Surveillance versus ablation for incidentally diagnosed small renal tumours: the SURAB feasibility RCT

Attitudes to a trial of active surveillance versus ablation

A total of 79% of the respondents said they would take part in a trial of ablation versus active surveillance in this patient group. Some of those who would not, or were unsure, gave their reasons in the open questions (Box 1).

Almost half (47.5%) of respondents anticipated problems in patients agreeing to a trial comparing ablation with active surveillance, 27.5% did not anticipate problems and 25% were unsure. The reasons given by those who believed it would be problematic to recruit were centred around patient preference for one particular option and patients not being at ease with active surveillance: ‘Some patients are uncomfortable with knowing they have a “tumour” and not having it “dealt with” ’. More than half (55%) of respondents did not envisage any other problems with a trial of this kind. Of the remaining 45%, the other problems given were related to the patients’ perspective, covered in the previous question, but the next most commonly mentioned problem was the lack of a surgical arm and what this would mean for patients who were eligible for surgery or ablation. There were comments that this is a much needed trial but consideration should be given to the potential problems within the clinical community.

Factors that influence which patients are approached

Age was explored as a factor that may have an impact on which patients are approached for the trial. Respondents were asked what age of patients they would feel comfortable recruiting to such a trial from an age range of 45 years to ≥ 80 years. The majority (80%) would be happy to recruit patients aged 75 years, but this dropped to 65% for patients aged 70–89 years and to 25% for patients aged ≥ 80 years; 47% would recruit patients aged 65 years and 15% would recruit patients aged 45–55 years.

When asked to rank the potential influential factors in the identification of patients for the trial of tumour size, age and presence of comorbidities, tumour size was ranked as the most important for 59% of respondents (Figure 1). The ranking of comorbidity was interesting, with a similar number of respondents ranking it first (n = 13), second (n = 15) or third (n = 12), indicating, perhaps, that opinions on the importance of comorbidities were divided.

FIGURE 1.

Factors ranked as most influential in identifying patients for the trial.

Vignettes

Respondents were asked to consider a series of vignettes (see Table 1, for example) with a range of ages (50–80 years), tumour sizes (2–3.5 cm) and level of comorbidity (Charlson score of 0 or 2 points) in relation to willingness to recruit to the trial.

Less than half (45%) of respondents would be willing to recruit patients in the 50–59 years and 60–69 years age groups with tumour size of 2–2.5 cm and no comorbidity (Figure 2). As tumour size increased in this age group, willingness decreased to 23% for tumour size of 3 cm and to 15% for tumour size of 3.5 cm. There was a marked difference in the views of urologists: 47% agreed to approach when tumour size was 2 cm, but figure this dropped to 8% for a tumour size of 3.5 cm (Figure 3). For radiologists this figure was 44% and 22%, respectively (see Figure 4).

FIGURE 2.

Willingness to recruit by age and tumour size and no comorbidity.

FIGURE 3.

Willingness to recruit by age and tumour size with comorbidity.

For patients aged 70–80 years, the majority (86%) of respondents would recruit if tumour size was of 2 cm and there was no comorbidity. This dropped substantially in the 70–79 years age group with increased tumour size: 48% at 3 cm and 30% at 3.5 cm. Similarly, for the 80–89 years age group, fewer (64%) would recruit if the tumour size was of 3.5 cm than of 3 cm (89%).

The presence of comorbidity (Charlson score of 2 points) had an influence. The majority (86%) of respondents would recruit patients aged 50–69 years with a tumour size of 2–2.5 cm or 3 cm (81%), but this figure dropped to 58% for tumour size of 3.5 cm (Figure 3). There was little or no difference in urologists’ willingness to recruit patients aged 50–79 years with tumours from 2 cm to 3 cm (89–100%) and comorbidity, but this proportion fell to 74% (aged 50–59 years) and 79% (aged 60–69 years) for tumours of 3.5 cm (Figure 4). For radiologists, this dropped from 83% for tumours of size 2–2.5 cm to 72% and 39%, respectively, for tumours of 3 cm and 3.5 cm in size (see Figure 4).

FIGURE 4.

Willingness to recruit by clinician type, age, tumour size and comorbidity. (a) Urologists’ willingness to recruit – no comorbidity; (b) urologists’ willingness to recruit – comorbidity (Charlson score of 2 points); (c) radiologists’ willingness to recruit – no comorbidity; and (d) radiologists’ willingness to recruit – comorbidity (Charlson score of 2 points).

There was an increase in the proportion of respondents who would recruit patients with comorbidity in the 70–79 years age group, to 91% for a tumour size of 3 cm, but this dropped to 64% for a tumour size of 3.5 cm (see Figure 3). For the 80–89 years age group with comorbidity, the percentages dip for a tumour size of 2–2.5 cm (77%), increase for a tumour size of 3 cm (86%) and drop for a tumour size of 3.5 cm (62%). The main difference between radiologists and urologists was in the percentage who would recruit patients aged 80–89 years with comorbidity and a tumour size of 2–2.5 cm (89% of radiologists, compared with 66% of urologists; see Figure 4).

Limitations

The response rate from urologists was disappointing and was a limitation of this survey as we were able to canvass the views of only half of this group of clinicians. The BAUS, unlike the BSIR, did not circulate the request on our behalf, which might have raised the profile of the survey and resulted in a better response rate. However, the number of interventional radiologists in the UK who conduct ablation is small, and we believe that we were able to capture the views of the majority of this group.

Contact information for activity: getting it right

Designing the information sheet about involvement in the development of the trial information was not straightforward. In order to explain the purpose of, and rationale for, this activity, it was necessary to mention the trial, and there was a concern that patients may think that they were being approached to be trial participants. We needed patient involvement in the design of our information sheet, particularly to ensure it was clear that those approached were not being asked to take part in a trial. However, after contacting two sources (PPI leads from local and national NIHR Clinical Research Networks and a local cancer patient group), support was not forthcoming. To avoid running into difficulties with the timetable for this activity, we had to accept that the best we could do was to rely on our SURAB lay coapplicant and a clinical colleague from a local trust, who commented on the content of the information sheet and suggested some changes.

Identification and contact of patients for the activity

It was considered important to involve patients who had been diagnosed with small cell renal cancer and successfully treated within the last 12 months. They would remember when they received their cancer diagnosis and understand how it would feel being approached to take part in a trial, particularly one that has a no-treatment arm. One hospital site had agreed to send out a letter from the clinician who had cared for the patient, with an information sheet to explain why they were being contacted, a reply sheet they could complete and return (to the researcher) if they were happy to be contacted and a pre-paid envelope. These patients were then contacted by the qualitative researcher who answered any questions they had and determined if they were willing to help. As this was PPI, consent was not required.

