The contribution of a negative colorectal screening test result to symptom appraisal and help-seeking behaviour among patients subsequently diagnosed with an interval colorectal cancer

2.1 Design

Semi-structured individual face-to-face interviews were carried out with patients diagnosed with an interval colorectal cancer following a negative gFOBt result through colorectal cancer screening (bowel screening programmes). Interviews were carried out within 6 months of patients receiving their cancer diagnosis.

2.2 Setting and recruitment

All patients newly diagnosed with a primary colorectal cancer and who had received a negative gFOBt result (within the preceding two years) from the Scottish bowel screening programme or the Midlands and North West screening programme hub were eligible to take part in the study. Eligible screening age for the two respective programmes was 50–74 years (Scotland) and 60–74 years (England). This study selected patients using a purposive approach, recruiting patients via the hospital clinic where they received or were currently receiving their cancer treatment in NHS Tayside, Scotland, and University Hospitals Coventry & Warwickshire NHS Trust, England. Lists of patients, reviewed by the multidisciplinary teams (MDT) following a colorectal cancer diagnosis, were sent quarterly to the local screening hub in Dundee, Tayside or Rugby, North West Midlands. Confirmation of receipt of a negative gFOBt result was confirmed through checking screening hub records, and the patient list was returned to the colorectal surgeon responsible for the patient's care.

Patient invitation letters were sent directly to the patient from the secondary care team on behalf of the research team, including a participant information booklet, response form and prepaid addressed return envelope. Once a response form was received by the researcher (KB), individuals were contacted by telephone to arrange a suitable time and date for the interview.

2.3 Exclusion criteria

Suitability based on clinical grounds (physical or mental health) was at the discretion of the secondary care team and primarily the patient's colorectal surgeon in charge of their care.

2.4 Data collection

Interviews were carried out by a female researcher, KB (MA, MSc, PhD), with previous training and experience in conducting qualitative research. Interviews took place between August 2013 and June 2014. Interviews lasted between 45 and 60 minutes, and signed consent was obtained before the interview commenced. All interviews were audio-recorded with the participant's permission, professionally transcribed verbatim and anonymized. Field notes were made both during and after each interview. Audio files were deleted once the written transcripts had been received and verified by the researcher. Interviews took place in the patient's home, and some patients were accompanied by their spouse or friend during the interview.

2.5 Topic guide

Interviews explored two key domains: (i) the participant's “pathway to diagnosis” including: initial symptom appraisal, symptom progression, first presentation to any healthcare provider and any perceived delays, and (ii) the contribution, if any, of their negative gFOBt result to their symptom appraisal or decision to seek help.

2.6 Ethical review

The project was granted ethical approval from the South East Scotland Research Ethics Committee 01 (11/SS/0006).

2.7 Analysis

All transcripts were read by both the project researcher (KB) and the principal investigator (CC). Emerging themes were discussed and a coding frame agreed with additional codes included when appropriate as data collection progressed. Final themes were agreed through an iterative process involving the core and wider research group. Thematic analysis of the data was undertaken using NVivo software (QSR International, V.10) and was ongoing throughout the study to allow emerging themes to be fed back into the data collection. An inductive reasoning approach was adopted where themes (or categories) were identified through careful examination and comparison of the data which involved 6 stages: familiarization of the data, generating initial codes, searching for themes, reviewing themes, defining and naming themes and producing a report.27

3.1 Recruitment

A total of 27 participants were recruited, with 17 interviews carried out in Tayside, Scotland, and 10 in the Midlands and North West, England. Recruitment was ongoing throughout the study; 50% of patients who received an invitation to take part in the study agreed to be interviewed. Twenty-one participants presented to their GP with symptoms prior to their diagnosis, two attended their GP for a routine diabetic health check, one participant was diagnosed following an emergency admission to hospital and two participants were diagnosed following a surveillance colonoscopy. One participant, wrongly identified, had received a positive gFOBt result in the screening programme, which led to their diagnosis: this interview was not included in the reported findings. A summary of study recruitment and participant demographics is provided in Table 1.

3.2 Reported symptoms

Participants described a range of symptoms and symptom characteristics, which differed in frequency, duration and severity prior to receiving their colorectal cancer diagnosis. Reported symptoms included: noticeable blood on toilet paper or in the toilet bowl or stool following a bowel movement, constipation and/or diarrhoea, change in bowel habit, weight-loss, indigestion, nausea and sickness, fatigue and pain. Participants who were diagnosed following a routine diabetic health check reported that they did not experience any symptoms preceding their cancer diagnosis.

3.3 Main themes

We found the relationship between a negative gFOB result and a patient's diagnostic journey to be multifaceted, with varying opportunities or scenarios in which screening participation and the receipt of a negative result had the potential to influence the appraisal interval following the onset of symptoms (Figure 1). We identified three key stages where a negative gFOBt result had the potential to contribute to the appraisal and/or help-seeking interval outlined in the Model of Pathways to Treatment,15 and a fourth stage where participants reflected on their screening participation following their interval colorectal cancer diagnosis. The potential impact of a negative gFOBt result at each stage was influenced by a number of emergent and cross-cutting themes: (i) receipt of a negative screening (gFOBt) result (Trust and Reassurance), (ii) recall of a negative test result following the onset of symptoms (Cancer Suspicion and Time Since Last Screen), (iii) the subsequent contribution, if any, of a negative gFOBt result to symptom response (Extended Symptom Appraisal & Delayed Help-seeking) and (iv) reflection on the bowel screening programme following an interval colorectal cancer diagnosis (Knowledge/Acceptance of the Limitations of gFOBt screening). These 4 key stages including the emergent themes and subthemes are shown in Figure 1.

