SEARCH

SEARCH BY CITATION

Keywords:

  • behavior;
  • colorectal cancer;
  • early detection;
  • screening

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

Australia is one of the first countries in the world to implement an organized whole-of-population screening program for colorectal cancer (CRC). Australians have made broad contributions to CRC in general ranging from primary prevention through genetics, secondary prevention and treatment, to palliation. This overview focuses on some of the contributions of direct relevance to population-based screening, stretching from technology development to population-based controlled studies and health services research. In terms of simple screening tests in a two-step screening strategy, the evidence is overwhelming that fecal immunochemical tests for hemoglobin (FITs) improve detection and are more acceptable. FIT-based screening is clearly acceptable to Australians and it has been demonstrated that a national organized screening program is feasible. In terms of benefit for Australians, with full roll out and high uptake by the population we could see the number of cases dying from CRC halved by this strategy. To this will be added the extra-screening benefits of improved diagnosis, improved treatment and improved public awareness, all benefits of other screening programs. CRC screening has progressed from a matter of irrelevance and distaste, to commonwealth government policy in the context of an organized program for all Australians.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

It is timely that we celebrate the Australian contribution 50 years after the foundation of the Society, now that we have initiated a national bowel cancer screening program (NBCSP), the first organized screening program applicable to men, and one of the first countries in the world to implement an organized whole-of-population program. Australians have made broad contributions to colorectal cancer (CRC) in general, ranging from primary prevention through genetics, secondary prevention and treatment, to palliation. This review focuses on some of the contributions of direct relevance to population screening but it is acknowledged that these broader contributions, as described in the National Health and Medical Research Council (NHMRC)/Australian Cancer Network guidelines,1 are equally important to control of CRC.

Goal of population screening

The goal of population-based screening for CRC is to reduce mortality from and/or morbidity due to the disease.2 The target of screening is therefore either curable cancer or those adenomas (advanced adenomas) most likely to progress to cancer. Presymptomatic detection of localized, curable cancer reduces morbidity and/or mortality, while detection of pre-invasive lesions, i.e. adenoma, will reduce cancer incidence.

Criteria justifying screening

The criteria that justify screening were defined for WHO four decades ago3 and can be categorized as follows:

  • 1
    The targeted disease must be a major health problem.
  • 2
    There should be a latent stage.
  • 3
    There should be a simple test to detect the latent stage.
  • 4
    Application of the test must be proven value at the population level on an intention-to-screen basis.
  • 5
    The screening test must be acceptable.
  • 6
    The screening process in its entirety must be feasible.
  • 7
    Cure of the screening-detected lesions must be possible, with benefit outweighing the harms.
  • 8
    The entire process must be cost-effective.

As we will see, CRC screening meets all of these criteria.

We can best understand the Australian contribution by working through the above criteria, but before doing so, we need to clarify what population screening involves.

The nature of population screening

Because screening is directed at reducing the population burden of a disease, it is ideal that screening is organized and screening tests should not be considered in isolation from the health services process that provides them.

The International Agency for Research on Cancer has more recently defined an organized screening program as one that has the following features: (i) an explicit policy with specified age categories, method, and interval for screening; (ii) a defined target population; (iii) a management team responsible for implementation; (iv) a health-care team for decisions and care; (v) a quality assurance structure; and (vi) a method for identifying cancer occurrence in the population.4

In contrast, opportunistic (otherwise referred to as ad hoc or case-finding) screening is done outside of an organized program and is often delivered through fee-for-service reimbursement of physicians. Compared with opportunistic screening, organized screening focuses much greater attention on the quality of the screening process, including follow up of participants.5 Consequently, organized screening provides greater protection against many of the harms of screening, including over-screening, poor quality and complications of screening, and poor follow up of those who test positive.

The screening test is just one event in a process that is simply represented in Fig. 1. ‘One-step screening’ is where the diagnostic procedure is used as the screening test, for example, colonoscopy for CRC. ‘Two-step screening’ is where an intermediate test (the ‘screening test’) is used to identify a subpopulation more likely to have neoplasia—those with the positive test proceed to the reference diagnostic standard. The screening test filters for those more likely to have neoplasia as indicated by a positive test6 and as quantified by the post-test probability.7 For screening to be effective, the entire screening process must be operational, accessible, and of adequate quality.

image

Figure 1. The screening pathway from subject engagement to rescreening and surveillance. One-step screening takes the subject straight to the diagnostic procedure. Two-step screening inserts a simple screening test which identifies who should proceed to the diagnostic test. Each event in the pathway must function well to ensure the best possible outcomes from screening.

Download figure to PowerPoint

A major health problem

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

Australia has one of the highest age-standardized rates for CRC on the world scene.8 Men are affected slightly more commonly than women, and over 90% of CRC occurs after the age of 50 years; the majority of diagnoses are after age 65. CRC is the second commonest cause of cancer-related death. Australian Institute of Health and Welfare estimates 8209 cases in men in 2010 (an average annual increase of 186, 2.3%, since 1996) and 6571 cases in women in 2010 (an average annual increase of 130, 2.0%, since 1996). Thus, we expect almost 15 000 new cases in 2010.

Colorectal cancer creates a major health burden on Australians, being the 10th highest single disease cause of health loss in 2003.9 Furthermore, the actual increase in cases continues to rise alarmingly; using the Australian Institute of Health and Welfare estimates the increase will be 25% in 10 years. This increase has occurred despite a substantial increase in the colonoscopy rate.10 One might assume that we are not colonoscoping the right people at the right time, or else we are not doing it well enough!

The profession's response

In response to this burden of disease and the potential for prevention through early detection, raised first by Greegor,11 the Society recognized the need for guidelines. While older members may recall earlier events, the main trigger for what we see happening today was the establishment of Guidelines in 1989 by the Gastroenterological Society of Australia.12 There was no level 1 evidence for population screening or high-risk surveillance at the time, but confusion in practice demanded that consensus was needed to guide practitioners. This was followed in 1994 by further development of these Guidelines in association with the Australian Cancer Society.13 Finally, engagement of a multidisciplinary team led to the publication of the first edition of the NHMRC world-acclaimed Guidelines in 1999, which have recently undergone a further revision.1 These Guidelines have been crucial in progressing screening, and indeed all organized approaches to managing CRC, because they demonstrated to funders and politicians that the Profession was in agreement about what should be done and how it would be done. Without this consensus, we would not be where we are today with an organized national screening program.