Response rate

In the first tranche, 20 letters were sent to patients in one hospital site. Unfortunately, none of those who returned the slip wished to be contacted. We then had to contact another hospital site to request help. One of the research nurses and the database manager from that site helped to identify eligible patients. A second tranche of 20 letters was sent, and 13 patients (nine men and four women) responded to say they were willing to be contacted. Telephone contact was made to answer any questions about the activity and to ascertain whether they had a preference for a group or one-to-one session. Only one person had misunderstood and thought they were being asked to take part in a trial. After learning more about what was required, one person declined any further involvement. Of the remaining 12, three expressed a preference for working in a group session and seven for a one-to-one session; two expressed no preference.

Stage 1: group session

Over several telephone calls with the five people willing to attend a group session, a convenient date and time were arranged. We asked if they would be able to attend for 3 hours. As we had no information other than their name and telephone number from the form they returned, we requested their home address (so we could send out an invitation letter and details of venue) and asked how they would travel to the venue (to organise parking spaces and permits), whether or not they wished to bring a partner and if they had any health or mobility problems we should be aware of. We also sent a copy of the information materials we would be working with should they have the time to look through them before the group met.

Three women and two men attended the group session. No one brought a partner. Ages ranged from 54 years to 77 years. Based on postcode and past or present occupation, four of the five group members’ socioeconomic status was categorised as A/B, which is described as professional and qualified to a high level.

The task began with the researcher (JL) briefly explaining the trial and the purpose of the activity. The participants gave some background on their experience of kidney cancer and what treatment they had received. The patient information materials were handed out and we began to go through each section of the patient information sheet (PIS) in detail.

Several general points were made: the sheer length was off-putting; it was repetitive; it had inconsistencies; and a higher than average level of literacy would be required to fully understand it. The key condition was described in different ways throughout the PIS as ‘renal tumour’, ‘kidney tumour’, ‘kidney cancer’ and ‘mass’. Similarly, the no-treatment arm was described as ‘active surveillance’ or ‘careful observation’. Some parts were considered superfluous or repetitive and some caused alarm. In response to the sentence ‘The tumour normally grows very slowly and can often be safely observed for a period of time’, the group wanted to know ‘how often?’ and ‘for what period of time?’. They thought that the positive aspects about being involved in the trial should be brought to the fore. It was suggested that bullet points of key information be given on the first page, then patients could make a decision if they want to read on. Alternatively, they thought that a short information sheet could be provided, with a longer, more detailed sheet given to patients who wish to know more and who agree to participate. A smaller booklet format was discussed, and it was thought that this might be less daunting. The bullet points should emphasise that patients would not have been approached if they were at risk, and the slow growth rate of this tumour, and should describe what action would be taken if patients in the active surveillance group patients experience any change in their condition. It was felt to be unnecessary to have the information sheet divided into parts 1 and 2. The group did not believe that there was a need to explain why patients are randomised. It was felt that there was too much detail about assessments and it was suggested that a flow chart be included. It was also suggested that information about the qualitative interviews could be reduced and simplified, and the need to call them ‘qualitative’ was queried.

The exercise took around 2 hours and attendees preferred to take refreshments (water, tea, coffee and biscuits) while they worked. Reimbursement was offered for travel and a gift voucher was given to each attendee as a gesture of thanks for their time.

Stage 2: one-to-one session

Amendments were made to the PIS in the light of the comments and suggestions of the group. The major changes were that bullet points were added to the front of the PIS and some of the information around the process for patient participants was replaced with a flow chart. As a result, the number of pages was reduced from 11 to nine.

The seven people who expressed a preference for one-to-one sessions were contacted again and a time was arranged to speak to them face to face. In six cases, the session took place in the participant’s home, and one session, at the request of the individual, was held at the university where the study team was based. The dates and times were confirmed in a letter, which was accompanied by the second version of the PIS so that participants could look through it beforehand.

Six of the sessions went ahead; one patient cancelled due to illness. Five participants were male and one female. This group spanned a broader range of socioeconomic status, categorised as C1/2 to D/E, which ranges from lower qualified professionals (e.g. nurses) to unskilled and casual workers.

As with the group exercise, the background to the study and the purpose of the visit was explained. Two people were slightly perplexed and bemused by the idea of PPI in research and some of the allotted time was taken to explain this in greater depth. The PIS was introduced and, as before, we went through it in detail, exploring each section using cognitive interviewing techniques. Most preferred the researcher to read each sentence and then explore the understanding and identify anything that was not clear.

Individuals tended to remark on similar points in the PIS. The main comments about the bullet points were to tighten up and simplify the text and give further reassurance about the active surveillance arm (that these tumours are slow growing). Comments on the ‘What is the purpose of the study?’ question can be summarised as ‘cut to the chase’ and include only the key points. This section was also perceived to indicate a great deal of uncertainty. Potential trial participants were mainly concerned about the risk of the cancer spreading if untreated and wanted to know how slowly the tumour normally grows, how long it can be observed for, how often these tumours can be observed and what the criteria for observation are. They also thought that the two interventions should be explained. There was a dislike of the term ‘randomisation’ and a preference for ‘random allocation’ or ‘allocated at random’. Some individuals did not understand the rationale for randomisation and thought this should be included. Patients liked the flow chart, and the only comment was about the term ‘randomisation’.

It was considered insensitive to mention in the risk information the minor risk of cancer associated with CT scans when the target population has just been diagnosed with cancer. It was also thought tha, in the same section, the text ‘some people may feel anxious about not receiving treatment and the possible risk of their cancer growing’ may cause concern in patients who had not considered this. One individual was participating in a cohort study and the study information was in a small (A5 size) booklet format that he found easy to read.

Stage 3: Newcastle Clinical Trials Unit and clinical team

The NCTU team reviewed the amended PIS following the group and one-to-one sessions to ensure that the PIS met the regulatory standards. The team members, in general, were very happy with the revised version but, as there is a requirement to inform potential trial participants of alternative treatments, requested that the section ‘What are the alternative treatments?’ be reinstated. The group had found this confusing (as two of the alternative treatments were part of the trial) and thought that this was already covered in the section on the purpose of the study. A compromise was to ensure that details of all potential treatment and management options were listed in the purpose of the study section at the front of the PIS. The clinical team was satisfied that the information around the condition and treatments and the rationale for the SURAB trial were fully explained. The amended PIS was submitted to the NHS REC, which accepted the changes without question.

