getting cancer and you are not doing anything about it‘ (Practitioner 2 [GP]: individual interview); ‘…you can probably make them more worried‘ (Practitioner 16 [Practice Nurse]: FG 3).
Over-referral and Over-burdening Services
Reflecting the CFIR construct of patient needs and resources (Damschroder et al., 2009, some participants expressed apprehension that increased referrals resulting from RAT use could place undue strain on already stretched healthcare services.
‘It could lead to over-referral as some people may have a certain risk but will not have cancer after they have been referred and tested‘ (Service User 17: individual interview); ‘It will put a strain on the NHS; you don’t want to over burden the services as well‘ (Practitioner 4 [GP]: individual interview).
Conversely, other practitioners believed that their clinical judgment, used in conjunction with the RAT, would mitigate the risk of over-referral:
‘We are not just referring but we are using our clinical judgements as well, so we would only refer those patients that need to be referred – so I don’t think there will be over-referrals‘ (Practitioner 1 [GP]: FG 1).
Practitioner Scepticism
Consistent with the CFIR construct of knowledge and beliefs of individuals involved in the implementation process (Damschroder et al., 2009), participants, particularly practitioners, highlighted the potential barrier of skepticism among colleagues who may doubt the value of new tools, especially if they lack understanding of their application. One practitioner admitted: ‘…until you said this thing, you know initially I was very sceptical about this tool‘ (Practitioner 3 [GP]: individual interview).
However, other practitioners demonstrated a more positive inclination towards using QCancer:
‘I believe it will be good to use a cancer risk assessment tool to facilitate earlier diagnosis of cancer, and as you know, earlier diagnosis will help with earlier treatment‘ (Practitioner 2 [GP]: FG 1).
Conflict with Existing Guidelines
Service users emphasized the need for consistent guidance, while practitioners expressed potential confusion arising from the integration of cancer RATs with existing guidelines, such as those from the National Institute for Health and Care Excellence (NICE). This aligns with the CFIR construct of complexity (Damschroder et al., 2009). Participants stated:
‘I think it is good for everybody to have the same sort of guidelines, so everybody should use the same sort of guidelines‘ (Service User 1: individual interview).
‘I will be quite confused about using the tool. With the NICE guidelines, you couldn’t focus on another criterion for any other risk here. I mean there are implications for investigations, referrals…, it has to be very much a repeated approach‘ (Practitioner 11 [GP]: FG 2).
High-Risk Symptoms Override Risk Scores
Participants articulated the view that the presence of symptoms strongly suggestive of cancer should necessitate referral for further investigation, irrespective of the risk score generated by the tool.
‘It doesn’t really matter about percentages; I know 1% is less risk. But the fact is the symptom is there, the coughing out of blood, which is quite worrying‘ (Service User 13: individual interview).
‘Regardless of what the tool said I will refer them for investigation with the symptoms. So, it doesn’t matter 1% or 0%, I will always do one thing, investigation if the symptoms are suggestive of cancer‘ (Practitioner 11 [GP]: FG 2).
Need for Training
A significant barrier identified by practitioners was a perceived lack of understanding regarding how to effectively integrate the tool into consultations, including accurate risk calculation, interpretation of risk scores, and communication of results to patients. They emphasized the critical need for comprehensive training on the practical application of the tool in patient consultations.
‘We don’t quite understand how to use that tool. I think we need to have proper education or training on using these tools‘ (Practitioner 2 [GP]: FG 1).
Establishing Effectiveness
Service users advocated for rigorous evaluation of QCancer’s effectiveness in patient consultations before widespread implementation.
‘I think if you are going to roll something out…. I would start with the doctors, see how the doctors do with it after evaluation and then move on to the practice nurses‘ (Service User 12: individual interview).
Practitioners similarly stressed the importance of evaluating the tool to enable comparison of its effectiveness with current clinical practice: ‘We have to make sure that it is better than what we are already doing‘ (Practitioner 13 [GP]: FG 3).
Facilitators to Implementation
Facilitators to QCancer implementation, aligned with the CFIR construct of relative advantage (Damschroder et al., 2009, encompassed perceptions that the tool could enhance clinical decision-making, promote positive modifications in patient health behaviors, streamline cancer assessment and treatment pathways, and facilitate personalized patient care.
Supporting Clinical Decision-Making
Both service users and practitioners expressed the belief that the tool could assist in making more informed decisions regarding cancer investigations and referrals. One service user stated the tool would help to, ‘make decisions appropriately‘ (Service User 1: individual interview).
A practitioner elaborated:
‘I think the tool will help to guide the clinician to see the broad level of differential diagnosis. It will also facilitate referral of patients by presenting a quantitative risk value to help explain risk and make a decision‘ (Practitioner 2 [GP]: FG 1).