The nature of disease—latency

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

Of crucial importance to the success of screening is that there is a window of opportunity to detect pathology before it presents itself as established disease and while it allows for cure.

By the early 1960s, the idea of the adenoma–carcinoma sequence was being enunciated.14 Around the world the idea was emerging that strategies for early detection were worth pursuing. By the mid-1990s the concept of the adenoma–carcinoma sequence was well established and fully described in a reference text book dedicated to CRC prevention.15

Unfortunately, symptoms do not occur early enough to enable worthwhile detection of latent stage disease—Greegor's early observations11 were confirmed in Australia. One Australian group showed that only 33% of rectal and sigmoid cancers and adenomas were accompanied by overt rectal bleeding.16

Bleeding from colorectal neoplasia

In view of the work by Greegor, several Australian gastroenterologists who had established a keen interest in gastrointestinal bleeding set about studying the volume and frequency of bleeding from colorectal neoplasms, and how this related to detection by guaiac-based fecal occult blood tests (GFOBTs).17 Using the crude method of radiochromium measurement of bleeding, they were able to show variation in volume of bleeding from cancers from day to day. Adenomas appeared to bleed only occasionally.

Soon after that followed a series of important contributions that further clarified the nature of bleeding due to colorectal neoplasia. The volume of blood lost in feces in normal subjects, based on studies using the radiochromium (51Cr-labeled red cell) method, is 0.5 ± 0.4 (SD) mL per day with an upper limit of 1.5.18 While fecal blood loss is common in patients with colorectal neoplasia,17 there is overlap with the normal range.17 Measurement of fecal blood loss by assaying heme plus heme-derived porphyrins (HDPs) presents the same problem.19 Thus a test which depends on blood volume loss or on fecal heme and its degradation products, will not completely discriminate between normal subjects and cancer-bearing patients.

The amount of bleeding from a cancer is not dependent on stage, but rather on site17 and possibly size. Proximal, large cancers bleed the most. Adenomas bleed less than cancers, as shown by the findings with HDPs and radiochromium, and bleeding from small adenomas (< 1 cm) is uncommonly outside the normal range.17,19

The quantity of bleeding varies from day to day in patients with large bowel neoplasia.17,20 Using GFOBT, testing of just one or two stools is of inferior sensitivity to testing of three, but testing more than three adds little benefit and worsens specificity.17,21 Thus it is normally recommended that GFOBT are performed on multiple, usually three, stools.

A simple test for detecting the latent period

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

The nature of the test to be used for CRC screening has been one of the most debated and misunderstood issues in gastroenterology, primarily because there is a range of potential screening tests which differ considerably in how they perform and what they detect.22 In addition, there is a failure to recognize that a screening test is just one element of the screening process (Fig. 1). The status of these different tests for use in screening has been recently reviewed.23

Importantly, a screening test should not be considered simply as diagnostic test. The WHO concept emphasizes the importance of using a simple cheap test that is acceptable and acts as an aid to early detection. Australians highlighted the fact that a screening test refined likelihood of the disease being present.6,7

Once selected on the basis of age, the initial screening test in two-step screening acts as the next filtering, or chance-predicting, step in the process of selecting who needs to undergo the diagnostic procedure, i.e. separation of those who at present are at high risk from those who are at a low risk. This concept is embodied in the pre-test : post-test likelihood ratio, that is, the chance of having CRC when one returns an abnormal test result (e.g. a positive FOBT), compared with the chance of having cancer when one has not done the test.24 This likelihood ratio derives from the same basic observations about a test performance as those used to calculate specificity and sensitivity; it is simply the sensitivity of a test divided by its false-positive rate.

In the mid-1960s Greegor described the detection of an unsuspected cancer in an asymptomatic patient using a fecal occult blood test.11 Soon after, he proposed that the test be used for screening, an idea that quickly caught the interest of groups around the world. This led to the establishment of randomized controlled trials (RCTs) across the USA and Europe (including the UK) using a simple occult blood test technology.23,25–28

At around that time, more sophisticated approaches were becoming available to detect blood in feces, such as immunochemical methods for globin and porphyrin assays for bacterially derived heme products.29–31 It was becoming apparent that we needed to understand a lot more about the fate of hemoglobin in the gut if we were to understand how best to apply these new technologies, and Australians participated in defining this.

Fate of hemoglobin in the gut

Heme is presented to the proximal gastrointestinal tract mainly as a hemoprotein in food (hemoglobin, myoglobin) or as erythrocytes derived from physiological or pathological bleeding. In the stomach, heme is rapidly released from the globin and precipitated. In small intestine, it is solubilized and around 5–15% is absorbed, so that the majority of this heme is passed to the colon.29

In the colonic lumen, heme and hemoglobin may derive from small intestine or from colonic bleeding. Heme may be modified by the luminal microflora in two ways: by removal of iron and/or by changes to the vinyl side-chains of the porphyrin ring.30 A range of heme-derived dicarboxylic porphyrins (HDPs) is produced as a result. The iron-free porphyrins have no peroxidase activity and so are not detectable by guaiac tests.

Hemoglobin is also degraded in feces but the process is slower than in the lumen of the stomach or small intestine. Using immunoquantifying methods, it has been shown that hemoglobin can lose as much as half of its immunoreactivity in human feces over 24 h.32,33 Such degradation would be expected to occur in vivo during transit around the large bowel.

There is no immunoreactive hemoglobin present in feces when proximal gastrointestinal bleeding is simulated in healthy volunteers who ingest their own blood (10–100 mL),32,33 even though heme can be readily detected (guaiac tests or heme-porphyrin assays). Clearly, heme is more resistant to degradation in the gut than is globin, and heme-dependent tests will be relatively more sensitive for detection of proximal bleeding.