Inclusion criteria

• Adults with a renal tumour < 4 cm (confirmation by radiology or by biopsy) in size.

• American Society of Anesthesiologists (ASA) physical status classification system grades 1 or 2.

• Aged ≥ 18 years.

• A CT/MRI scan of abdomen/chest with no evidence of metastases.

Exclusion criteria

• Patient’s clinician does not feel they would be suitable for the trial (e.g. because of concomitant disease).

• Multiple SRMs in one kidney.

• Coagulopathy that cannot be corrected.

• Previous participation in this study.

• Inability to give informed consent (carer/proxy consent will not be allowed in this study).

Primary

The primary outcome of the trial is feasibility defined as:

1. willingness of patients to be randomised, assessed by reviews of patient screening logs and defined as:

   1. number of patients consenting to be randomised as a proportion of all eligible patients approached about the trial, with reasons for non-consent

   2. qualitative assessment of barriers to, and facilitators of, recruitment

2. willingness of clinicians to randomise patients, assessed via qualitative interviews

3. assessment of retention and dropout rates, defined as:

   1. number of patients who started randomised treatment as a proportion of the number randomised, with reasons for early dropout

   2. number of patients who completed randomised treatment as a proportion of the number randomised, with reasons for early dropout (including death)

   3. qualitative assessment of barriers to, and facilitators of, data collection and participant retention.

Secondary

Compliance with interventions and trial processes, defined as the number of patients who completed patient-reported outcomes at each time point as a proportion of the number randomised, including baseline, with reasons for non-compliance. Outcome measures for a definitive trial will be tested in the feasibility study and applied before treatment, at baseline and 3 months and 6 months after treatment.

The following measures were collected:

• The Short Form questionnaire-36 items (SF-36) score. The SF-36 is a 36-item questionnaire that measures QoL across eight domains, which are both physically and emotionally based. A single item is also included that identifies perceived change in health, making the SF-36 a useful indicator for change in QoL over time and treatment. A summary of physical QoL [physical component summary (PCS)] and emotional QoL [mental component summary (MCS)] is produced. The percentage scores range from 0% (lowest or worst possible level of functioning) to 100% (highest or best possible level of functioning).

• The Functional Assessment of Cancer Therapy – General (FACT-G) score. The FACT-G is a 27-item compilation of general questions divided into four primary QoL domains: physical well-being, social/family well-being, emotional well-being and functional well-being. It is considered appropriate for use with patients with any form of cancer.

• The State–Trait Anxiety Inventory (STAI) score. The STAI is a measure of the severity of current symptoms of anxiety. There are two subscales: (1) state evaluates the current state of anxiety and subjective feelings of apprehension, tension, nervousness and worry; and (2) trait evaluates relatively stable aspects of ‘anxiety proneness,’ including general states of calmness, confidence and security. The range of scores for each subtest is 20–80 points, with a higher score indicating greater anxiety.

We also developed and tested health economic data collection tools in the form of a participant costs questionnaire (PCQ). The PCQ has two parts: (1) part A, to be administered at 3- and 6-month follow-ups, and (2) part B, to be administered at 6 months only.

The end of study was the last participant’s final study contact at their 6-month post-treatment follow-up.

Identification and screening of patients for the feasibility study and parallel qualitative component

The trial flow chart (Figure 5) illustrates this process as well as data collection time points.

FIGURE 5.

Trial flow chart.

Potential participants were identified in the renal cancer clinics at participating sites. Site principal investigators (PIs) and/or clinical colleagues with documented delegated responsibilities for patient identification and screening performed this task.

An eligibility screening log was completed by the investigator to document participants’ fulfilment of the entry criteria for all patients considered for the study and parallel qualitative component and, subsequently, included or excluded. This information was anonymised and transcribed into screening logs on an ongoing basis via a secure online database.

Patients who agreed or declined to take part in the feasibility study were given an information sheet about the qualitative study and their permission to pass on their contact details to the qualitative researcher was requested. These patients were then contacted by the qualitative researcher, who answered any questions they had and, if they were happy to participate, arranged a convenient time to conduct an interview.

Clinical trial consent

Eligible patients were contacted by the centre PI/nurse lead/nurse specialist to invite them to participate in the trial and parallel qualitative component.

Informed consent was undertaken by appropriate site staff, as per the site delegation log. The delegated staff member (usually the centre PI/nurse lead/nurse specialist) explained the trial and parallel qualitative component to the patient, gave them the information leaflet and answered any questions they had.

Patients were encouraged to take the information leaflet home to discuss it with family and friends and asked to arrange a suitable time for a second meeting (allowing at least 24 hours for this). However, if participants had travelled a long distance to the hospital and would not be returning until an intervention visit, and if returning to hospital for the consent process would be a burden, consent was taken on the same day as information provision. In this case, a member of the local study team phoned the participant 48 hours later to confirm that they still wished to take part. This conversation was documented in the patient’s medical notes.

If a patient declined to participate in the trial and/or qualitative component, the study team (with permission from the patient) documented any reasons available for non-participation in the eligibility screening form and transferred this to the anonymised site screening log. The screening forms and logs ensured that potential participants were approached only once.

After ensuring that the patient had understood the information provided, the research team member delegated to take consent asked the patient to sign and date the consent form agreeing to participate in the feasibility study and/or the parallel qualitative component. Consent by the patient was witnessed and dated by the delegated research team member taking consent.

Written informed consent was always performed before randomisation or any other study-specific procedures/investigations.

The original signed consent form was retained in the investigator site file, with a copy in the clinical notes, a copy faxed to the NCTU (for centralised monitoring) and a copy provided to the participant.

For patients who agreed to take part in the qualitative study, a different consent form was completed at the time of the interview. Owing to the small subject population, the information sheet and consent forms were available only in English.

Randomisation

When all eligibility checks had been made and written informed consent had been obtained, participants in the feasibility study were randomised in a 1 : 1 ratio (stratified by centre) to either active surveillance or ablative treatment (RFA, CRYO or MWA as per centre practice). Randomisation was undertaken using the central web-based randomisation service provided by the NCTU.

The PI at the site or the individual with delegated authority accessed the web-based randomisation system. Patient screening ID, initials and centre (the stratifying variable) were entered into the web-based system, which then returned the allocation status (successful randomisation was followed up by an automated confirmatory e-mail to the site and relevant staff).

Participants were then informed of their allocated treatment group by the site PI or delegated individual following randomisation.

Following allocation, the site organised:

• the procedure date for those allocated for ablative treatment

• the active surveillance protocol for those allocated to the active surveillance arm.

