Article Text

Optimising colorectal cancer screening acceptance: a review
1. Carlo Senore1,
3. Nereo Segnan1,
4. Cristina Bellisario1,
5. Cesare Hassan3
1. 1Centro di Prevenzione Oncologica (CPO Piemonte), AOU Città della Salute e della Scienza, Turin, Italy
2. 2Digestive Disease Center, University of Washington, Seattle, Washington, USA
3. 3Unit of Gastroenterology, Ospedale Nuovo Regina Margherita, Rome, Italy
1. Correspondence to Dr Carlo Senore, AOU Città della Salute e della Scienza, Centro di Prevenzione Oncologica, Via S Francesco da Paola 31, Turin 10123, Italy; carlo.senore{at}cpo.it

## Abstract

The study aims to review available evidence concerning effective interventions to increase colorectal cancer (CRC) screening acceptance. We performed a literature search of randomised trials designed to increase individuals' use of CRC screening on PubMed, Embase, Cochrane Database of Systematic Reviews and Database of Abstracts of Reviews of Effects. Small (≤100 subjects per arm) studies and those reporting results of interventions implemented before publication of the large faecal occult blood test trials were excluded. Interventions were categorised following the Continuum of Cancer Care and the PRECEDE–PROCEED models and studies were grouped by screening model (opportunistic vs organised). Multifactor interventions targeting multiple levels of care and considering factors outside the individual clinician control, represent the most effective strategy to enhance CRC screening acceptance. Removing financial barriers, implementing methods allowing a systematic contact of the whole target population, using personal invitation letters, preferably signed by the reference care provider, and reminders mailed to all non-attendees are highly effective in enhancing CRC screening acceptance. Physician reminders may support the diffusion of screening, but they can be effective only for individuals who have access to and make use of healthcare services. Educational interventions for patients and providers are effective, but the implementation of organisational measures may be necessary to favour their impact. Available evidence indicates that organised programmes allow to achieve an extensive coverage and to enhance equity of access, while maximising the health impact of screening. They provide at the same time an infrastructure allowing to achieve a more favourable cost-effectiveness profile of potentially effective strategies, which would not be sustainable in opportunistic settings.

• COLORECTAL CANCER SCREENING
• COST-EFFECTIVENESS
View Full Text

## Statistics from Altmetric.com

### Key messages

• Substantial differences in colorectal cancer (CRC) screening use may result in higher disease burden, lower quality of life, health inequities and increased healthcare cost. Non-adherence to recommended protocols represents an important attributable factor in CRC mortality, particularly among lower socio-economic status populations.

• Available evidence indicates that multifactor interventions targeting multiple levels and considering factors outside the individual clinician control represent the most effective strategy to enhance CRC screening acceptance, supporting the WHO recommendation to implement population-based organised programmes.

• Removing financial barriers, offering all subjects in the target population the option to adhere to high-quality, screening interventions, using personal invitation letters, preferably signed by the reference primary care practitioner, and reminders mailed to all non-attendees, represent a priority for interventions aimed to enhance acceptance of CRC screening.

• Physician reminders may represent an additional support to the diffusion of screening in areas with low baseline uptake rates, but in any case they can be effective only for individuals who have access to and make use of healthcare services.

• Educational interventions for patients and providers are effective, but the implementation of organisational measures may be necessary to favour their impact. Integrating qualitative research results in the design of educational interventions represents a challenge for future research.

## Introduction

The uptake rate represents a critical determinant of the magnitude of the health impact of colorectal cancer (CRC) screening at the population level. Non-adherence to recommended protocols represents an important attributable factor of CRC mortality,1 particularly among deprived groups. However, in spite of the strong evidence2 ,3 supporting the screening effectiveness and of a general consensus of guidelines recommending screening of average-risk individuals over age 50,3 ,4 CRC screening rates remain low. Inequalities in screening use represent an important component of the wide variability observed within and across countries5–9 and may result in higher disease burden, lower quality of life, increased healthcare costs and health inequities.

Screening is a complex process of care consisting of several steps and interfaces between patients, providers and health organisations, which may be promoted, or impeded, at different levels by a range of factors. An extensive literature exists concerning predictors of uptake and the effect of strategies promoting attendance. However, the variability in the screening protocols and the mix of programmatic and non-programmatic delivery systems are limiting the comparability and the transferability of the results.