The nature of hemoglobin derivatives found in feces can be predicted to vary according to the site of bleeding (Table 1).32 The more proximal the bleeding, the more likely are HDPs to be the principal products. The more distal the bleeding, the more likely it is that intact hemoglobin will be present.34 As a consequence, immunochemical tests, which require hemoglobin to be relatively intact, would be more selective for large bowel bleeding than tests which detect heme or HDPs. Importantly, each product varies in its capacity to be detected by a particular FOBT technology (Table 1).

Table 1.  Predicted balance of hemoglobin derivatives in feces, depending on site of blood loss
DerivativeSite of bleedingTest technology
GastricProximal large bowelDigital large bowel
  1. Adapted from Young and St John.32

  2. FIT, fecal immunochemical tests for hemoglobin; GFOBT, guaiac-based fecal occult blood test; HDP, heme-derived porphyrin.

Intact hemoglobin+++FIT
Intact heme++++GFOBT, HDP
Heme-derived porphyrins++++++HDP

It is noteworthy that even then we knew a lot about issues such as stability of hemoglobin in feces. Neither heme nor hemoglobin is stable in feces (see above). Immunochemical tests are potentially vulnerable to the effects of degradation of hemoglobin during transit through the colon.32,33 There is no published information on whether the different antibodies used by different manufacturers in different fecal immunochemical tests for hemoglobin (FITs) vary in their ability to detect partially degraded forms of hemoglobin. The implications of analyte instability in feces for clinical practice have not been fully studied but have posed problems for the national program in 2009.

Dietary factors were also a problem for some of the technologies. Dietary peroxidases31,32 interfered with GFOBT and HDP assays,30 creating difficulty to control effects on specificity. Also, because GFOBT and HDP assays could detect bleeding from any level of the gastrointestinal tract, they could influence GFOBT and HDP assays, although the effect on the former was relatively small.35

Implications

Thus, simple tests for detecting blood products in feces were available, but the technologies differed. At a technological level, FITs seemed likely to be more useful.

Evidence that screening tests are effective

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

For CRC, the extent of the evidence has been recently reviewed.23 Two-step screening using GFOBT as the initial test is proven to be effective in reducing CRC-specific mortality by multiple RCTs measured on an intention-to-screen basis. The concept of intention-to-screen outcome assessment is the same as for intention-to-treat analyses where the intervention is a drug. Results in all people allocated to the intervention group are analyzed regardless of test participation because the usefulness of the intervention is initially dependent on its acceptability.

To summarize these RCTs, the relatively insensitive GFOBT Hemoccult II used biennially returned intention-to-screen reductions in CRC mortality of 15–20%.25–28 This measure improves to a 33% reduction with rehydration of Hemoccult II used annually, a process that increases sensitivity but results in considerable deterioration in specificity.25 This increased sensitivity achieved with rehydrated Hemoccult is also associated with a reduction in CRC incidence when followed up for 18 years,36 presumably resulting from increased detection and removal of adenomas. It should be noted that intention-to-screen end-points underestimate or dilute the screening effect in individuals when there is nonparticipation in the cohort invited to screening, or when there is screening outside of the trial in the non-invited group.37

We can best summarize this overwhelming evidence for effectiveness as indicating that early detection of colorectal neoplasia (down-staging) using a method that detects the pathological event of bleeding results in reduced mortality and incidence as long as one tests repeatedly (a fundamental requirement of any screening program using a simple test as the initial screen).

A crucial contribution from Australia was the meta-analysis that pooled the data from the three earliest trials to provide us with the highest level of evidence for effectiveness. This meta-analysis indicated a reduction in CRC mortality of 16% on an intention-to-screen basis, although it underestimates benefit because it pools the data from annual and biennial results and includes both sensitive and less-sensitive GFOBT types.38

Moving beyond the evidence to action

Several issues became topical in Australia in terms of translating these findings into action: impact of screening for participants, the need for an organized national program involving retesting at appropriate intervals, and improving screening test sensitivity.

Participant benefit

The mortality benefit, ranging from 15% to 35%, was both welcomed and derided depending on one's perspective. A key point missed by the critics of benefit being too low was that the benefit was expressed on an intention-to-screen basis, taking a population perspective. If one looked at benefit to actual participants, this almost reached 40% reduction using just biennial screening with Hemoccult II, the least sensitive of the GFOBT available; specifically 34% in the Danish and 39% in the UK trials.23 The individual stood to gain a lot by doing a simple test.

Need for an organized program

Ad hoc screening was not seen as being advantageous. The proof came from highly organized screening programs (as RCTs) where all aspects of the screening process were managed, and, as indicated above, the incidence of disease was rising despite a much faster rise in the colonoscopy rate. As outlined in The nature of population screening, the International Agency for Research on Cancer has recently defined an organized screening program and called for this to become the norm as it provides greater protection against many of the harms of screening, including over-screening, poor quality and complications of screening, and poor follow up of those who test positive.

Consequently, efforts were made to move beyond the Guidelines to effective, strategic implementation. An Australian Health Technology Advisory Committee was instigated in 1995 to look at the evidence and consider if Australia should establish an organized screening program. After a year's thorough analysis and debate, it concluded in 199739 that:

  • 1
    On the basis of published evidence, and subject to favorable preliminary testing, it is recommended that Australia develop a program for the introduction of population screening for CRC by fecal occult blood testing for the average risk population (well population aged over 50).
  • 2
    Given the uncertainties relating to the most effective means of implementing such a program and to the feasibility, acceptability, and cost-effectiveness of such a program in the Australian setting, the program should commence with preliminary testing involving a number of pilot and feasibility studies.

But there was no public lobby and the profession remained skeptical. Approaches to government fell on deaf ears given the absence of a lobby and the apparent professional disagreement.

Two strategies finally swung the pendulum in favor of an organized program. First was the NHMRC guidelines, which signaled unity of opinion within the profession.1

Second was a politically crafted submission direct to federal cabinet from the National Cancer Control Initiative in 1999 that resulted in a promise of funds to initiate the pilot screening program; this was funded in the commonwealth budget of 2001.

Improving screening test sensitivity

It is interesting to note that the reduction in CRC-specific mortality was achieved with a GFOBT (Hemoccult II) for which its once-only sensitivity for cancer (testing at just one point in time collecting three fecal samples) was only about 30–35%.23 This promised considerable benefit from a better technology with greater sensitivity, and it was generally considered that a better test was needed, especially if we were to include capacity to detect adenomas and thus prevent cancer, in addition to detecting curable cancer.