3.4 Response to receipt of a negative gFOBt result

3.5 Recall of negative gFOBt result following the onset of symptoms

3.6 The subsequent contribution, if any, of a negative gFOBt result to symptom response

3.7 Reflection on the bowel screening programme following an interval colorectal cancer diagnosis

4.1 Principal findings

In this study, we examined whether the receipt of a negative gFOBt result contributed to the pathway to diagnosis among screening participants subsequently diagnosed with an interval colorectal cancer. Patient accounts highlighted the variability and complexity of presenting symptom type and contributing factors that influenced the patient interval. Participants placed a high level of trust in their screening gFOBt result, interpreting it as an “all clear” and seldom questioned the reliability of the result or the limitations of the screening test when they received their result letter. The majority of participants did not consider their previous gFOBt result (nor consider a potential colorectal cancer diagnosis) following the onset of symptoms, and therefore, reassurance from a negative gFOBt did not contribute to a delay in the patient interval. However, for a smaller group of patients who did consider colorectal cancer as a possible cause of their symptoms (or were experiencing symptoms intermittently or at the time they completed their screening test), a negative gFOBt result did influence the patient interval and contributed to a delay in seeking help. A negative gFOBt result led participants to “downplay” the seriousness of their symptoms by reducing their cancer suspicion and reinforcing “benign” causal beliefs having been reassured by a recent “all clear” test result. Pre-existing conditions, particularly when the characteristics of any new symptoms were similar to those previously experienced, led to some delay across the symptom appraisal, help-seeking and diagnostic interval.15

4.2 Comparison with existing literature

The processes within the appraisal interval as part of the Model of Pathways to Treatment,15 namely “patient appraisal and self-management,” draw on a number of psychological theories including Leventhal's Common Sense Model of Illness self-regulation and the Andersen Model of Total Patient Delay.16, 28 Consistent with our findings, Leventhal's model explains that initial symptom appraisal, or recognition of a bodily change, when symptoms do not exceed a certain level of interference, leads to symptoms being normalized or dismissed. However, when symptoms persisted, or increased in severity (eg exceeded a threshold of interference), patients adopted a different approach where they self-monitored their symptoms or sought self-treatments prior to making the decision to seek medical advice. Furthermore, as reported in the literature, we found that many participants did not attribute their symptoms to a potential cancer diagnosis despite some experiencing a colorectal cancer “alarm” symptom, for example blood in stool.2

Lack of recognition of the seriousness of a symptom has been reported extensively in the literature as one of the main contributing factors to a delayed patient interval.18, 19 In both the present study and the study carried out by Solbjor and colleagues examining interval breast cancers following mammography screening, a negative screening result was shown to impact on the patient interval primarily through reassurance which led some participants to “downplay” the seriousness of their symptoms and to assume a benign cause.26 Hall and colleagues found that a negative screening result, along with a number of other factors including absence of blood or pain, or “feeling well,” was viewed as reassuring.23 In the same way, reassurance from a previous “all clear” investigation has been associated with a delay in help-seeking even when symptoms persisted or occurred a number of years later with some patients believing that they would not be taken seriously if they returned with similar symptoms.20, 21 The added complexity of the presence of comorbidities (or competing diagnoses) and the associated delay on the diagnostic pathway potentially resulting in late-stage disease is an important and recognized challenge for healthcare providers,29, 30 particularly in the light of the recent evidence published by Torring and colleagues who report any delay to the primary care interval to be associated with more advanced colorectal cancer.31

Interval colorectal cancer has been shown to have an adverse effect on trust in FOBt screening, with the same study reporting poorer quality of life among interval cancer patients when compared to those with screen-detected disease.32 Similarly, this has been demonstrated in mammography screening where a diagnosis of an interval breast cancer influenced patient's trust, although not to the point of creating distrust, where women instead saw themselves as exceptions in an otherwise beneficial screening programme.33 We also found that despite some interviewees suggesting a loss of confidence in the screening programme following their diagnosis, participants generally remained positive towards the overall benefits of screening and expressed a willingness for continued participation.

4.3 Strengths and limitations

Participants were recruited from hospitals in both Scotland and England, providing narratives from cancer patients who participated in different area-based bowel screening programmes. Carrying out semi-structured interviews within six months of diagnosis permitted an in-depth exploration of the participant's pathway to diagnosis; nevertheless, accounts are retrospective and therefore subject to recall and framing bias. We did not have access to any clinical or demographic information for those who declined to be interviewed, nor the number that the hospitals did not consider eligible to contact: their experiences and perspectives may differ.

4.4 Implications for practice and policy

Although reassurance from a negative gFOBt result is appropriate for most screening participants, there is a risk of over-reassurance for some. Methods of increasing understanding of the limitations of gFOBt screening (and indeed of FIT screening) and the ongoing risk of developing cancer despite receiving a negative test result are needed; the significance of any associated delay in seeking medical advice in terms of clinical outcomes also merits further exploration. Similarly, finding effective ways to engage with screening participants that complement existing initiatives in primary care with regard to symptom vigilance and prompt help-seeking behaviour remains an important challenge; there is a need to develop nuanced health messages to inform screening participants of the importance of help-seeking for vague as well as alarm symptoms following a negative test result, particularly in the presence of comorbidities. The use of FIT as a diagnostic tool in primary should allow more streamlined diagnostic pathways for all patients.34