Modifying Patient Health Behaviors
While designed as a RAT for symptomatic individuals, participants recognized the potential of the tool to raise awareness and encourage modifications in health behaviors.
‘I think it might be just raising awareness, so people realise what’s happening, and what can go wrong with them and where the risks are and may be, they can reinforce them. Where someone else like the young person who has given up smoking it might be used to reinforce by saying well, you’ve got a very low risk, so if you’ve given up smoking carry on with that. Rather than saying you’ve got a very high risk later‘ (Service User 5: individual interview).
‘I also feel the tool will help in terms of using the risk generated to advise patients who need behavioural changes. If their risk was small, I would tell them to maintain healthier lifestyles by exercising, eating a healthy diet, less alcohol and to stop smoking if they were smoking. Yes, as I said, this tool can help to empower patients to take control of their risk factors and live healthier lifestyles‘ (Practitioner 2 [GP]: FG 1).
Improving Processes and Speed of Cancer Care
Service users and practitioners anticipated that the tool could expedite cancer diagnosis by improving the efficiency of assessment and treatment processes.
‘I do think it will be a useful idea, yeah. I think my first worry is that I may have cancer and most of us will like to know early so they can get it sorted. But a lot of things can be picked up, can’t they, if they spot check risk‘ (Service User 4: individual interview).
‘I think when the tool is fully integrated in our IT systems and every practitioner gets familiar with using it, it will be time saving in the long term, as the consultation, the assessments, investigations and referral processes will be faster‘ (Practitioner 1 [GP]: FG 1).
Personalizing Patient Care
Participants perceived that the tool would facilitate more patient-centered care by providing risk assessments tailored to individual patients, enabling personalized care plans rather than generalized approaches.
‘I think it will make the care more patient-centred because you’re presenting them with their own risk not a general risk, it’s personal to them and it will just make the consultation more patient focused, and I think it will make patients feel more involved in the consultation and just feel more cared for‘ (Service User 12: individual interview).
‘Patients will go away with a lot more targeted information about their personalised risk of cancer rather than a vague statement‘ (Practitioner 1 [GP]: individual interview).
Discussion
This study identified a spectrum of barriers and facilitators influencing the potential implementation of QCancer in primary care. Barriers included concerns about increased consultation times, patient anxiety, over-referral, practitioner skepticism, integration with existing guidelines, the need for training, and demonstrating effectiveness. Facilitators centered on the tool’s perceived ability to support clinical decision-making, accelerate assessment and treatment, promote healthy behaviors, and personalize patient care. A key strength of this study is its exploration of perspectives from both primary care practitioners and service users, providing a comprehensive understanding of the multifaceted challenges and opportunities associated with cancer RAT implementation. The finding that cancer RATs may contribute to personalized patient care and health behavior modification adds valuable insights to the existing literature on cancer RATs.
Strengths and Limitations
This study is among the first to directly compare the perspectives of both service users and practitioners regarding QCancer, a cancer RAT specifically designed for symptomatic individuals in general practice settings. The use of both individual interviews and focus groups enriched the data, capturing diverse perspectives and facilitating in-depth exploration of key themes. Furthermore, the achievement of data saturation, in terms of both codes and meaning (Hennink et al., 2016), strengthens the robustness of the findings.
However, the study also has limitations. Notably, all service user participants were of White British ethnicity, despite efforts to publicize the study broadly. This lack of ethnic diversity may limit the generalizability of findings to other populations, as cultural and linguistic factors could influence perceptions of cancer RATs. Previous research suggests that individuals from ethnic minority groups may be underrepresented in research due to factors such as limited awareness, language barriers, or differing levels of research engagement (Gill et al., 2013; Lo and Garan, 2008; Redwood and Gill, 2013).
Another limitation is the lack of consideration for participants’ educational or socio-economic backgrounds, particularly among service users. These factors could potentially shape their views on cancer RATs. However, the study’s primary focus was on comparing the perspectives of service users as a group to those of practitioners, rather than exploring variations within the service user group itself.
Comparison with Existing Literature
The concern regarding additional consultation time required for risk calculation, patient communication, and discussion of further investigations aligns with existing literature highlighting the time constraints in primary care and the potential for new technologies to add complexity (Damschroder et al., 2009). Integration of cancer RATs into existing general practice IT systems, coupled with effective training, is crucial to mitigate time-related barriers and streamline tool utilization.
While participants expressed concerns about potential patient anxiety related to cancer investigations, a systematic review of cancer RATs in primary care found no evidence of increased cancer worry (Walker et al., 2015). This discrepancy may reflect anticipatory anxiety among participants who had not yet experienced RAT use, highlighting the importance of clear and sensitive communication strategies to address patient concerns and manage expectations (Akanuwe et al., 2020; Kim et al., 2018).