Blinding

This was a feasibility study with the primary outcome defined as recruitment and retention rates and qualitative exploration of the patients’ experiences and understanding of the randomisation process and treatment options. Randomised treatments could not be blinded owing to the type of interventions being researched. For this reason, staff were not blinded for the follow-up assessments.

The baseline data capture assessments were to be completed before randomisation in order to reduce any bias in terms of patient attitude to allocated treatment affecting baseline data.

Baseline

Following written informed consent, a routine, standard care biopsy was performed post consent as part of standard care prior to randomisation in study participants who had not already received this clinically. In sites that did not perform this procedure as standard care, the participant consented to the procedure and the study prior to the biopsy. A compulsory renal biopsy at the core of the ablated lesion was to be performed on all patients who had received ablative treatment. There were three possible biopsy outcomes based on the histology and immunohistochemistry interpretation:

1. failed biopsy

   1. no tissue or non-renal tissue

   2. normal renal tissue

2. fibrosis

   1. suggestive of ablation site (inflammation, haemorrhage or haemosiderin)

   2. not suggestive of ablation site (old fibrosis, no evidence of recent inflammation or haemorrhage).

   3. inconclusive

3. renal tumour (classification and grade)

   1. tumour cells with Ki67 reactivity

   2. tumour cells without Ki67 reactivity

The baseline (pre-randomisation) visit involved collection and retrospective collation of the following data:

• Demographics (e.g. age, sex).

• Medical history.

• Blood tests and urinalysis (taken as per local policy, with no requirement to report these to the study team or record them on the study eCRF).

• SF-36.

• STAI.

• FACT-G.

• Biopsy results were recorded in the baseline eCRF to document relevant details and presence of cancerous growth. If a routine diagnostic biopsy was not standard care at site, consent for the biopsy and the study was given prior to the biopsy procedure.

• Tumour size and volume were recorded from routine CT/MRI scans already obtained during diagnosis of the tumour. Size was recorded in three planes of measurement (volume = 0.5326 × diameter 1 × diameter 2 × diameter 3). The volume calculation was not to be completed by site.

• Confirmation was recorded in the eCRF that study inclusion criteria was fulfilled and no exclusion criteria applied.

Patients were randomised only after the baseline visit, review of biopsy results and checking of inclusion/exclusion criteria. Patients were not randomised if the biopsy showed a non-cancerous result.

Three-month follow-up (for all participants)

It was planned that the 3-month follow-up of patients in the ablation arm would take place 3 months (SD 14 days) after the treatment date. In the case of patients in the active surveillance arm, it was planned that this should take place 3 months post randomisation (SD 14 days).

At the 3-month follow-up, the following information was requested for all patients:

• results of blood tests and urinalysis (taken as per local policy, with no requirement to report these to the study team or record them on the study eCRF)

• SF-36 score

• STAI score

• FACT-G score.

Imaging results were obtained in order to document any changes in tumour progression as follows:

• For patients receiving ablation (using any of the methods), data were to be captured from a routine CT/MRI of the abdomen performed within 3 months of the ablative procedure.

• For patients receiving active surveillance, a 3-month post-randomisation scan of the abdomen was optional depending on routine practice at site.

The imaging should allow capture of the following data on the eCRF:

• Tumour size and volume. Size was recorded in three planes of measurement (volume = 0.5326 × diameter 1 × diameter 2 × diameter 3). The volume calculation was not to be completed by site.

• Confirmation of any progression of the tumour.

The following additional data were requested for all patients:

• any changes in treatment plan (e.g. switching from active surveillance to ablation or surgical excision) with dates, times and types of procedures recorded and reasons for a switch in treatment (i.e. reason for crossover)

• complications of treatment and AEs

• results of the administration of the health economics questionnaire: PCQ (part A).

Six-month follow-up (for all participants)

It was planned that the 6-month follow-up of patients in the ablation arm would take place 6 months (SD 14 days) after the treatment date. In the case of patients in the active surveillance arm, it was planned that this should take place 6 months post randomisation (SD 14 days).

At the 6-month follow-up the following was requested for all patients:

• results of blood tests and urinalysis (taken as per local policy, with no requirement to report these to the study team or record them on the study eCRF)

• SF-36 score

• STAI score

• FACT-G score.

For patients who received an ablative procedure, data were captured from a routine follow-up CT/MRI scan of the abdomen performed at 6 months after the procedure in order to document any changes in tumour progression.

For patients receiving active surveillance, data were captured from a routine follow-up CT/MRI scan of the abdomen performed at 6 months after randomisation in order to document any changes in tumour progression.

The imaging allowed capture of the following data on the eCRF:

• Tumour size and volume. Size was recorded in three planes of measurement (volume = 0.5326 × diameter 1 × diameter 2 × diameter 3). The volume calculation was not to be completed by site.

• Confirmation of any progression of the tumour.

Tumour volume was calculated from the three-dimensional diameters using the formula to calculate an ellipsoid volume:

If there was progression of the growth rate or the size of the tumour in patients on the active surveillance arm, ablation or partial nephrectomy would be offered depending on the facilities available in the participating centre.

Progression was to be considered to have occurred if growth rate exceeded 2.5 mm per 6 months or tumour volume had doubled by 6 months.

The following additional data were captured for all patients:

• any changes in treatment plan (e.g. switching from active surveillance to ablation or surgical excision) with dates, times and types of procedures recorded and reasons for a switch in treatment (i.e. reason for crossover)

• complications of treatment and AEs

• results of the administration of the health economics questionnaire: PCQ (parts A and B).

Progression criteria to main trial

Recommendation to move to a definitive, fully powered, multicentre trial is based on the following calculations:

1. The upper 90% confidence interval for the proportion of patients recruited should exceed 50%. The trial should recruit at least 49 patients from 120 approached. This was based on a requirement that the underlying recruitment rate should be at least 50%. The exact 90% confidence interval corresponding to 49 successes from 120 Bernoulli trials is from 32.0% to 50.2% (with 48 successes from 120 trials, the upper interval drops to below 50%). Should recruitment be < 49 patients, this would provide evidence that the recruitment rate is too low and the feasibility of a RCT would be questionable. The upper 90% confidence interval for the retention rate (the proportion of patients recruited who have been followed up) should be ≥ 80%.

2. Economic assessment should suggest that further research likely to be worthwhile.

Adverse events

All non-serious AEs during study participation were to be reported on the study eCRF and sent to the trial manager within 1 month of the form being due. Severity of AEs was graded on a three-point scale (mild, moderate or severe). Relation (causality) and seriousness of the AE to the treatment was to be assessed by the investigator at site in the first instance. The individual investigator at each site was responsible for managing all AEs according to local protocols.

