Clinical evidence

Two meta-analyses summarised the results of randomised trials of the use of intensive follow-up after curative resection for colorectal cancer (Tjandra and Chan, 2007; Jeffery et al., 2007). A protocol for intensive follow-up was not defined because studies in the meta-analysis used different protocols. Thus, the results of the meta-analyses should be interpreted as evaluating the principle of intensive versus less intensive follow-up rather than the assessment of specific follow-up regimens as the included studies are heterogeneous in this regard (Table 5.1).

Table 5.1

GRADE profile: In asymptomatic patients who have undergone treatment with curative intent for colorectal cancer is intensive follow up more effective than less intensive or no follow-up.

There is moderate quality evidence of significant overall survival benefit at 5 years with intensive follow-up (Tjandra and Chan, 2007; Jeffery et al., 2007). Low quality evidence suggests that there is uncertainty as to whether more intensive follow-up confers a disease specific survival benefit when compared with less follow-up (Jeffery et al., 2007).

There is moderate quality evidence that the number of all recurrences detected is similar with both intensive and minimal follow-up (Jeffery et al., 2007 and Tjandra and Chan, 2007). There is low quality evidence that significantly more asymptomatic recurrences are detected in the intensively followed-up group.

The time to recurrence is significantly less with intensive follow-up but the evidence is of low quality (Jeffery et al., 2007; Tjandra and Chan, 2007). There is low quality evidence that the number of curative procedures attempted for recurrence is significantly more with intensive follow-up (Jeffery et al., 2007; Tjandra and Chan, 2007).

A single prospective comparative cohort study was identified during update searches (Laubert et al., 2010) which reported that 5-year overall survival was significantly better in the more intensively followed group versus the minimally followed group and the no follow-up group (p<0.001), though no statistically significant difference was observed in the rates of R0 resection of recurrent disease between the groups.

Economic evaluation

A systematic search of published cost-effectiveness studies was undertaken to inform this topic about follow up of patients with colorectal cancer who have undergone treatment with curative intent. Studies published prior to 1995 were excluded as they are unlikely to have relevance to current NHS practice and costs. The review identified six potentially relevant published economic evaluations (Borie et al., 2004; Hassan et al., 2010; Macafee et al., 2008; Michel et al., 1999; Norum and Olsen, 1997; Renehan et al., 2004). Following quality assessment, two of these studies (Borie et al., 2004; Michel et al., 1999) were deemed to have very serious limitations and were therefore excluded from further consideration. Two other studies (Norum and Olsen, 1997; Hassan et al., 2010) were also excluded as they were conducted in Norway and the USA respectively and were considered by the GDG to be less relevant for informing the cost effectiveness of follow up in the UK because of possible differences in clinical practice, costs and healthcare provision between countries. Therefore two studies (Macafee et al., 2008; Renehan et al., 2004) were included in the review of economic evidence. Both of the included studies were conducted from the perspective of the UK NHS, but differed in most other respects (Table 5.2).

Table 5.2

Modified GRADE profiles for included economic studies.

In Renehan et al. (2004) five randomised trials, each comparing a form of intensive follow up to conventional follow up, were meta-analysed to obtain estimates of health effects expressed in terms of life years gained. Details of the various follow up strategies and the frequency and type of surveillance tests from each trial were not reported in full in the reviewed publication. Costs of both follow up and treatment of recurrences were included in the analysis. Costs were based on the study-specific treatments and as these trials predated the routine use of adjuvant chemotherapy, cost of chemotherapy was not included. Across the five trials, the mean per patient cost of follow up in the intensive arm ranged from £3,388 to £6,509.

Macafee et al. (2008) compared an intensive follow-up regimen (based on one arm of the Follow Up after Colorectal Surgery [FACS] trial) with standard follow up (based on the principles of the British Society of Gastroenterology). Only hospital-based costs during follow up and the cost of surgically treating resectable recurrences were included in the analysis; costs of further elective operations for bowel continuity, chemo/radiotherapy and costs to primary care were not considered. The time horizon for the analysis was limited to 5 years and results were reported in terms of cost per additional resectable recurrence identified.

One additional relevant paper (Tappenden et al., 2009) was identified during the search. This paper was itself a systematic review of UK economic evaluations of colorectal cancer interventions and identified the same individual studies (Macafee et al., 2008; Renehan et al., 2004) related to the topic of follow up that have been included in the current review.