The framework of the quality in the continuum of cancer care model10–12 provides a systematic approach for assessing factors that influence delivery, access and quality of screening. Recognising that cancer care delivery occurs in a multilevel environment (figure 1), this model highlights those levels of the healthcare system representing potential targets for intervention, including policy, practice, provider and patient level. It may therefore offer useful clues to identify potential targets of interventions, to assess their potential reach, as well as the determinants of their success or failure, and to identify directions for further research.

Figure 1

Levels of intervention related to cancer screening interventions in health service settings.

Several factors interact in a synergistic manner at these different levels of care to affect provider delivery and patients' use of screening. The PRECEDE–PROCEED model,13 explaining behavioural change as the result of the interplay of predisposing, enabling and reinforcing factors, represents a useful tool to assess the mechanisms through which these factors can affect screening behaviours.

The success of a behavioural intervention requires the identification of the level that must be influenced to achieve an impact and of the optimal strategy to address the relevant factors, which can favour behavioural change. Underuse of CRC screening among uninsured groups can be influenced, for example, by availability of free-of-charge screening programmes funded by national or regional governments, while participation among insured subjects may be enabled by mailing of the faecal occult blood test (FOBT) kit, at the organisational level.

These theoretical frameworks (box 1) will be used to examine existing evidence of effectiveness of interventions aimed to enhance use of CRC screening, identifying the level of care and the behavioural mechanisms targeted by each intervention. Taking into account the public health perspective, we also assessed evidence concerning cost-effectiveness of successful interventions (box 2).

Box 1

### Conceptual framework

The PRECEDE–PROCEED model13 explains behavioural change as the result of the interplay of predisposing factors, providing the rationale, or the motivation, for the behaviour, of factors enabling actual realisation of the decision, and of factors reinforcing individuals' decision to adopt the desired behaviour and supporting its maintenance over time.

Interventions that focus on patient predisposing factors attempt to increase awareness and knowledge of colorectal cancer (CRC) and CRC screening, identify and address health beliefs, or values and fears that impede cancer screening. Interventions focused on patient enabling factors address costs of screening, diagnostic testing and cancer treatment, access to CRC screening and barriers impacting the acceptability of CRC screening. Patient reinforcement interventions may include reminders to complete screening or provide encouragement through peer or professional counselling.

Interventions that focus on provider predisposing factors increase knowledge and enhance agreement with clinical practice guidelines for CRC screening, or dispel assumptions about subjects' health beliefs and likelihood of screening completion. Provider interventions addressing enabling factors include the provision of sufficient time to determine CRC screening eligibility and needs, computer or manual reminders to recommend screening, or sufficient support staff and technology. Interventions of provider reinforcing factors may include the institution of payor requirements, leadership expectations, collegial norms and performance reports.

Box 2

### Directions for future research

As long as maintaining subjects' engagement over several screening rounds following initial attendance is required to ensure screening effectiveness, additional research is needed to identify interventions that can enhance sustained attendance.

Qualitative research can offer insights on barriers to screening acceptance, which can be used to tailor educational messages. Integrating findings from qualitative studies into the design of interventions that can be sustainable in the context of large-scale population programmes represents an important aim to be addressed in future studies.

Developing valid measures to evaluate the impact of specific interventions on the adoption of decisions consistent with individuals' values and preferences represents a challenge to be addressed in future studies.

## Methods

### Literature search

We searched in first instance recent systematic reviews (SR) and meta-analyses addressing the same topics of our study on PubMed, Embase, Cochrane Database of Systematic Reviews and Database of Abstracts of Reviews of Effects. The search was based on the strategy used in the SR by Jepson et al15 with appropriate changes according to our inclusion criteria and databases. Relevant original articles published before August 2012 (search update of the most recent review16) were retrieved through already published reviews. We subsequently searched PubMed, Embase and CENTRAL for randomised controlled trials (RCTs), reporting comparative evaluations of interventions, published until 15 October 2014. See online supplementary appendix 1 for the details of the search strategies.

We excluded small-size trials (≤100 subjects per arm) and those reports preceding the publication of the European population-based trials17 ,18 documenting the effectiveness of guaiac FOBT (gFOBT) screening. The findings concerning the impact of interventions conducted when evidence of screening effectiveness was still lacking might not be comparable to those reported from more recent trials, implemented following endorsement of CRC screening by professional societies and governmental institutions. Observational studies were considered when assessing policy options, not easily amenable to testing by RCTs.