Serious adverse events

All SAEs during study participation were to be reported to the chief investigator (CI) within 24 hours of the site learning of their occurrence. The initial report could be made by telephone or fax. In the case of incomplete information at the time of initial reporting, all appropriate information was to be provided at follow-up as soon as it became available. The relationship of the SAE to study procedures was to be assessed by the investigator at site, as was the expected or unexpected nature of the AE.

Ethics consideration

Ethics approval was sought and granted by the Newcastle and North Tyneside 2 Committee, NHS Health Research Authority (reference 14/NE/0155).

Clinical trial assessments/data collection

Resource use

Resource use and patients’ out-of-pocket expenses during follow-up

The rates of response to the PCQ are presented in Table 12. Although the response rate is higher in the control arm than in the treatment arm, this comparison is not meaningful given the very small sample size. The loss to follow-up of two patients in the intervention arm was due to a staffing issue at one of the participating sites, which led to patients not being followed up by the research team.

In a definitive study, data collected on PCQs would be presented as average costs for each item. These costs would be estimated by multiplying the number of times each individual used a particular service by the number of individuals that gave that response, then summing them across individuals, which would then be divided by the total number of respondents contributing data (examples are given in the footnote of Table 13). However, for this study, the data (see Tables 13 and 14) are presented in their raw form for easier interpretation of responses. Furthermore, owing to the small sample size, any calculation of average costs would not be meaningful.

Table 13 summarises participants’ use of different health services collected from part A of the PCQ. The most common use of health services during the trial’s follow-up period appeared to be outpatient visits. Informal care was given by a relative or friend for the participant in the treatment arm. Planned inpatient stay and GP telephone consultation were also observed for one patient each. However, given the extremely small sample, no meaningful conclusion can be drawn.

Table 14 presents patients’ time and travel information on attending each type of health service recorded on part B of the PCQ. With the information recorded on part B of the PCQ, the average costs for an individual to attend each type of the health services could be calculated, which would then be combined with information collected on part A to derive a total costs for each individual’s out-of-pocket expenses.

There seem to be some inconsistencies between responses to parts A and B of the PCQ. Only one patient in the control group reported an inpatient stay in part A. However, in part B, a patient from the treatment arm also filled in their travel information for an inpatient stay. No GP/nurse visits were reported in part A, whereas two patients reported travel information for such visits in part B. This type of inconsistency may occasionally occur as a result of a reporting error or problems in the questionnaire design. One of the aims of a feasibility study is to examine and improve processes before conducting a definitive trial. When a larger sample is achieved, such inconsistencies are likely to have an insignificant effect on the results. In this study, we do not have an adequate sample size and the inconsistencies highlight how unreliable it would be to conduct analysis based on such data.

The aim of the PCQ used in the feasibility study was to gather information on the pattern of patients’ use of health services to help determine how the relevant data should be collected in the definitive trial. However, the small sample size did not allow this to be done.

Outcome measured by utility

The response rates of SF-36 are presented in Table 15. Two participants in the intervention arm were lost to follow-up because of a staff issue at one of the participating sites.

Table 16 presents the utility scores calculated from the SF-6D derived from responses to the SF-36. Average utility at each of the time points (baseline, 3 months and 6 months) and quality-adjusted life-year (QALY) gain over the 6 months period were presented. QALYs were calculated using the area under the curve method (1) with three time points of baseline, 3 months and 6 months. The area under the curve method calculates QALYs by multiplying the QoL with the duration of the life defined by the time points that form the curve. As with the case of resource use data, the sample size for calculating the utility score is too small to draw any meaningful conclusion or make any comparison between trial arms.

Objectives

To explore patients’:

• views on aspects of the informed consent process including method of approach and time to make a decision

• views on the content and amount of written and verbal trial information and their understanding of it, what they believe to be the purpose of the trial, what the two ‘treatment’ arms involve, what are the risks and benefits and what is required of them

• reasons why they declined/agreed to participate and under what circumstances would they participate/not participate

• experiences of participating in the trial, including their views on the arm they were randomised to, whether or not they underwent any further treatment and completion of the outcome measures (SF-36, STAI, FACT-G)

• reasons for not continuing in the trial until final follow-up.

To explore recruiting staff’s views on:

• participating in, and recruiting to, the trial

• the trial patient information

• cases in which they were unwilling or unhappy to recruit certain eligible patients to the trial.

In this section, we outline the findings of the qualitative parallel study: the barriers to, and facilitators of, the conduct of SURAB; specific facilitators raised by interviewees; interviewees’ views on the early closure of SURAB; and the key issues to be addressed in a future full trial. In Discussion, the key barriers to the trial are summarised.

Patients

When patients were approached to participate in the trial, they were also asked if they would be willing to be contacted by a member of the research team (JL) to discuss taking part in an interview. This invitation included those who declined to participate in the main SURAB trial. At that point, they gave consent to be contacted by a member of the research team about taking part in an interview. The names and contact numbers of patients were passed via the clinical trials unit to the qualitative researcher (JL). Patients were contacted by telephone to further explain the study, answer any questions they may have and set up a time to conduct the interview if they were happy to proceed. Consent to participate in an interview would be obtained at the time the interview was conducted.

Recruiting clinicians

At the study launch and during site initiation visits, clinicians and research nurses responsible for approaching and recruiting patients to the SURAB trial were informed that they would be asked to participate in an interview to explore their views around trial feasibility. A list of recruiting clinicians and research nurses was obtained from the NCTU that was managing the trial. At the end of the trial, the clinicians and research nurses from this list were approached via e-mail to request that they participate in a face-to-face (Newcastle-based staff only) or telephone interview and agree a date and time for the interview.

Data collection

Interviews were digitally recorded with the permission of the interviewees and transcribed verbatim. A topic guide for the interviews was developed from discussions with the wider team and from literature around trial participation. Although a topic guide was used, interviewees were encouraged to speak freely about any other issues relating to the feasibility study. The guide was revised as new issues emerged in consecutive interviews.

Analysis

Transcript data were managed using NVivo, version 10 (QSR International, Warrington, UK). A thematic framework was derived from the data through a process of data familiarisation to look for emergent themes. The framework was tested and refined and data were coded using the final framework. Data were analysed using a form of constant comparison method. 51

Approach to participate in the SURAB trial

The initial approach to take part in the trial was made by a consultant urologist. Those who agreed then met with a research nurse. All of the patients who were interviewed were happy with the approach to take part in the trial and, when applicable, the consent process. The timing of approach about the trial was said to be appropriate. Even considering the fact that the discovery of the tumour was incidental, no one thought it unreasonable to be asked about the trial at the time, or soon after, they received the diagnosis. They all felt they had been given sufficient time to make a decision; in some cases, the discussion about the trial had spanned three outpatient appointments. In a number of cases, the patient had a spouse or family member with them.