During the 1980s and 1990s different FOBT technologies had become available and our understanding of the fate of hemoglobin in feces improved dramatically (see above). It seemed likely that the newer more sophisticated technologies would be more precise and more acceptable. The challenge was how to compare these with the proven technology to determine if they provided an advantage in screening.

This led to a series of recommendations on studies comparing different test technologies.32,40 The challenge was formidable; there were no real precedents for this and the clinical epidemiological theory underpinning evaluation of diagnostic and especially screening tests was in its infancy. Recently, Australian researchers have led a world initiative (significantly involving the Asia–Pacific region, where incidence of CRC is rising) in defining the stepwise process needed to evaluate new screening tests in a way that provides population proof of their benefit.22

Australian contributions to improvement of FOBT have been substantion and the reader is referred to several studies41–43 and a recent review of the value of FITs.44 Among other international studies, these showed categorically that FITs were better at cancer detection than GFOBT, which would translate into a better impact on mortality reduction, and that this was achieved without the major specificity problem that arose when increasing the sensitivity of GFOBT as occurs with the Hemoccult Sensa test.45

It was also apparent that some of the FIT tests available were quantitative. Two things arose from this. The first was the demonstration that some adenomas definitely bleed.46 This, in turn, was followed with proof that FITs do have a capability to detect advanced adenomas, certainly much better than GFOBT;43 this will enhance ability to reduce incidence when undertaking population screening. The second was that this provides considerable advantage for health-care systems such that the cut-off for positivity can be chosen to suit circumstances, achieve the preferred balance of true- and false-positive results, and control the colonoscopy follow-up rate according to available resources.

The fact that FITs can better detect adenomas is likely to improve professional acceptance, given the capacity for this to protect against development of cancer. We already knew from the Minnesota population screening trial that high-sensitivity FOBT screening reduced CRC incidence.36

As will be seen later, FITs also presented other benefits in terms of population participation.

Other screening tests

Flexible sigmoidoscopic screening has now been underway, primarily in Western Australia, for over a decade. Its use as a screening tool is supported by case–control studies.23 Population-based RCTs with mortality as an end-point are underway in the UK, Italy, and USA.

In the Western Australia study, between 1995 and 2005, 3402 people underwent an initial flexible sigmoidoscopy screening examination and 1025 had a 5-year recall examination. Fourteen percent of initial screenings detected at least one adenoma. Over a mean follow-up time of 8 years, invasive cancer was detected in 0.4% of participants; 0.7% of those with an initial normal screening later developed CRC, with 75% of these occurring proximal to the splenic flexure. They concluded that flexible sigmoidoscopy is a viable screening method but with clear limitations in terms of reach within the colon. Despite this evidence for feasibility and capacity to detect lesions, it is not popular as a screening tool elsewhere in Australia.47

Computed tomographic colonography (CTC) has undergone some evaluation. One report compared it with colonoscopy as a follow-up tool for patients with a positive fecal occult blood test. The authors concluded that CTC appears less accurate, less effective, and potentially more costly than colonoscopy.48 It has been acknowledged that there are big challenges in implementation of CTC as a screening tool in Australia.49 One concern has been about the impact of finding extracolonic pathology; a cost-effectiveness analysis found 118 of 432 subjects (27%) had extracolonic lesions.50 In 7.4% of cases it was felt that findings were clinically relevant and in 2.1% it was felt that subjects may derive a clinical benefit from treatment. Clinical follow up added just 14% to the cost. Such findings suggest that extracolonic lesions might not pose a major problem. Another study was undertaken to assess population acceptance of CTC screening in Western Australia; overall participation was 18% and the yield of advanced colorectal neoplasia was 8.7% (95% confidence interval 5–14%).51

Test acceptability

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

It is self-evident that for population-based screening to be affective, the test must be acceptable, or it will not get done. CRC screening has a dimension shared by no other cancer screening in that the initial screening test in two-step screening, specifically an FOBT, can be done conveniently in the privacy of one's own home. Behavioral research has been undertaken to identify and overcome the barriers peculiar to this type of screening. In an early (1984) study of 501 people visiting their general practitioner (GP), the Health Belief Model was found to inadequately explain screening behavior by accounting for only 12% of the variance in screening behavior.52 Perceived barriers to taking the test and perceived susceptibility to bowel cancer were the only components contributing significantly in the multiple regression analysis in that study.

In the mid 1990s, Weller studied awareness in nearly 2000 participants in a CRC screening program.53 They concluded that while there were high levels of awareness of fecal occult blood test screening, most respondents had not had a test, nor did they intend to take a test in the future. Important determinants of participation in screening were a family history of CRC, a belief that bowel cancer can be cured if detected at an early stage, a perception of personal susceptibility to bowel cancer, and an acceptance of the technique. Unfortunately, this did not provide information about non-participants.

Larger-scale research projects involving screening participants and non-participants have now been conducted and are close to being reported.54 These are based on better developed behavioral models and include actual participants and non-participants, rather than simply examining intentions. The significant behavioral determinants are relatively few: fecal aversion, knowledge, and perceived value (S Cole, pers. comm., 2009). The significant demographic variables are gender, age, and socioeconomic status. When studying readiness of the general population to screen, the majority were found to have not progressed beyond contemplation; half of these were still in the precontemplation stage.55

Recently, the strategy of providing an advance notification letter a few weeks prior to offering the screening test was trialed as a means to advance people through the stages of readiness to screen. In an unbiased population-based RCT, this was shown to increase participation from 40% to 48%.56 This strategy is now incorporated into the NBCSP.

Test technology and screening behavior

The FIT technology, through its analytic specificity for human hemoglobin, seemed likely to simplify stool testing and provide behavioral benefits. It had been shown that FITs were selective for colorectal bleeding and probably removed the need for diet and drug restrictions. The more recent branded versions also used different stool-sampling methods and evidence was accumulating that they required only one or two stool samples rather than three.57 This led to some important behavioral research exploring the benefit gained in population participation in testing using this new technology.