The outcome of interest in the context of organised programmes was the participation in the first-level screening test, defined as the proportion of invited subjects who underwent the test. In opportunistic settings, the outcome considered was test coverage, defined as the proportion of the target population who had a test within the recommended interval. Interventions have been categorised according to the targeted level of the healthcare system, taking into account the specific behavioural mechanism involved and the screening model (opportunistic vs organised).

## Results

The SR and meta-analyses literature search yielded 112 papers. These were screened for relevance based on titles and abstracts and 21 of them were judged potentially relevant. Full text could not be acquired for one review19 and 17 of the remaining 20 were included: 716 ,20–24 considered interventions addressing several levels of care, 610 ,25–29 were focused on provider, or organisation level, 330–32 on screening modality and 133 on community level.

The RCTs literature search yielded 170 papers. These were screened for relevance based on titles and abstracts: 67 were obtained in full text and 44 were included.

Reasons for exclusion are reported in online supplementary appendix 1. PRISMA flow charts are reported in online supplementary appendix 1.

### Levels targeted by interventions

#### Policy

Two main policy options define the context within which screening is currently provided. Screening tests may be recommended by primary care practitioners (PCPs) during routine medical consultations for unrelated conditions, or on the basis of a possible increased CRC risk (family history or other risk factors, including symptoms), or as result of a patient's request. This approach is defined as opportunistic screening, as opposed to organised screening, where all eligible subjects are actively invited, following an explicit and prespecified protocol stipulating testing and assessment procedures, offered free of charge, within an established organisational context ensuring systematic monitoring of the quality of the whole process.34

CRC screening rates in opportunistic settings differ by type of insurance and are influenced by the amount of patient copayment.35 Individuals without access to primary care are excluded from participation in those settings where most CRC screening relies on office-based interventions delivered by PCP36 and insurance status represents the most important determinant of screening coverage in the USA.37

Policy measures characterising the delivery of screening within organised programmes, including the reduction, or elimination, of costs for the participant and the implementation of a system of patients' reminders, were identified as the most effective interventions aimed to enhance preventive services use in a comparative analysis of the relative impact of different strategies.38 Higher screening rates, with reduced inequalities across social groups, have been reported in France and Italy, in areas where organised CRC screening programmes had been introduced, as compared to areas where only opportunistic screening was available.39 ,40 Recent observational studies showed that the introduction of organised screening is associated with a substantial reduction in CRC mortality among populations served by the programmes, as compared with people not routinely invited.41–43

#### Practice organisation

Service delivery: According to the results of a recent review,38 establishing separate clinics devoted to screening, involving nursing or clerical staff in the delivery of services and adopting monitoring and quality improvement approaches can have a strong positive impact on screening use. These intervention components characterise the infrastructure of screening delivery within organised programmes.

A positive impact of these measures was observed also in opportunistic settings. Staff delivered interventions, designed to identify subjects eligible for screening and to offer counselling about FOBT testing,44 or assistance to fix a TC appointment,45 were associated with an increase in the rates of FOBT and TC screening, in primary care clinics. A survey of family medicine practices reported that those using nurses, or health educators, to provide behavioural counselling were also more likely to show higher screening rates than those not involving staff in counselling activity.46 The introduction of a monitored institutional directive was associated with a significant increase in the screening rates in a managed care setting.47

Invitation: The offer of a pre-fixed appointment in the invitation letter represents an effective strategy to enhance uptake,15 currently adopted in sigmoidoscopy (FS)-based programmes,48 although tested in RCTs only among women invited for breast and cervical cancer screening. Compliance with FS screening was improved by timing invitations in proximity to annual milestones, or invitees' birthday,49 while having a nurse calling invitees to schedule an appointment was not more effective than offering an open appointment.50