Interviewees were positive about both verbal and written trial information, finding both clear and easy to understand. They appeared to understand that they could withdraw from the trial at any point. The patient information booklet was thought to be helpful, and some had referred to it throughout the trial:

If you go to the middle section, it’s more or less a summary that explains it all. So now and again I have a look at that.

Pt03

When asked what they understood to be the purpose of the trial, no one mentioned it was to determine the feasibility of a full trial. Their focus tended to be on the treatment and management of the tumour with comments such as ‘no one knows the best way to treat it’. Only one interviewee misunderstood the purpose of the trial:

It’s to see if it’s frozen if it keeps growing.

Pt01

Randomisation is often said to be difficult for potential trial participants to comprehend, but all interviewees were clear that, if they agreed to take part, the group they were allocated to would be determined at random, by a ‘toss of a coin’. The trial was perceived to be of low risk and a few patients mentioned that the only risk they were aware of was related to the follow-up biopsy and a small chance of bleeding from the biopsy site. There was no perceived benefit, apart from the view that being part of a trial meant ‘better treatment’.

It did explain that there were two options, being left alone and monitored or the ablation and it probably wouldn’t benefit me as such but would benefit others in the future.

Pt02

Interviewees who agreed to the trial understood what was required of them in terms of treatment/management and follow-up visits related to the trial. There were no suggestions for changes or areas of improvement to the approach and consent process.

Decision to participate in the SURAB trial

All interviewees had discussed participation with family and, in one case, their GP. In most cases, their families were important in the decision-making process:

To be honest I was undecided at the time . . . one surgeon told me that I could live for ages with it and I could die with . . . the tumour because it was slow-growing which is encouraging . . . but my family they had spoken to people who had had the operation [ablation] and had been successful and they encouraged me to have the operation.

Pt04

One interviewee, based on a conversation with another patient who had undergone a total nephrectomy, decided that he would prefer ablation. He agreed to the trial and was randomised into the ablation group, but he said that he would have seen the trial through to the end had he been randomised to the active surveillance group. Another said he knew of people who had been on active surveillance for many years without any detrimental effect and this was a factor contributing to his decision to participate.

Attitudes to clinical research generally were explored and, although none of the interviewees had taken part in research previously, they were clearly positive about it and its importance. There was a perception that being part of a trial meant a slightly better standard of treatment:

My father had a heart thing and he took part in a trial and you know people have said if you take part in a trial you get . . . very good treatment . . . if I hadn’t had the [comorbid condition] I would have 100% gone down the trial road.

Pt02

For those who agreed to participate in the trial, the main reasons were altruistic, to help others in the future or because they thought highly of the consultant who approached them about the trial:

Unfortunately I lost my wife last year through cancer so . . . I thought that I should take part in this because if it helped me it would help other people in the future.

Pt02 surveillance

They say it’s for the future and I’ve got four great-grandkids, I’ve got five grandkids you know so if it helps somebody in the future, fair enough.

Pt01

This was often coupled with other factors such as the opportunity to have better treatment or a higher level of monitoring. Also mentioned were the fact that it was a low-risk trial and concerns about the alternative interventions such as surgery and risks of infection:

To be perfectly honest because I know at the end of the day they’re not going to say ‘Alright that’s it’ . . . [the information] explains that if you need any treatment of any sort after the study is finished you get it. It’s like Big Brother in a way.

Pt03

Reasons for declining to participate in the trial varied. One person, mentioned earlier, would have been happy to take part and be randomised but his family persuaded him to request ablation. Another person who had a comorbid condition that warranted continuous treatment thought that the additional visits would be too onerous:

I said . . . I would want to get rid of the cancer if possible rather than be monitored and be backwards and forwards.

Pt02

Only one person stated that they did not feel comfortable with the risk of being randomised to the active surveillance arm and that informed their decision to decline the trial. They were not happy to continue without treatment knowing that they had a cancerous tumour on their kidney and wanted to have the tumour removed immediately.

Views on the arm randomised

Of those who agreed to participate in the trial, none was disappointed with the arm they were randomised to. Prior to randomisation, they did not have a preferred arm and said that they would have seen the trial to its conclusion if they had been in the alternative arm. They were confident that they would be well cared for in either arm. One person also cited as a reason for agreeing the very slow growth of tumour mentioned in the trial information and the fact that the trial was voluntary and participants could withdraw at any point and undergo treatment if in the active surveillance arm.

Trial follow-up and outcome measures

The follow-up visits at 3 and 6 months after random allocation in the active surveillance arm and after treatment in the ablation arm were considered reasonable and not onerous. One interviewee commented that after agreeing to participate in the trial he had thought ‘Oh God, what have I let myself in for’, but, in effect, felt that it was an ‘easy’ trial to take part in and there was ‘not much involved’ (PT08).

Trial participants were required to complete three questionnaires at baseline – (1) a general health survey, (2) health states and QoL and (3) level of anxiety – and the same again at 3 and 6 months with the addition of a PCQ. Interviewees did not find these problematic, but one person commented that the level of anxiety measure was difficult to complete as they could not determine if the stress they were feeling was induced by their work situation or the presence of the tumour.

Response to request for interviews

Nine PIs and 16 research nurses (sometimes more than one per site) were e-mailed from a list provided by the NCTU (Table 18).

The site PI and research nurse in Southampton did not respond to requests to participate in an interview; therefore, their views on the feasibility of the SURAB trial in their site are unknown. In Oxford, the research nurse responded to say that the study had not progressed beyond the site initiation visit and, therefore, she felt unable to comment; no response was received from the PI. In Leeds, Glasgow and Stevenage, the research nurses did not respond to e-mails and, therefore, only the opinions of the PIs are represented. In University College London, the PI did not respond to e-mails and, therefore, only the views of the research nurses are represented.

A total of six PIs and five research nurses representing seven sites were interviewed.

Themes and subthemes

The aim of the parallel qualitative study was specifically to explore the barriers to, and facilitators of, the conduct of the SURAB trial. Interviewees’ opinions on the essential attributes of a future trial in terms of design and process were also explored.

What were barriers to the conduct of the trial for some recruiters were not barriers (or were even sometimes facilitators) for others. Therefore, as a constant comparative method of analysis was used, in which similar as well as contrasting ‘deviant’ cases are explored, the barriers and alternative views have been reported together. When interviewees have explicitly voiced what they considered to be facilitators in the conduct of the trial, these have been summarised in a separate section.