Linking evidence to recommendations

Overall survival was the most consistently reported outcome. Survival with good quality of life was considered the endpoint of most importance to patients and health professionals. All outcomes were considered useful but evidence was limited for some outcomes (for example quality of life and late effects of treatment).

The GDG assessed the benefits of intensive follow-up versus less intensive and found evidence that improved survival is associated with more intensive follow-up. However there was variability in the components and frequency of the different intensive protocols in the evidence.

The GDG considered detection of recurrence to be a critical goal in follow-up because doing so would enable some patients to be cured. Whilst CEA will not detect all recurrence, the GDG considered that its use would be beneficial in achieving this goal. The GDG were also concerned that the use of CT scans and CEA tests for follow-up was inconsistent across cancer networks, and thus, a minimum standard of care had to be recommended based on clinical experience and the need to promote patient confidence. The GDG were not able to recommend one specific protocol from the evidence. Instead they elected to recommend a pragmatic protocol of follow-up.

The review of clinical and cost-effectiveness literature shows that there is no consistent definition of what constitutes intensive follow up for colorectal cancer patients. The various studies included in this review differ in terms of the types of tests and interventions included and the frequency of surveillance, therefore no single recommendation for a specific protocol for intensive follow up can be recommended. Caution should therefore also be exercised when pooling studies or making generalisations about both the effectiveness and cost effectiveness of different protocols for intensive follow up over conventional (or less intensive) follow up.

The GDG also assessed the potential harms, namely increased patient anxieties from intensive testing. The number of studies of this endpoint were few but there was no strong evidence that the intervention of intensive follow-up is associated with increased anxieties across a wide range of patients.

The GDG chose not to recommend further research on oncological outcomes related to follow-up because they were aware that relevant trials were already underway.

The use of intensive follow-up may incur an increased cost on resources, particularly imaging. The trade-off is improved survival and probable improvement in quality of life.

The overall quality of evidence on was assessed as low to moderate by GRADE methodology.

Neither of the cost-effectiveness studies included in the economic evidence review reported an incremental cost-effectiveness ratio (ICER) in terms of cost per QALY. In the absence of information about what represents a reasonable cost per additional resectable case identified, it is difficult to interpret the results of the Macafee et al. (2008) analysis and therefore this study has limited relevance for informing the current Guideline topic. The results of Renehan et al. (2004), although expressed in terms of cost per life year gained, suggest that intensive follow up is cost effective when compared to conventional follow up. There is some uncertainty about the impact that quality adjustment of survival would have on the ICER reported in Renehan et al. (2004), but it is unlikely to change the main conclusion of the paper.

Clinical evidence

There were a small number of studies directly investigating the information needs of patients with colorectal cancer (Nikoletti et al., 2008; Lynch et al., 2008; Persson et al., 2005; Broughton et al., 2004; Kerr et al., 2003; Sahay et al., 2000). All included studies employed qualitative methodology to assess and investigate patient information needs and included studies investigating the population of interest (colorectal cancer patients); few included studies identified specific inclusion or exclusion criteria with the majority specifying only that patients were colorectal cancer patients with the ability to understand/read the language in which the study was being conducted. There was one study conducted in the UK which included not only colorectal cancer patients but their carers too (Broughton et al., 2004).

The number of patients in each study ranged from 20 (Sahay et al., 2000) to 1,966 (Lynch et al., 2008) and all studies included patients treated for colorectal cancer with few specific restrictions to inclusion.

The included studies may be at risk from recall error due to the differing points in the treatment pathway at which each participant took part in a study. Studies may also be at risk from selection bias with response rates from 5 studies ranging from 32–86% (Nikoletti et al., 2008; Lynch et al., 2008; Persson et al., 2005; Broughton et al., 2004; Kerr et al., 2003).

Included studies addressed factors such as the specific information requirements of participants, the source of information and modes of delivery, the timing of information provision and the impact of information provision on wellbeing and quality of life. There appeared to be a high degree of dissatisfaction with information provided on specific areas across the studies, particularly related to bowel function. In one study more than 50% of patients were not happy with the information provided in relation to bloating, wind/gas, difficulties emptying bowels, medication, the use of pads and other unspecified bowel problems (Nikoletti et al., 2008). In one study 59% of responders reported not being instructed in stoma irrigation techniques and more than 80% of respondents were dissatisfied with information received during chemotherapy and radiotherapy (Kerr et al., 2003).