Test delivery (table 1): organised programmes: Mailing of the test kit with instructions, together with the invitation letter, has been adopted in several organised programmes to maximise accessibility. Subjects receiving an invitation including an immunochemical FOBT (faecal immunochemical test (FIT)) kit showed a twofold increase in the likelihood to participate, as compared to subjects receiving an invitation letter with the indication to pick up the kit by their general practitioner (GP), in a population-based programme in Belgium,52 and a more modest, though statistically significant, increase in attendance (30% vs 28%), compared to those receiving the indication to pick up the kit at a pharmacy in an Italian pilot screening study.53 Mailing of FIT kit, compared with the invitation to pick up the kit at a pre-fixed date at a primary clinic, was associated with a statistically significant increase in the screening rates among attendees in previous screening rounds of two Italian population programmes (overall response: 63.0% vs 56.8%).55 Mailing of the gFOBT test kit, as opposed to a simple mail reminder, was associated with an increase from 9.6% to 20.1% in the screening rates among previous non-responders, in a population-based study in Canada,51 while a similar increase was observed only in one centre in the Italian FIT-based programmes.55

Table 1

Interventions targeting organisation of practice

Delivery of the kit by the GP, or by non-medical staff, who may provide counselling, enhances acceptance. Completion of FIT and gFOBT was higher among subjects required to pick up the kit at the GP's office than among those referred to the local hospital.54 Screening activity reports indicated that involvement of trained community volunteers,62 or of pharmacies,63 in the kit distribution is associated with high participation rates, although these options were not tested in comparative trials.

Opportunistic setting: Mailing of gFOBT kit, resulting in a modest (4%), but statistically significant, absolute increase in participation of primary care patients, compared with mailing of a kit request card.61 Mailing of the kit emerged as the most effective component of interventions using reminders to enhance screening rates among subjects overdue for screening:57–59 compared with mailing of the kit alone, no additional effect could be observed when adding an educational brochure,58 ,61 a brochure followed by a follow-up phone call,59 or a web-based educational tool.57 A larger effect on FIT uptake has been observed in studies where mailing of kit was associated with multifaceted outreach interventions, providing multilingual, or low-literacy, educational brochures and using reminder phone calls, both automated and performed by screening navigators.45 ,56

The potential positive effect of face-to-face interactions is supported by the results of an RCT among primary clinic attendees, showing a higher FIT uptake when a trained non-health professional delivered the kit and collected the sample from participants' home, as opposed to direct mailing of the kit.60

Screening modality (table 2): Multiple tests70 ,81 ,82 have been validated for CRC screening, showing a different effectiveness, acceptability, safety and cost profile.

Table 2

Comparative trials of different CRC screening strategies

The findings of a recent survey among women who had never participated in CRC, cervical or breast cancer screening programmes83 suggest that although more global barriers, such as cancer fatalism, may explain non-participation in more than one programme, the uptake of CRC screening might be influenced by specific characteristics of recommended options as dislike of the test appears to be a stronger barrier to CRC screening. Differently from clinical practice, where decision-making process mostly depends on test accuracy, acceptance represents a critical factor to orient the choice in screening programmes and in the past 20 years several studies compared the relative uptake of competitive strategies.30 ,31

Organised programmes with FOBTs have been implemented in several countries,84 with a trend in most recent years favouring the adoption of FIT, as opposed to gFOBT. Despite the choice between the two tests depending on several organisational and economic factors, uptake and effectiveness represent key elements in the decision-making process. The uptake was substantially higher among people invited for FIT than among those offered gFOBT, in two large population-based RCTs in the Netherlands71 ,85 and in a previous pilot population-based intervention in Italy.67 A large study conducted in the context of an established population programme has confirmed the findings of these pilot RCTs, showing that the adoption of FIT is associated with an increased uptake over gFOBT and that it may also result in a reduction in the participation gap for age, gender and deprivation.79

Screening rates were increased when using FIT as compared to gFOBT also in a US study in a clinic setting,77 but not in an RCT in primary care setting in Israel.78 However, in this latter study, three stool samples—rather than one, as in all previous studies—were required. A recent RCT in a managed care setting showed a lower adherence among people invited to perform a two-sample FIT, or a three-sample gFOBT, as compared to those allocated to a single-sample FIT regimen80 and indeed, the higher compliance achieved when using FIT instead of gFOBT has been explained by the lack of dietary restrictions, the easier and less unpleasant sampling methods and the lower number of samples.54

Inconsistent findings have been reported when comparing FS and faecal tests. Two large Italian population-based RCTs55 ,86 reported a similar uptake with FIT as with FS, while the attendance rate was substantially higher both with FIT and with gFOBT than with FS in a Dutch population RCT. Compliance was also significantly lower with screening FS than with gFOBT in two average risk population cohorts, in Italy and Sweden,65 ,68 while another study, conducted in a single UK practice, where the GP was performing FS, reported a higher uptake with FS than with gFOBT.76 FS would appear to be particularly penalised in settings with a high FIT, or gFOBT, adherence, while a low CRC screening uptake would marginalise such difference.