Over-referral emerged as another significant concern. Current guidance emphasizes that cancer decision support tools should prompt consideration of cancer risk but ultimately rely on clinical judgment for referral decisions (Macmillan Cancer Support, 2015). While some practitioners in this study acknowledged the importance of combining RAT outputs with clinical judgment, the potential for increased referrals and associated costs remains a valid consideration.
Practitioner skepticism, observed in this study and in simulation studies (Chiang et al., 2015), underscores the need for robust evidence demonstrating the clinical effectiveness of cancer RATs in improving patient outcomes. Lack of trust in risk calculations, particularly when they conflict with clinical intuition, can hinder implementation (Chiang et al., 2015). Demonstrating clear benefits and providing adequate training to enhance understanding and confidence in RATs are essential to address practitioner skepticism.
The need for training and guidance highlighted by practitioners in this study echoes findings from Dikomitis et al., (2015), who reported challenges faced by practitioners in using cancer RATs in routine practice. Addressing clinicians’ learning needs through comprehensive information and training programs (Jones et al., 2000; Sowden et al., 2001) is critical for successful implementation.
The observation that high-risk symptoms may override RAT scores reflects the importance of clinical judgment and guideline adherence in cancer referral pathways. Macmillan Cancer Support (2015) suggests that cancer decision support tools can complement NICE guidelines by providing prompts and alerts, while still allowing clinicians to exercise their judgment based on clinical presentation and NICE criteria.
Facilitators identified in this study, such as improved decision-making and faster pathways to diagnosis and treatment, are supported by evidence from Green et al., (2015), who found that RATs aided GPs in symptom recognition and referral decisions for lung and colorectal cancer. Furthermore, embedding decision support tools to augment, rather than replace, clinical judgment is crucial for successful adoption (Green et al., 2015).
The potential of cancer RATs to promote positive health behavior changes, identified as a facilitator in this study, is also supported by systematic review evidence suggesting that health promotion messages within RATs can influence behavior change (Walker et al., 2015).
Finally, the personalized care aspect of cancer RATs aligns with the broader shift towards person-centered healthcare, emphasizing individual patient needs, values, and preferences in care planning (Gill et al., 2013; Sepucha et al., 2008).
Implications for Practice and Further Research
Addressing the identified barriers is crucial for successful implementation of Primary Care Cancer Risk Assessment Tools. Strategies to mitigate time constraints, such as streamlining tool integration into IT systems and optimizing consultation workflows, are necessary. Providing comprehensive training and ongoing support for practitioners is essential to enhance their confidence and competence in using these tools. Clear communication strategies are needed to address patient anxieties and ensure realistic expectations regarding risk assessment and investigation pathways.
Quantitative research is warranted to rigorously evaluate the impact of cancer RATs like QCancer on referral rates, diagnostic yield, overdiagnosis, and, most importantly, patient outcomes compared to current practice. Further research should also explore strategies to optimize the integration of RATs into diverse primary care settings and address potential disparities in access and utilization across different population groups.
Conclusion
This study provides valuable insights into the perceived barriers and facilitators to implementing QCancer, a primary care cancer risk assessment tool. While the identified facilitators offer encouraging pathways for adoption, the barriers must be carefully considered and proactively addressed to ensure successful and equitable implementation in primary care settings. Ultimately, realizing the full potential of primary care cancer risk assessment tools to improve early diagnosis and patient outcomes requires a multifaceted approach that addresses both technical and human factors influencing their integration into routine clinical practice.
Acknowledgements
We extend our sincere gratitude to the service users and primary care practitioners who generously participated in this study and shared their valuable perspectives.
Supplementary Material
For supplementary material accompanying this paper visit http://dx.doi.org/10.1017/S1463423621000281.
phcsup.zip (42KB, zip) click here to view supplementary material
Authors’ contributions
ANS conceived the original study idea. JA and ANS designed the study, with support from SO and SB. JA conducted the fieldwork and analysis, supported by ANS, SO, and SB. JA drafted the initial manuscript, and all authors contributed to editing and approving the final version.
Financial support
This research was supported by a grant from Lincolnshire Partnership NHS Foundation Trust.
Conflict of interests
The authors declare no competing interests.
Ethics approval and consent to participate
The University of Lincoln School of Health and Social Care Ethics Committee approved the study. All participants provided informed consent prior to participation. The study was conducted in accordance with the principles of the Declaration of Helsinki.
References
[References will be added here, mirroring the original article’s references]
Associated Data
Supplementary Materials
For supplementary material accompanying this paper visit http://dx.doi.org/10.1017/S1463423621000281.
phcsup.zip (42KB, zip) click here to view supplementary material