Australian research made several important contributions. Removing diet and drug restrictions was shown to significantly improve participation (by 12% in one study).58 A subsequent unbiased population RCT with participation as the key end-point demonstrated that, by moving from a stick-sampling GFOBT (three-sample) to a brush-sampling FIT (two-sample), participation rose from 23% to 40% in a largely screen-naïve population that did not have great awareness of the value of screening for CRC.59

There has been a push around the world to give people a choice of screening test, assuming that offering choice removes barriers and increases participation. In practice, a choice of screening test did not improve participation.60 Participation with FOBT screening was higher than with other tests, including colonoscopy and CTC. This has crucial implications for detection of advanced neoplasia in a screening population as the screening test must get done.

Participation with screening CTC is reported as modest; the overall participation rate is just 18% (184/1009) although in those who did it, it had a high level of acceptability: participants found colonoscopy (87%) and CTC (67%, P < 0.001) less unpleasant than expected.61 About 29% (26/89) of CTC subjects had a positive screening test meaning that the colonoscopy follow-up rate will be high and cost-effectiveness uncertain. The yield of advanced colorectal neoplasia was 8.7% (95% confidence interval 5–14%). They also concluded that providing a choice of test did not increase participation.

Limited outcomes of population FIT screening in Australia were first available in 1992.62 The paper reported an evaluation of the first 2 years of a South Australian CRC screening program that was established in 1988. Based on 1 year of follow up, over the period of this analysis there were 24 cancers and 99 adenomas detected in 6208 participants using the FIT Detectacol test (Institute of Medical and Veterinary Science, Adelaide). The authors estimated sensitivity and specificity of the test (for cancer) at 83% and 96%, respectively. The estimated predictive value of a positive test for CRC in this population was 7.5%. They found that higher-than-average socioeconomic groups participated, and almost one-third had suffered from bowel symptoms in the 6 months before taking the test. This limited the generalizability of results to the wider population.

Doctors and screening behavior

It is not just the public who pose a challenge but doctors themselves. A study in 1982 concluded that recognition of subjects at high risk for CRC, and knowledge about the proper use, follow up, and potential for screening of fecal occult blood tests needed emphasis in medical educational programs.63 Population research in Adelaide has shown that GPs are credible advocates for screening. The linking of a GP's name to the screening offer increases participation by approximately 30%.64

Cost-effectiveness

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

Ultimately, the incorporation of population screening into the health-care agenda is dependent on demonstration of cost-effectiveness. By the mid-1990s there had been several attempts internationally to model the cost-effectiveness of CRC screening but all modeling was based on assumptions and estimates and not on real screening data and actual costs. In 1996 Salkeld et al. published on the modeling of cost-effectiveness in Australia using Australian costs and incidence together with test performance and effectiveness from the Minnesota trial.65 Even with outcomes from rehydrated Hemoccult, a poor specificity test with resultant high false-positive rate, the results were excellent with a cost per life-year-saved of $AUD24 660. This was of the same order of magnitude of benefit from breast and cervical cancer screening.

So the question became, not should we screen, but why are we not screening? This paper was influential in persuading fund-holders to support screening.

More recently, cost-effectiveness of screening using flexible sigmoidoscopy compared with annual and biennial rehydrated Hemoccult fecal occult blood testing and colonoscopy has been undertaken using Australian costs and a Markov model.66 Colonoscopy averted the greatest number of cases of CRC (35%), followed by flexible sigmoidoscopy (25%) and annual fecal occult blood testing (24%). Colonoscopy averted the greatest number of deaths from CRC (31%), followed by annual fecal occult blood testing (29%) and flexible sigmoidoscopy (21%). Flexible sigmoidoscopy was the most efficient in terms of cost per life-year saved ($AUD16 801), followed by colonoscopy ($AUD19 285), biennial ($AUD41 183), and annual ($AUD46 900) fecal occult blood testing.

These costs for FOBT would be greatly improved if findings with FITs were used to model the cost.

Feasibility—the Australian pilot screening program

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

While much of this research was being undertaken, lobbying the Commonwealth Government finally led to the AHTAC-recommended pilot study getting underway in late 2003 with the view of testing feasibility within Australia. Results of this single round of screening were reported in 2005 (see below for some detail). The initial plan was to offer several rounds of screening to about 50 000 people in Adelaide, Mackay, and Melbourne. The main focus was to be on the general population aged 55–75 years. It was considered that high-risk groups needing colonoscopic surveillance could be handled through the medical system but that organized screening was needed to ensure the best outcomes for the general population.

Following extensive consultation with all the stakeholders it was decided to adopt a centralized process for mailing out invitations with FIT kits to people identified from the Medicare database. The Pilot Program used two different brands of FITs. Immunochemical FITs were selected as they have no potential for interference by diet or medication. Positive results were notified to participants and their nominated GP, and participants were advised to see the GP for initial assessment and referral on to a colonoscopist. It was considered essential that there was adequate quality across the pathway, that data were consistent and accurate, and that a quantitative FIT be included to allow flexibility in the cut-off for positivity if the positivity rate became unacceptably high relative to colonoscopic resources. A safety net process was put in place to maximize the likelihood that people got to colonoscopy if they tested positive.

The Final Evaluation Report of the Bowel Cancer Screening Pilot Program showed that a NBCSP would be feasible, acceptable, and cost-effective in Australia.67 A total of 56 907 men and women from Mackay, Adelaide, and Melbourne were invited to participate in the pilot. The overall participation rate was 45.5%, which compares well with participation rates in other, longer-established screening programs. Participation in the Pilot was higher among women (47.4%) than men (43.4%). The major reasons stated for participating in the Pilot were precaution; prevention; early detection; health check important, and peace of mind. Of the 25 688 correctly completed FITs, 2317 returned a positive result, giving an overall positivity rate for the Pilot of 9.0%, with a higher rate in men and among older participants.

The Program detected 176 advanced adenomas and 67 suspected cancers. However, it should be noted that these figures were collected for the Evaluation Report prior to all the results being recorded by the Register; therefore, the actual figures could be higher. Overall, workloads were manageable but Reporting requirements were problematic: data collection was dependent on doctors completing forms, tracking and completion rates of forms were inadequate, accuracy of completion was unclear, and inability to link to state population registers for cancer greatly inhibited assessment of the major outcomes. This meant that of the 2308 positive cases, colonoscopy results were reported in only 1273 and histopathology reports lagged further behind. Results were not retrieved for 1035 due to poor data collection processes and very poor reporting rates.