Different combinations of FS and faecal tests have been tested to improve screening efficacy or uptake (table 3). The addition of FS to gFOBT was associated with a substantial decrease in the uptake, as compared to gFOBT alone, in three population-based RCTs.65 ,66 ,76 A similar trend was observed when adding FS to FIT in a small Australian population RCT69 and in a recent population-based RCT in Norway.73

Table 3

Patient level interventions—organised programmes

Offering people the option to choose between FS and FIT did not improve uptake as compared to an invitation for one of the two tests.55 ,69 Offering FIT to those who refuse FS screening represents an effective alternative approach: the sequential offer of FIT retrieved 19% of those declining FS invitation in an Italian population-based programme and similar findings have been reported in a Dutch pilot screening study.72

The use of colonoscopy (TC) as a primary screening tool has also been advocated, despite the lack of experimental evidence of efficacy. The uptake was lower with TC than with FIT in two large population-based RCTs, in Italy86 and Spain,74 and in a small Australian population study.69

The uptake was higher when offering screening with CT colonography (CTC) than with TC in a Dutch population-based RCT.75 Two studies are ongoing to compare the uptake of CTC and of FS or FIT.98 ,99

#### Provider

The involvement of GPs was shown to be highly effective in improving compliance, both in the context of organised100 ,101 and opportunistic screening settings;102 ,103 prompting subjects invited to seek advice might facilitate GPs' counselling,104 which may be particularly important for less educated, or older, people, less likely to use mailed information material.105

GPs may need adequate information, focused on the programme effectiveness on the accuracy of the adopted test, in order to effectively promote screening,106 and the implementation of organisational measures may be necessary to optimise the impact of such educational efforts.38

Reminder systems have been shown to increase the use of preventive services,15 ,38 ,107 computerised systems being somewhat more effective than chart-based manual reminders.

In opportunistic screening settings, the impact of reminders is higher for tests that can be directly delivered by the physician: PCPs referring their patients to other providers were less likely to recommend FS.108 Also, a reminder note to the physician to direct his patients to perform an FOBT was more effective than educational seminars offered to physicians and as affective as a phone reminder for the patients.109

Giving feedback to providers about their screening rates and their relative performance may reinforce their commitment to promote FOBT screening,27 although it was suggested38 that organisational changes may have a larger impact and that provider feedback should be part of multifactor interventions. In the context of organised programmes, the computerised information systems implemented to manage and monitor screening activity can ensure timely feedback to PCPs, as well as to other professionals involved in the process.

A recent review10 concluded that incentives may be more likely to reinforce provider referral, if targeted to individual physicians, if they are large enough to be noticed and if they do not require referral to other providers/facilities (as it is the case for endoscopic screening). A recent analysis of time trends in FOBT screening rates in Canada would suggest that the introduction of financial incentives for PCP resulted in an increase in population coverage.110 Reimbursement policies may affect providers’ behaviour: the low FS screening rates and the parallel increase in the TC screening rates in the USA are likely explained by the low clinician reimbursement rates for FS and by the coverage of TC by the public payer.111

Providers' characteristics may influence acceptability of the test: the reported trend towards a lower participation of women in FS screening55 ,73 ,85 is likely related to women's preference for having the exam performed by a female endoscopist and in fact no gender difference was observed in a pilot English study of nurse-delivered FS screening.112

#### Patient

Organised programmes (table 3): Most programmes provide written information in the form of leaflets to people invited for screening. Information leaflets represent a valuable tool in addition to the invitation letter:97 ,113 ,114 a recent trial conducted within the UK screening programme showed that an enhanced procedural leaflet was associated with a 6% absolute increase in CRC screening uptake, which was maintained when combined with a screening endorsement letter signed by the GP.92

Evidence concerning the incremental value of tailored leaflets in this context is inconsistent. A brochure specifically designed to address known barriers to FOBT screening was not associated with an increase in the uptake in the English screening programme,90 while mailing of a psycho-educational booklet to subjects indicating low interest in having FS screening was associated with a moderate increase (from 50% to 54%) in the attendance rate.94 The format of the information material may influence subjects' decision: using visual instruments to enhance appeal and clarity may enhance uptake.115

An advance notification letter was associated with an increased participation in gFOBT and FIT screening,88 ,89 ,93 as well as in FS screening,87 although the size of the effect was not consistent across studies.