A number of themes, categorised broadly under barriers and facilitators, were identified from the interview data. Themes identified as barriers were linked to trial processes, trial design and ‘selling’ (a term used by a few of the interviewees) the trial. Subthemes were identified within trial processes, trial design and ‘selling’ the trial. The themes and subthemes for barriers to, facilitators of and features of a future trial are mapped in Figure 7. The arrows illustrate where the themes and subthemes relate to the key elements to address in the design and conduct of a future full trial.

FIGURE 7.

Themes and subthemes for barriers to, facilitators of and features of a future trial.

Barriers and facilitators: trial processes

Issues around delays or difficulties that had an impact on when recruitment started or the effectiveness of recruitment were raised. These were linked to processual factors such as completion of documentation at sites, sites and trial pathway and management issues linked to the clinical trials unit and trial team.

Barriers and facilitators: trial design

As might be expected, not all of those interviewed were completely happy with the design of the trial. Concerns around trial design related to issues initially with the trial protocol; inclusion and inclusion criteria, namely the need for biopsy and clinician views on whether or not the patient was suitable for the trial; and approach to participate, which includes identification and selection for the trial.

Patient’s clinician does not feel would be suitable for trial

Patients were excluded if their physician considered that they would not be a suitable candidate for the trial. On the eligibility form, the example reason given was ‘due to concomitant disease’. This was also interpreted by the sites recruiting to the SURAB trial as excluding those patients who, under non-trial conditions, would not be deemed suitable for active surveillance and ablation. This was the most common reason why patients screened for the trial were deemed ineligible (45/92; Figure 8), followed by the presence of comorbidities (21/92; see Figure 8). Twenty-seven out of the 119 ‘ineligible’ patients in the Consolidated Standards of Reporting Trials (CONSORT) diagram were actually not approached because the trial closed early.

FIGURE 8.

Reasons why patients were deemed ineligible for SURAB.

Despite the inclusion criterion of age ≥ 18 years, recruiting clinicians were not comfortable offering the trial to relatively young patients. One PI stated that these patients want ‘closure’; they either want to be discharged from hospital if the tumour is found to be benign or have it quickly removed and ‘they don’t tend to want to be followed up for a very long period of time’ (RS03). In the MDT meetings, it was often decided that it would not be best practice to put younger patients on active surveillance, and, therefore, they would not be offered the trial:

You say ‘Yes, it’s kidney cancer and we’re going to watch it’, that’s fine if you’re 68 or 78 because there’s a very real chance that it will not cause any problems during your lifetime . . . in a 40 year old . . . that’s going to catch up with you at some point and it’s going to need some treatment at some point . . . we’re going to have to do CTs every 6 months for the next 30 years. Is that realistic? No probably not. Is it ever going to go away? No. ‘Am I always going to have this cancer in my kidney doctor?’, ‘Yes’. Can you assure me that it isn’t going to spread?’, ‘No’.

RS09

As it was generally believed to be wrong to recruit younger or fitter patients to the trial and offer random allocation to ablation or active surveillance, the SURAB trial was felt to be a trial primarily for a much older population: ‘definitely a large cohort that we see who are 70 plus, those two options are fine’. However, some clinicians would prefer not to offer ablation to older, less fit people:

When I see an old person who has got co-morbidities and I’m thinking ‘surveillance’, I’m hoping I’m never going to have to treat them.

RS06

One interviewee explained that the patients who present with renal cancer typically fall into two main groups: elderly patients who would not be considered for treatment and would normally be offered active surveillance and people fit for surgery ‘who automatically wanted surgery’ (RS06). There was a view that, particularly in older patients with comorbidities, clinicians would hope to never have to treat them:

Were there any patients that you were unwilling or unhappy to recruit?

Yes . . . we had patients who were too old and frail we had patients who were too young and fit. We would offer typically the younger patients partial nephrectomy and the older patients we wouldn’t be happy to treat them at all.

Despite the mention of a large group of patients in their seventies who would be ideal for SURAB, most interviewees believed, at the time of agreeing to recruit to the trial, that they had underestimated the potential numbers who would be eligible:

If you look at the numbers of tumours that we see less than 4 cm, whatever centimetres, then yes, on paper you think ‘Well even if we pick up 25% of them we’re there. We’re laughing. But for each particular one, it’s quite a big sell for various reasons.

RS09

Although there was no shortage of patients with renal masses (three to five per week was the number quoted by one interviewee), very few (one per fortnight) were in the older age group that clinicians consider suitable for active surveillance. There was a recognition that, in reality, the pool of potential recruits is quite small, if the trial is deemed to be suitable only for older, fitter patients, and this was something not thought of before the trial began:

So we find that patients, if they’re fit enough, they want to have a treatment, if they’re not fit enough, they don’t want anything and the number that are in between that could be randomised is quite a small number.

RS03

Barriers and facilitators: ‘selling’ the trial

Four factors were identified that had an impact on recruitment to the trial. These were (1) patient route into the trial, (2) clinician as recruiter, (3) patient preferences and (4) trial intervention options.

Patient preference

It was stated to be a common occurrence for patients who attend the clinic to have a clear idea of what they want in terms of treatment or management of the tumour. A number of patients declined to take part in the trial as they already had decided on treatment. Out of 36 patients provided with study information, 26 declined as they had a preference for ablation (n = 16) or active surveillance (n = 10). In situations in which patients had decided on ablation, bringing in the possibility of active surveillance and then the existence of the SURAB trial was particularly difficult:

. . . if they’ve said ‘I want to have the radio-frequency ablation’ then you can mention ‘OK, well radio-frequency ablation is an option in these circumstance and another option . . .’ . . . you have got to introduce the option of observation and then you’ve got to jump from that to get them to agree to be randomised between the two. So it’s quite a big ask.

RS09

Some patients, even if they had not specifically decided on a particular treatment, wanted something done, which, again, made it difficult to convey the trial with an active surveillance arm as a viable option:

I’m not good at sales, apart from convincing people that I’m a surgeon they can trust and I think filling this trial, it was a difficult sell, . . . they come wanting treatment for their cancer . . . So I think the observation was a difficult sell.

RS04

Further linked to the fact that the SURAB trial was offering treatment versus no treatment, one interviewee said that patients clearly fell into one or the other group with few open to either option:

Even when you had the discussion . . . there are some people who definitely want to avoid any form of intervention and there are some people who definitely want something done no matter what you say to them . . . those people obviously weren’t consented into the main part of the trial.