On the other hand, quality measures for colonoscopy were promising; bowel preparation was adequate in 89.6%, and the self-reported cecal intubation rate was 99.3%. GPs referred more than 90% of positive cases for colonoscopy.

The pilot was judged a success, primarily because of the initial high participation rate given that no such program had ever been offered before.

The rollout to the national program

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

After a major lobbying effort to increase cabinet-allocated funds, the first phase of the rollout of the national program commenced in 2006 for those turning 55 and 65 years of age in the selected timeframe. The invitation letter recommends individualized triage to specialist if family history, symptoms, personal history, or major illness are present. Dealing with these is funded from usual care, and it is hoped that this advice maximizes the chance that someone who needs colonoscopic surveillance seeks this.

The second phase of the NBCSP commenced on 1 July 2008 and offers testing to people turning 50, 55 or 65 years of age between January 2008 and December 2010. So far, retesting is not being offered to prior invitees except for pilot invitees. The Program is being phased in gradually to help ensure that health services, such as colonoscopy and treatment services, are able to meet any increased demand.

People eligible to participate in the program are receiving an invitation with an FIT kit as in the pilot but with an advance notification letter. The FIT kit remains free. Participants with a positive FOBT result are advised to discuss the result with their doctor, who will generally refer them for further investigations, usually a colonoscopy. Thus, only the offer and development of kits is centrally organized. The process of follow up of a positive test proceeds outside of the screening program according to usual care in either the public or private sectors. Some states (Queensland and South Australia) are using nurse pathway coordinators as these have been shown to improve adherence to the evidence-based pathway, especially of cases who return a positive test. Unfortunately, little has been done to improve data acquisition although more efficient and user-friendly data sheets are now available.

Considerable effort has been put into establishing a quality framework (see article by Michael Bourke in this Supplement), and efforts to improve training in colonoscopy are proceeding with full national cooperation.

For those not invited, NHMRC Guidelines1 provide for screening from age 50 years and older but these people will need to pay for the initial screening test. The NBCSP register tracks participants through pathway, offers FITs, provides reminders, monitors results, and makes payments to clinicians for information on events although this had limited success in the Pilot.

The minimum data set covers over 100 items with approximately 25 address outcomes related to success, failure, and quality. Unfortunately, these are not available in real time to pathway coordinators and are very limited in their value for detailed research.

The key outcomes that will become available include:

  • • 
    Cancer or adenomas
    • ○ 
      Mortality
    • ○ 
      Detection rates, stage, site, missed lesion
  • • 
    Screening test/screenee
    • ○ 
      Participation, reminders
    • ○ 
      Attendance at GP clinic
    • ○ 
      Colonoscopy done if test positive
    • ○ 
      Demographics, disability
  • • 
    Colonoscopy/histopathology
    • ○ 
      Findings, quality, complications
    • ○ 
      Action based on findings

Of ongoing concern to the profession is the failure to yet fund and organize rescreening, the failure to make screening available for those aged 65–75 and the generally slow rate of the rollout. It remains unclear as to the adequacy of training capacity as well.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References