Evidence concerning effectiveness of mass media campaigns and community-based interventions is inconclusive,20 although some reports indicated that they might have a positive impact,110 especially among previous non-responders.116

Reminders enable individuals to plan attendance to their screening appointment. Available evidence29 indicates that several kinds of reminders, including letters, postcards, telephone, auto-dialler or provider delivered face-to-face reminder, are effective, with telephone reminders being most effective, but also most costly.24

Adding a collection paper to the mailed FIT kit, to reduce unpleasantness of sampling, did not result in an increase in the participation in a pilot population programme, as compared to standard mailing.91

Several studies showed a positive impact on screening uptake of a letter signed by the GP, reinforcing invitees' commitment to attend, compared to letters from different authority sources.48 ,67 ,92 ,95 Support by other trusted PCP may also be effective.117 Organisational barriers, related to the need to plan an additional visit, may limit, however, the impact of GPs' availability for personal encounters: subjects offered a prior consultation with their GP showed a lower response to an FS invitation than those receiving an invitation letter with leaflet.96

Married adults are more likely to participate in CRC screening than non-married, and inviting both members of a couple together significantly increased attendance in both genders in the UK FS pilot screening study.118 Financial incentives to subjects invited for screening were not found to have an impact on participation.15

Opportunistic setting (table 4): Several educational interventions were tested among people attending routine office visits, including the use of decision aids and personalised risk assessment tools,128 ,120 telephone assistance to encourage screening, delivered by trained staff,124 ,129 ,134 or through interactive voice response (IVR) systems,121 ,131 tailored print, or web-based, communication, or literacy appropriate pamphlets,123 ,127 ,130 ,133 ,135 ,136 nurse support, including personal encounter and follow-up phone call,122 ,119 tailored navigation126 and videos.137 ,138 Telephone support conveying educational messages and providing motivational counselling was generally associated with an increase in the proportion of subjects performing any test within the 12-month interval following the intervention, also among uninsured groups. However, these interventions are more labour-intensive than mail reminders and their reach may be limited: only 60% of the targeted subjects could be contacted in one study among uninsured women.139 Tailored navigation offered by nurses, or trained professionals, was also associated with increased uptake in most studies, although peer and professional navigators were not more effective than reminder calls in a clinic setting.125

Table 4

Patient level interventions—opportunistic setting

Mailing or delivery via the web of tailored printed educational material, or the use of decision aids, did not show an effect on screening uptake, compared to usual care; a moderately positive influence of educational videos was reported only for FS in one study.138

The crucial role of organisational measures, enabling subjects to adopt the recommended behaviours, is confirmed by reports showing that screening rates were increased when the physician was delivering an FOBT request form, together with an information pamphlet, while no effect of the leaflet was observed when people interested in screening were requested to actively seek a referral to screening from their providers.132 ,140 Similarly, the inclusion in the reminder letter of instructions for fixing an appointment for endoscopy did not result in an increase in the uptake.58

### Cost-effectiveness of interventions to promote uptake

Organised screening programmes, dealing with a massive number of eligible subjects, already entail substantial investments, driven not only by the adopted test but also by organisational costs and by the cumulative cost of disposable and non disposable material (table 5). Thus, any additional intervention aimed to increase participation should prove to be not only effective but also sustainable from an economic perspective.

Table 5

Cost-effectiveness studies of interventions aimed to enhance screening acceptance