The desired source of information and modes of deliveries varied across studies although common themes did appear with doctors, specialist incontinence advisors, nurses, surgeons and relatives all identified as possible sources of information. Modes of delivery included one to one teaching by a health professional, leaflets, pamphlets/booklets, discussion groups, and internet.

The timing of information provision was addressed in two studies (Broughton et al., 2004; Nikoletti et al., 2008). The best time for the provision of information was considered to be either before surgery (32.9%) or after surgery while still in hospital (37.2%) (Nikoletti et al., 2008). Carers appreciated the time spent when specialist nurses provided information and several patients and carers would have appreciated more information when being discharged, in particular relating to what symptoms were considered normal after bowel surgery (Broughton et al., 2004).

From one study, bivariate analysis indicated a poorer quality of life was associated with communication problems for men and younger patients, though on multivariate analysis, controlled for clinical and demographic differences, no interaction was observed between communication and gender or age. For patients that completed the questionnaire over 3 years, differences in quality of life between clear and unclear communications groups remained. The difference was statistically significant for emotional (p<0.02) and social functioning (p<0.05) and for sleep problems (p<0.02) (Kerr et al., 2003).

Two studies which considered patient perspective were identified on update searches (Beaver et al., 2010; O’Connor et al., 2010).

From Beaver et al. (2010) is was reported that although patients saw a nurse specialist while they were a hospital inpatient, they were unsure of what to expect once the returned home; this was particularly true of patients without a stoma as they did not usually receive a visit from the nurse specialist once discharged home. Patients also reported that doctors did not address their concerns or provide information at follow-up appointments and this left them feeling uncertain about their condition and what to expect. This was again particularly true of patients without a stoma.

Patients without a stoma reported more feelings of isolation, though this was not limited solely to this group of participants. There appeared to be an expectation from patients that the nurse specialist would visit them at home following discharge and a feeling of disappointment when this was not the case.

Patients with a stoma frequently commented that they learned about stoma care through ‘trial and error’ as they felt that follow-up care did not provide sufficient information on provision of stoma bags and care (Beaver et al., 2010).

Patients experiencing nurse led follow-up reported favourably on their outpatient experience in terms of information, support, knowing what to expect and what was ‘normal’ in their situation.

Written information was considered beneficial, particularly diagrams nurses drew for each patient, tailored to their own surgical procedure and pitched at their own level of understanding. Leaflets were perceived to be helpful, providing useful future points of referral.

O’Connor et al. (2010) reported that males felt it was more important to know where their family could go to get help with dealing with their illness. The study also reported statistically significantly higher satisfaction levels with information on where family could get help dealing with the patient’s illness, whether they could wear normal clothing, how treatment works against cancer, if they were going to need help taking care of themselves and how to prepare for the investigative tests.

Younger patients expressed significantly higher information needs regarding changes in the things they can do with and for their family, who to talk to about alternative therapies, where the family could go to get help dealing with the patient’s illness, if treatment would alter the way they looked, what type treatments are available, how to prepare for the tests, what to do if they felt uncomfortable in social situations, if the illness was hereditary, if treatment would affect their relationship or sex life and if they could continue with their job after surgery and treatment. Older patients expressed higher information needs only in knowing who to call if they had questions while still undergoing treatment.

No significant difference in information needs or how these needs were met were observed in relation to length of time since diagnosis, type of treatment and whether or not a patient had a stoma. Comparison of perceptions of the importance of items of information with perceptions of how these needs were met showed a statistically significant difference, indicating that patients felt that information needs with ratings of a high level of importance were not adequately addressed (O’Connor et al., 2010).

Stoma care nurse specialists were reported to be the most common source of information, with other healthcare professionals such as ward nurses, chemotherapy nurses, colorectal consultants and GP mentioned. One patient cited the internet as the preferred source of information. Interpersonal communication with a healthcare provider was cited as the most common and preferred source of information (O’Connor et al., 2010).

Linking evidence to recommendations

The GDG looked for evidence primarily concentrating on the patient perception of information needs, not needs reported by health professionals. There was surprisingly little evidence with only a small number of studies, all of which were qualitative in design and with heterogeneous inclusion and exclusion criteria. Older patients and those with more severe disease were under represented, they may have different needs but the evidence is not clear.

The GDG looked at what information was useful, who should deliver it, when and in what format. The impact of information delivery and quality of life outcomes were also examined. The recommendations are based on the available evidence and the expertise of the GDG. Patients’ lives are profoundly altered by the diagnosis and treatment of their bowel cancer. There was a strong agreement in the GDG that more needs to be known from patients about their own information requirements.