Australians have made major contributions to all facets of research and development leading to population screening for bowel cancer, stretching from technology development to population RCTs and health services research. In terms of simple screening tests in a two-step screening strategy, the evidence is overwhelming that FITs improve detection and are more acceptable. FIT-based screening is clearly acceptable to Australians and a national organized screening program is feasible. The NBCSP is here to stay. In terms of benefit for Australians, with full rollout and high uptake by the population, we could easily see the number of cases dying from CRC halved by this strategy. To this will be added the extra-screening benefits of improved diagnosis, improved treatment and improved public awareness, all benefits of other screening programs. CRC screening has progressed from a matter of irrelevance and distaste to Commonwealth government policy in the context of an organized program for all Australians.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. A major health problem
  5. The nature of disease—latency
  6. A simple test for detecting the latent period
  7. Evidence that screening tests are effective
  8. Test acceptability
  9. Cost-effectiveness
  10. Feasibility—the Australian pilot screening program
  11. The rollout to the national program
  12. Conclusions
  13. References
  • 1
    Australian Cancer Network Colorectal Cancer Guidelines Revision Committee. Guidelines for the Prevention, Early Detection and Management of Colorectal Cancer. Sydney: The Cancer Council Australia and Australian Cancer Network, 2005.
  • 2
    Young GP, Levin B, Rozen P. Prevention and Early Detection of Colorectal Cancer. London: WB Saunders, 1996; 384.
  • 3
    Wilson JMG, Jungner G. Principles and practice of screening for disease. WHO Public Health Papers No. 34, 1968.
  • 4
    International Agency for Research on Cancer. Use of screening for cervical cancer. In: Cervix Cancer Screening. IARC Handbooks of Cancer Prevention Vol 10. Lyon: IARC Press, 2005; 11762.
  • 5
    Miles A, Cockburn J, Smith RA, Wardle J. A perspective from countries using organized screening programs. Cancer 2004; 101 (Suppl. 5): 120113.
  • 6
    Young GP, Macrae FA, St John DJB. Clinical methods of early detection: basis, use and evaluation. In: YoungGP, LevinB, RozenP, eds. Prevention and Early Detection of Colorectal Cancer. London: WB Saunders, 1996; 24170.
  • 7
    Deeks JJ. Systematic reviews of evaluations of diagnostic and screening tests. BMJ 2001; 323: 15762.
  • 8
    Australian Institute for Health and Welfare. Cancer in Australia—An Overview. Canberra: Australian Institute of Health and Welfare, 2008.
  • 9
    Begg SJ, Vos T, Barker B, Stanley L, Lopez AD. Burden of disease and injury in Australia in the new millennium: measuring health loss from diseases, injuries and risk factors. Med. J. Aust. 2008; 188: 3640.
  • 10
    Bell JC, McCredie M, Coates MS, Armstrong BK. Trends in colorectal cancer incidence and mortality in New South Wales, 1973. Med. J. Aust. 1997; 166: 17881.
  • 11
    Greegor DH. Occult blood testing for detection of asymptomatic colon cancer. Cancer 1971; 28: 1314.
  • 12
    Dent O, Goulston K, St John DJB, Selby W, Young GP. Guidelines for Screening for Colorectal Cancer. Sydney: Australian Gastroenterology Institute, 1991.
  • 13
    Goulston K, St John DJB, Bokey L et al. Guidelines for Early Detection, Screening and Surveillance for Colorectal Cancer. Sydney: Australian Cancer Society, 1994.
  • 14
    Morson BC. Precancerous conditions of the large bowel. Proc. R. Soc. Med. 1971; 64: 95962.
  • 15
    Hamilton SR. Pathology and biology of colorectal cancer. In: YoungGP, RozenP, LevinB, eds. Prevention and Early Detection of Colorectal Cancer. London: WB Saunders, 1996; 322.
  • 16
    Chapuis PH, Goulston KJ, Dent OF, Tait AD. Predictive value of rectal bleeding in screening for rectal and sigmoid polyps. Br. Med. J. (Clin. Res. Ed.) 1985; 290: 15468.
  • 17
    Macrae FA, St. John DJB. Relationship between patterns of bleeding and Hemoccult sensitivity in patients with colorectal cancers or adenomas. Gastroenterology 1982; 82: 8918.
  • 18
    Pierson RN, Holt PR, Watson RM, Keating RP. Aspirin and gastrointestinal bleeding. Am. J. Med. 1961; 31: 25965.
  • 19
    St John DJB, Young GP, McHutchison JG, Deacon MC, Alexeyeff MA. Comparison of the specificity and sensitivity of Hemoccult and HemoQuant in screening for colorectal neoplasia. Ann. Intern. Med. 1992; 117: 37682.
  • 20
    Ahlquist DA, McGill DB, Fleming JL. Patterns of occult bleeding in asymptomatic colorectal cancer. Cancer 1989; 63: 182630.
  • 21
    Thomas WM, Pye G, Hardcastle JD, Mangham CM. Fecal occult blood screening for colorectal neoplasia: a randomized trial of three days or six days of tests. Br. J. Surg. 1990; 77: 2779.
  • 22
    Young GP, Mandel J, Sung J et al. Guidelines for evaluation of new screening tests: application to colorectal and gastric cancers. Submitted.
  • 23
    Young GP, Allison J. Screening for colorectal cancer. In: YamadaT, AlpersD, KaplowitzN, LaineL, OwyangC, PowellD, eds. Textbook of Gastroenterology, 5th edn. Philadelphia, PA: Lippincott Williams and Wilkins, 2008; 17082.
  • 24
    Sackett DL, Haynes RB, Tugwell P. Clinical Epidemiology. A Basic Science for Clinical Medicine. Boston/Toronto: Little, Brown and Company, 1985; 76126.
  • 25
    Mandel JS, Bond JH, Church TR et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. Minnesota Colon Cancer Control Study. N. Engl. J. Med. 1993; 328: 136571.
  • 26
    Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sondergaard O. Randomised study of screening for colorectal cancer with fecal-occult-blood test. Lancet 1996; 348: 146771.
  • 27
    Hardcastle JD, Chamberlain JO, Robinson MH et al. Randomised controlled trial of fecal-occult-blood screening for colorectal cancer. Lancet 1996; 348: 14727.
  • 28
    Faivre J, Dancourt V, Lejeune C et al. Reduction in colorectal cancer mortality by fecal occult blood screening in a French controlled study. Gastroenterology 2004; 126: 167480.
  • 29
    Young GP, Rose IS, St John DJB. Hem in the gut. I: fate of hemoproteins and the absorption of hem. J. Gastroenterol. Hepatol. 1989; 4: 53745.
  • 30
    Rose IS, Young GP, St John DJB, Deacon MC, Blake D. The effect of ingestion of hemoproteins on fecal excretion of hemes and porphyrins. Clin. Chem. 1989; 35: 22906.
  • 31
    Young GP, St John DJB. Fecal occult blood tests: choice, usage and clinical applications. Clin. Biochem. Rev. 1992; 13: 1617.
  • 32
    Young GP, St John DJB. Selecting an occult blood test for use as a screening tool for large bowel cancer. Front. Gastrointest. Res. 1991; 18: 13556.
  • 33
    McDonald CA, Walls RS, Burford Y, Goulston KJ, Yuen AC. Immunochemical detection of fecal occult blood. Aust. N. Z. J. Med. 1984; 14: 10510.