Most organised programmes55 ,71 ,74 ,85 routinely adopt a recall strategy using one or two mail reminders for non-participants. Despite a specific estimate being not available, the cost of this approach may be approximated. In an Italian programme based on biennial invitation with FIT, the overall cost of the programme resulted to be €3.2 and €2.3 per person invited at first and subsequent rounds, respectively.148 Assuming equivalent numbers between invitations and reminders, no more than 4–10% of such cost appeared to be related with mail reminders. In a Dutch population study on CTC, mail reminders accounted for only 1% of the overall screening cost.149 The fact that mail reminders accounted only for a marginal fraction of the overall programme cost was not unexpected since reminders-related costs are minimised by the pre-existing organisational platform, including IT infrastructure, governing the programme flow. Despite these studies not providing an estimate of reminders' efficacy, it may be reasonably assumed that they are nearly cost-neutral in an organised setting.148 ,149 Thus, it may be assumed that any intervention to increase uptake that is able to exploit the already implemented infrastructure of an organised programme, and that is not more costly than a mail reminder, may be considered sustainable from an economic perspective. Of note, one cost-effective model assessing the cost and benefit of an advanced notification letter within the Australian organised programme showed its convenience with an incremental ratio of $3976 per life-year gained.150 In a randomised trial within the Italian organised screening programme, an advance notification letter was associated with a 20% increase in the attendance for FS with an incremental cost per additional screenee of €9–10.87 Mailing of the FOBT kit was associated with an increase in participation, but it may not always represent a cost-effective option, as the incremental cost per additional participant is dependent on the mailing and kit costs and on the expected increase in uptake.55 Mailing a recall letter with a test order coupon resulted in a substantial decrease in the costs, but also in participation, compared to mailing a second FOBT kit to all non-responders, in the French screening programme.151 An SR on the cost-effectiveness of interventions aiming to reduce structural barriers in CRC screening in opportunistic primary care settings showed that mailing the FOBT kit, or supplying prepaid return envelopes,24 represents cost-effective interventions, with costs per additional screenee ranging between$0.6 and $32. Kit delivery by non-health professionals, or by the GP, although effective, may not represent a sustainable strategy, considering the reported low proportion (25%) of collaborative GPs reported in a recent pilot screening trial,54 the staff costs when using nurses, or on-health professional, or the difficulties in maintaining volunteers’ involvement over time. The cost per additional screenee of client-directed interventions, such as telephone/mail/web reminders, or navigators, was below$100 in two studies,141 ,146 but substantially higher, up to $2602, in the other trials conducted in primary care settings in the USA.122 ,141–146 The reasons for the unfavourable cost-effectiveness ratio of these interventions rely on multiple factors, such as the relatively small size of the target population, the personnel cost, the low increase in uptake and, to a lesser extent, the cost for the supply. The increased costs of such interventions in this setting are also related to the fact that their implementation requires the creation of those organisational platforms already incorporated in organised programmes. Indeed, the larger target population and the automatisation of telephone calls likely explains the convenience of the client-directed interventions in the two series reporting the lowest incremental cost.141 ,146 The only study addressing the costs of a provider-directed intervention showed an incremental cost of$21 124 per percentage point increase in screening rate.144

Cost-effectiveness of an awareness campaign for early diagnosis of CRC has been assessed using a simulation model in the UK:147 the intervention appeared to be cost-effective, with an additional cost of £13 496 per saved quality-adjusted life-year, but the effectiveness was related only to an increase in GP referrals, as CRC screening rate were practically unaffected.

## Discussion

The delivery of CRC screening is the result of a multidisciplinary undertaking, involving several health professionals, as well as technical and administrative staff, and requiring the implementation of procedures and infrastructures facilitating exchange and sharing of information. Successful completion of this process requires an active engagement of the screenees, who are expected to carefully follow faecal sampling instructions, to comply with specific bowel preparation regimens and to plan their participation taking into account the specific features of the screening protocol (ie, need for sedation, influencing driving ability). Opportunistic screening, which depends on individuals requesting screening, or on health professionals recommending it, cannot adequately support the progression of screenees through the continuum of such process, where several critical steps are taking part outside the traditional patient–provider encounter.

The results of this review confirm that both policy measures and organisational changes, ensuring that all eligible subjects in the target age range are actively and systematically invited at regular intervals, following an explicit protocol, need to be implemented in order to optimise CRC screening acceptance. The decision of some European countries to reorient their screening policies from opportunistic screening to organised programmes,152 as well as the adoption of strategies based on active invitation of the target population in underserved communities,35 ,45 to address inadequate population coverage, would suggest that poor screening acceptance is mainly attributable to lack of offer and/or of inadequate consideration of those factors that might enable individuals' participation.

Reports from cervical and breast cancer screening programmes also showed that organised programmes can determine a larger reduction of cancer incidence153–155 or mortality,156 compared with opportunistic screening, and that they can be effective in reducing pre-existing survival differences across socio-economic groups.157 ,158 Their greater attention on the quality of the entire process can ensure greater protection against over-screening and complications,159–161 resulting in a more favourable cost-effectiveness profile.