RS10

One interviewee commented that the population are more informed, can access information via the internet and have a greater awareness of what treatment is available. The information provided as part of the trial gives ‘patients more information than they’ve ever had’ (RS02), and more than they receive routinely.

Intervention options

The intervention options were problematic for some interviewees as they believed there was an imbalance between the two arms of ablation and active surveillance. It was also believed to be difficult to ask patients to agree to be randomised when one of the arms was ‘to do nothing’:

One could imagine a trial where you are comparing two different types of ablation and you say ‘We don’t really know which of these two is the best’ . . . with this one you’re saying ‘Look, I don’t know whether you’re going to get just watched, or we’re going to treat you’, which is quite a big leap between those two things.

RS09

However, other interviewees felt that if the emphasis was placed on the fact that the active surveillance was only for 6 months after which time patients would return to normal care and have treatment if they wished, then the trial could potentially be more appealing:

I think had we pushed a bit harder that they were just going to be observed for 6 months and then after that they could come out of the trial and go for a treatment. I think that might have been something that would have been a bit more attractive for recruitment actually.

RS03

A number of the SURAB sites had participated in a previous trial: CONSERVE [A Feasibility Study for a Multicentre Randomised Controlled Trial to Compare Surgery With Needle Ablation Techniques in People With Small (4 cm) Renal Masses. Soomro N, Newcastle Hospitals NHS Foundation Trust, and Wood J, CTU Newcastle University, 2014]. 52 CONSERVE was mentioned by a number of interviewees in comparison with the SURAB trial in terms of its challenges. On one hand, CONSERVE was believed to be ‘more attractive to patients and maybe clinicians because at least you had two active elements’. On the other hand, it was mentioned as being difficult to recruit to because of the risks with partial nephrectomy; in the light of this, the SURAB trial was initially thought to be an easier trial because the eligible patients would be ‘an older population’ but:

. . . people either come with a preconceived idea . . . or when you tell them what the options are they really like that option. If they really like that option then why would you consider going into something that might not give you that option . . . but might give you an option that you’re really not desperately keen on.

RS09

The alternative view was that there was not a problem with the two arms of ablation and active surveillance but that the target group for such a trial would be older, less fit patients. The short period of active surveillance (6 months) was believed to be an attraction as it meant delaying treatment for a relatively short period:

I couldn’t understand why people were struggling because the two options of surveillance versus cryotherapy and particularly with only a six month mandatory window, it was very easy, I thought, to . . . offer people the two choices. If it was ‘You’re going to be under surveillance until we felt it was necessary to treat’ and that could be years later then I’m not sure patients would have been that easy to randomise.

RS06

There was also the opinion that patients are more interested in trials in which they have a chance of receiving a new intervention that is not available as part of standard care. It was questioned why patients would bother taking part in the trial when they could choose whichever option outside the trial:

What I’ve experienced over all my years in research is once something is available outside the setting of a clinical trial your uptake will be, it used to be about 20% agree to be on the study, I think it’s lower now.

RS02

Facilitators

The key facilitators in response to a question around ‘What worked well in SURAB?’ were patient route into the trial, systems of patient identification and approach and treatment options. It appeared to be ideal if the first contact eligible patients had about their condition was with the recruiting clinician. Interviewees mentioned the benefits of ‘the blank canvas’: patients who, at the point of the trial being discussed, do not have preconceived ideas about which management option they would prefer. To maximise the likelihood of patients not being advised about management by another urologist, a centralised referral system and a ‘totally consultant-led clinic’ with the consultant being the sole clinician recruiting to the trial appeared to be the answer. Linked to this was a keen research nurse and a system to ensure eligible and potential eligible patients do not slip through the net:

There was complete capture from MDT through to [research nurse] who would tell me, ‘Three patients are coming this afternoon, look out for them.’ I knew their names, I would then speak to them, [research nurse] would come and help and we’d have a chat and that was it, you know.

RS06

The other facilitator mentioned was that the SURAB trial was a simple trial with clear inclusion and exclusion criteria, and the treatment options, particularly with only a 6-month active surveillance arm, were attractive.

Treatment options

Interviewees were divided on the value of a trial of ablation versus active surveillance. Opinions ranged from the view that ‘it’s an incredibly important question that we need to answer’ (RS10) to the view that CONSERVE was a ‘much more valid question’ (RS03). The value to the clinical community of observing patients for only 6 months was also raised:

I’m still not sure what the point of that was, if you’re only going to watch them for 6 months, these things are so slow-growing what benefit are you going to really get to show that patients didn’t come to any harm by waiting 6 months? You know, you need to know they don’t come to any harm by waiting 10 years doing nothing.

RS03

However, one other interviewee commented that asking patients to take part in a trial in which they may be randomised to a longer period of observation may have made it more difficult to recruit patients.

It was also believed that, since the inception of the SURAB trial, things have moved on in the treatment of small cell kidney tumours, particularly with the adoption of robotic partial nephrectomy:

The treatment has evolved over time to really suit a different set of patients . . . So you’re almost trying to organise a study and you’re randomising people when we’ve already sort of already boxed them into three different sets of treatment . . . CP-cryo is a really good treatment for those patients that are maybe less suitable for having a proper op, and the reason we have surveillance programmes is . . . quite a lot of tumours in the elderly it’s a waste of time treating them. So when you then try and randomise them it’s bizarre because, you know, we’ve already boxed them into three different groups already.

RS03

A trial tailored to the patient was suggested, in which certain patients would be approached to be randomised to surgery or ablation and other patients to ablation or active surveillance. It was thought that this would be a more attractive option to patients. However, this may not address the issue raised by one interviewee that these trials are not novel enough to attract patients to participate when they can receive the trial options as part of usual care.

One interviewee raised the point that a key question that needs to be addressed is whether or not active surveillance is safe, and they quoted an Italian study under way exploring this. 53 They believe that, once this is determined, it may be easier to recruit patients to a trial of active treatment versus active surveillance.

Limitations

Although valuable and useful data have been obtained from recruiting staff who agreed to be interviewed, the response to the qualitative study was disappointing from some sites, particularly in the two sites at which neither the PI nor the research nurse agreed to be interviewed. We have the perspectives of both PI and research nurse in only three of the nine SURAB sites. Not having the views of any recruiting staff in two sites and only the PI or research nurse in four sites is a definite limitation of this qualitative study. However, we were able to capture the views of recruiting staff with a range of experience in relation to SURAB: from sites that failed to recruit any patients, from sites that managed to recruit a small number of patients and from sites that achieved their target recruitment rate.