  • 34
    Young GP, St John DJB, Rose IS, Blake D. Hem in the gut. Part II. Fecal excretion of hem and hem-derived porphyrins and their detection. J. Gastroenterol. Hepatol. 1990; 5: 194203.
  • 35
    Lynch NM, McHutchison JG, Young GP, Deacon M, St John DJB, Barraclough D. Gastrointestinal blood loss from a new buffered aspirin (Ostoprin): measurement by radiochromium and HemoQuantTM techniques. Aust. N. Z. J. Med. 1989; 19: 8996.
  • 36
    Mandel JS, Church TR, Bond JH et al. The effect of fecal occult-blood screening on the incidence of colorectal cancer. N. Engl. J. Med. 2000; 343: 16037.
  • 37
    Mandel JS. Sigmoidoscopy screening probably works, but how well is still unknown. J. Natl. Cancer Inst. 2003; 95: 5713.
  • 38
    Towler BP, Irwig L, Glasziou P et al. Screening for colorectal cancer using the fecal occult blood test, Hemoccult (review). Cochrane Database Syst. Rev. 1998; 2.
  • 39
    Australian Health Technology Advisory Committee. Colorectal Cancer Screening. Canberra: Publications Production Unit, Commonwealth Department of Health and Family Services, 1997.
  • 40
    Young GP, St John DJB, Winawer SJ, Rozen P. Choice of fecal occult blood tests for colorectal cancer screening: recommendations based on performance characteristics in population studies. Am. J. Gastroenterol. 2002; 97: 2499507.
  • 41
    St John DJB, Young GP, McHutchison JG, Deacon MC, Alexeyeff MA. Comparison of specificity and sensitivity of Hemoccult and HemoQuant in screening for colorectal neoplasia. Ann. Intern. Med. 1992; 117: 37682.
  • 42
    St John DJB, Young GP, Alexeyeff MA et al. Evaluation of new occult blood tests for detection of colorectal neoplasia. Gastroenterology 1993; 104: 16618.
  • 43
    Smith A, Young GP, Cole SR, Bampton P. Comparison of a brush-sampling fecal immunochemical test for hemoglobin with a sensitive guaiac-based fecal occult blood test in detection of colorectal neoplasia. Cancer 2006; 107: 21529.
  • 44
    Young GP, Cole S. Which fecal occult blood test is best for colorectal cancer screening? Nat. Clin. Pract. Gastroenterol. Hepatol. 2009; 6: 1401.
  • 45
    Allison JE, Tekawa IS, Ransom LJ, Adrain AL. A comparison of fecal occult blood tests for colorectal cancer screening. NEJM 1996; 334: 1559.
  • 46
    Young GP. Fecal Immunochemical Tests (FIT) vs Office-Based Guaiac Fecal Occult Blood Test (FOBT). Pract. Gastroenterol. 2004; 28: 4656.
  • 47
    Viiala CH, Olynyk JK. Outcomes after 10 years of a community-based flexible sigmoidoscopy screening program for colorectal carcinoma. Med. J. Aust. 2007; 187: 2747.
  • 48
    Walleser S, Griffiths A, Lord SJ, Howard K, Solomon MJ, Gebski V. What is the value of computered tomography colonography in patients screening positive for fecal occult blood? A systematic review and economic evaluation. Clin. Gastroenterol. Hepatol. 2007; 5: 143946.
  • 49
    Mendelson R, Ramsay D, Edwards J, Forbes G. Virtual colonoscopy: issues in implementation. Australas. Radiol. 2005; 49: 814.
  • 50
    Chin MWS, Mendelson RM, Edwards J, Foster NM. Clinical significance of extracolonic findings in a community screening programme. Am. J. Gastroenterol. 2005; 100: 27716.
    Direct Link:
  • 51
    Scott RG, Edwards JT, Fritschi L, Foster NM, Mendelson RM, Forbes GM. Community-based screening by colonoscopy or computed tomographic colonography in asymptomatic average-risk subjects. Am. J. Gastroenterol. 2004; 99: 114551.
    Direct Link:
  • 52
    Macrae FA, Hill DJ, St John DJ, Ambikapathy A, Garner JF. Predicting colon cancer screening behavior from health beliefs. Prev. Med. 1984; 13: 11526.
  • 53
    Weller DP, Owen N, Hiller JE, Willson K, Wilson D. Colorectal cancer and its prevention: prevalence of beliefs, attitudes, intentions and behaviour. Aust. J. Public Health 1995; 19: 1923.
  • 54
    Young G, Cole S, Roosa N et al. Differences between participants and non-participants in colorectal cancer (CRC) screening. Gastroenterology 2004; 126: S1319.
  • 55
    Duncan A, Wilson C, Cole SR, Mikocka-Walus A, Turnbull D, Young GP. Demographic associations with stage of readiness to screen for colorectal cancer. Health Promot. J. Austr. 2009; 20: 712.
  • 56
    Cole SR, Smith A, Wilson C, Turnbull D, Esterman A, Young GP. An advance notification letter increases participation in colorectal cancer screening. J. Med. Screen. 2007; 14: 735.
  • 57
    Cole S, Smith A, Upton J, Young GP. Is single stool sampling as effective as two stool sampling in fecal immunochemical test based screening for colorectal cancer? Gastroenterology 2007; 132: A314.
  • 58
    Cole SR, Young GP. Participation in fecal occult blood test-based screening for colorectal cancer is reduced by dietary restriction. Med. J. Aust. 2001; 175: 1958.
  • 59
    Cole SR, Young GP, Esterman A, Cadd B, Morcom J. A randomized trial of the impact of new fecal hemoglobin test technologies on population participation in screening for colorectal cancer. J. Med. Screen. 2003; 10: 11722.
  • 60
    Multicentre Australian Colorectal-neoplasia Screening (MACS) Group. A comparison of colorectal neoplasia screening tests: a multicentre community-based study of the impact of consumer choice. Med. J. Aust. 2006; 184: 54650.
  • 61
    Scott RG, Edwards JT, Fritschi L, Foster NM, Mendelson RM, Forbes GM. Community-based screening by colonoscopy or computed tomographic colonography in asymptomatic average-risk subjects. Am. J. Gastroenterol. 2004; 99: 114551.
    Direct Link:
  • 62
    Weller D, Thomas D, Hiller J, Woodward A, Edwards J. Screening for colorectal cancer using an immunochemical test for fecal occult blood: results of the first 2 years of a South Australian programme. Aust. N. Z. J. Surg. 1994; 64: 4649.
  • 63
    Macrae FA, Hill DJ, Dent O, Goulston KJ, St John DJ. Colorectal cancer: knowledge and attitudes of doctors in Victoria. Aust. N. Z. J. Med. 1982; 12: 27883.
  • 64
    Cole SR, Young GP, Byrne D, Guy J, Morcom J. Participation in fecal occult blood test-based screening for colorectal cancer is improved by endorsement by the primary care practitioner. J. Med. Screen. 2002; 9: 14752.
  • 65
    Salkeld S, Young G, Irwig L, Haas M, Glasziou P. Cost-effectiveness analysis of screening by fecal occult blood testing for colorectal cancer in Australia. Aust. J. Public Health 1996; 20: 13843.
  • 66
    O'Leary BA, Olynyk JK, Neville AM, Platell CF. Cost-effectiveness of colorectal cancer screening: comparison of community-based flexible sigmoidoscopy with fecal occult blood testing and colonoscopy. J. Gastroenterol. Hepatol. 2004; 19: 3847.
  • 67
    Monitoring and Evaluation Steering Committee. Australia's Bowel Cancer Screening Pilot and Beyond. Final Evaluation Report. Bowel Cancer Screening Pilot. Aust Government, Department of Health and Ageing. October 2005; Screening Monograph No. 6/2005.