Cost-effectiveness evaluations are context-related as the organisation of service delivery and the costs of the procedures differ across jurisdictions. However, the estimates of the relative costs are likely more generalisable than the absolute costs. Also, once a model has been built up, it may be considered as a matrix that may be easily converted in country-specific or context-specific scenarios. In fact, in most comparative studies, the single components of the final cost are provided, so that they may be simply converted, or updated, to the relative costs of different countries. Similarly, efficacy end-points, such as participation, may be reassessed with simple prospective studies, allowing the redefinition of the incremental cost-effectiveness ratio in a different scenario.

A range of approaches may be required to improve uptake. Subjects who are already considering to take part in screening in the future may be more responsive to interventions using PCP endorsement, or addressing enabling factors, such as periodic invitations, reminders and aids to making the test more practical.162 On the other hand, the persisting differential uptake of CRC screening across educational level and/or socio-economic status (SES), also in the context of established population-based programmes,8 ,9 highlights the need to develop more effective approaches to address individuals' values and preferences, their knowledge, risk awareness and beliefs about CRC, and their attitudes and expectations towards screening.

Most studies have been assessing the effect of educational interventions focused on social cognition variables (ie, knowledge, risk awareness, attitudes). These were shown to be strongly associated with intention, but only weakly with action, which is better predicted by factors related to life difficulties.163 Qualitative research findings indicate that several cognitive factors, including fatalistic beliefs about CRC164 and individuals' evaluation of the relative weight of short-term inconveniences and long-term benefits (consideration of future consequences (CFC)),165 are associated with SES and mediate the negative impact of social deprivation on attendance. Interventions taking CFC into account, by emphasising short-term benefits, could, for example, promote equality in screening participation.166

Reliance on printed communication to convey educational messages may be problematic, particularly as subjects with lower health literacy are less likely to seek information and show a lower confidence to participate in screening.167 Studies conducted in opportunistic settings would suggest that interventions involving the offer of phone support by nurses, or other non-health professionals, as well as of tailored patient navigation, may increase uptake. However, the size of those studies is often relatively small, they rarely take a pragmatic approach and most cost-effectiveness analyses showed that their incremental costs would not be sustainable. Therefore, interventions shown to be effective in opportunistic settings may not be easily implemented in large population-based mes. Developing sustainable strategies to deliver tailored educational messages adds to the challenge of integrating the findings of qualitative research in the design of educational interventions.

Concerns have been raised recently about the potential conflict between promoting uptake and the spirit of enabling people to make an informed choice about whether or not they want to be screened. Indeed, people may have different risks of contracting the disease, while some of the harms related to screening are equitably distributed among the screenees.168 Therefore, supporting individuals making an informed choice about entering screening is regarded as a desirable goal.169 However, available evidence supporting the effectiveness of educational interventions in promoting the adoption of decisions consistent with individuals' values and preferences15 ,114 is limited. Developing valid measures to evaluate interventions impact on informed decision making and designing studies incorporating these dimensions represents a challenge to be addressed in future research.

The impact of most interventions in organised settings was assessed using participation in a single invitation round as the main indicator of effect. Initial uptake may represent a process indicator, measuring efficiency, or acceptance of the programme, or reflecting the impact of organisational choices.170 However, if regular retesting is required to achieve the expected benefit, the response in each single round may overestimate the success of the programme. Sustained attendance over several rounds represents a more appropriate indicator of impact.

In conclusion, available evidence indicates that organised programmes allow to achieve an extensive coverage and to enhance equity of access. The adoption of a public health perspective involves a commitment to ensure sustainability of the intervention over time. Achieving these goals requires the adoption of policy measures establishing a supportive context for screening delivery. Organised programmes provide an infrastructure facilitating the introduction into routine care of evidence-based approaches to enhance screening acceptance and allowing to achieve a more favourable cost-effectiveness profile of potentially effective strategies.

View Abstract

• ## Supplementary Data

This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Files in this Data Supplement:

## Footnotes

• Contributors CS performed the review of interventions in opportunistic and organised settings, in collaboration with NS and with other coauthors. CB performed the literature search and the evaluation of retrieved papers. CH and JI performed the review of cost-effectiveness studies and of trials comparing different strategies. JI suggested studies focused on vulnerable groups. All authors have read and approved the final version of the manuscript.

• Competing interests None.

• Provenance and peer review Commissioned; externally peer reviewed.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.