## Abstract

Background

Despite the higher risk of colorectal cancer (CRC) in people with chronic kidney disease, it remains uncertain whether early detection through screening is cost-effective in this setting. We aimed to determine the costs and health benefits of CRC screening in people on dialysis or who have received a kidney transplant.

Methods

Using a government health perspective, three probabilistic Markov models were constructed to compare the cost-effectiveness and cost-utility of annual immunochemical faecal occult blood test (iFOBT) screening against no-screening in a cohort of 1000 patients (age 50–70 years) on dialysis and with kidney transplants. A series of one-way, multi-way and probabilistic sensitivity analyses were conducted to assess the robustness of the model structure and the extent in which the model's assumptions were sensitive to the uncertainties within the input variables.

Results

The incremental cost-effectiveness ratios (ICERs) of CRC screening compared with no-screening were $138 828 per quality-adjusted life year [QALY;$122 977 per life year saved (LYS)], $121 973 per QALY ($ 85 095 per LYS) and $44 790 per QALY ($25 036 per LYS) for dialysis patients not listed on the transplant waiting list, patients on the transplant waiting list and patients with kidney transplants, respectively. The test specificity of iFOBT, the starting age of screening and cancer prevalence were influential factors that determined the overall cost-effectiveness of screening in this setting.

Conclusion

Screening for CRC using iFOBT may reduce cancer-specific mortality in patients on dialysis and with kidney transplants. However, the benefits and costs of screening CRCs in patients on dialysis, especially for those deemed not suitable for transplantation, greatly exceeded the typical thresholds for acceptable cost-effectiveness.

## INTRODUCTION

Colorectal cancer (CRC) will soon surpass lung cancer as the leading cause of death due to cancer in the Western world [1, 2]. Screening for CRC is now standard clinical practice in most developed countries, because there is high quality evidence to suggest early cancer detection through screening and the subsequent resection of pre-cancerous polyps reduces cancer-specific mortality in the general population [3–5]. Faecal occult blood testing (FOBT) is the most commonly used screening tool and has the strongest trial base for mortality benefits in the general population [6–11]. Screening for CRC using immunochemical FOBT (iFOBT) (also known as faecal immunohistochemical test) is also cost-effective [3], reasonably accurate [4, 5], well-tolerated and acceptable by the general community [12, 13].

The costs and consequences of screening in people with chronic kidney disease (CKD) is likely to be different. Screening using iFOBT in CKD may incur a greater risk of false-positive results because of minute gastrointestinal bleeding from dysfunctional platelets secondary to uraemia, a higher risk of colonic angiodysplasia and the greater use of anti-platelets and anti-coagulants during dialysis [14–16]. The harms of subsequent testing may also be greater. People with CKD may be more susceptible to electrolyte derangement during bowel preparation, and more likely to have colonoscopy- and polypectomy-related complications such as sedation-related events, perforation and postpolypectomy bleeding [17, 18]. In addition, over-detection of inconsequential disease may carry financial and health burdens, which will adversely impact the patient's quality of life.

Unlike for the general population, there are no randomized controlled trials of CRC screening in people with CKD, and such trials will probably never be done because of limited feasibility. Previous studies have examined the likely impact of CRC screening in the transplanted setting using modelled analyses [19, 20], but did not assess the benefits and the quality of life impact of screening in other non-transplanted CKD populations. In this study, we aim to estimate the health outcomes [in both life years and quality-adjusted life years (QALYs)] and costs of screening CRC in the dialysis and kidney transplant populations.

## MATERIALS AND METHODS

Using a government payer perspective, three probabilistic Markov models were constructed to simulate the natural history of screening for CRC in hypothetical cohorts of kidney transplant recipients, dialysis patients on the transplant waiting list and dialysis patients who were not on the transplant waiting list (n = 1000). The models included all the potential consequences of screening, the diagnostic procedures and the treatment of CRC in kidney transplant recipients and patients on dialysis. We populated the models using the best available evidence and tested the uncertainties within the input variables and the model structures using one-way and probabilistic sensitivity analyses.

### Model structure

Full details of the model structure are published elsewhere [19]. In brief, three different analyses were conducted to compare the health benefits and costs of iFOBT screening with no screening in the three different populations: dialysis patients not on the transplant waiting list, dialysis patients on the transplant waiting and the kidney transplant populations. Further refinements of the previously published models were made [4] that took into consideration the events that followed transplant failures such as returned to dialysis and the possibility of receiving another live donor or deceased donor transplant. A schematic representation of our model for the listed dialysis population is shown in Figure 1. The screening and diagnostic pathways for dialysis patients not listed on the transplant waiting list and kidney transplant recipients were similar to those on the transplant waiting list. However, we assumed that all dialysis patients not listed on the transplant waiting list would not have the chance of receiving a transplant.

FIGURE 1:

Schematic diagram of screening CRC in dialysis patients on the transplant waiting list.

FIGURE 1:

Schematic diagram of screening CRC in dialysis patients on the transplant waiting list.

Patients had the choice of receiving screening or no screening. Those who received screening and were screened positive underwent colonoscopy. In the no-screening arm, cancers could only be diagnosed clinically. Patients with undiagnosed cancers either survived with the disease undiagnosed, died with the disease undiagnosed or died from other causes. Our model also took into consideration the potential complications of diagnostic colonoscopies such as colonic bleeding and perforations.

Of those diagnosed with CRCs, treatment modality was dependent on the cancer stage at the time of diagnosis. Individuals with localized CRCs without lymph node involvement received surgical management only, whereas those with lymph node involvement received surgical and adjuvant therapy. Adjuvant therapy included chemotherapy, radiotherapy or a combination of both. Patients with widespread metastatic disease received palliative care only.

Patients who were cured from their CRCs were shifted to the surveillance pool, where annual colonoscopies were conducted for the first 2-year post-operative period and then every 5 years hence. Any cancer recurrences were detected through surveillance colonoscopies or clinical diagnoses. Dialysis patients on the waiting list with previously diagnosed CRCs were eligible for transplantation after a cancer-free period of 5 years. Patients without cancers and remained alive at the end of year one returned to the beginning of the screening/no-screening decision node. At the end of each annual cycle, dialysis patients who were not on the waiting list survived on dialysis with or without cancer, or died from cardiovascular disease and/or other causes. Dialysis patients who were on the transplant waiting list survived on dialysis with or without cancers, received a transplant, or died from cardiovascular disease and/or other causes.

Transplanted patients survived with or without cancers, experienced transplant graft failure and returned to dialysis, or died from cardiovascular disease and/or other causes. At the end of each annual cycle, the costs and health benefits of each individual were accrued for the specific health state and the cumulative costs and health benefits were estimated over the lifetime of the patients. We assumed screening stopped at aged 74 years. The model terminated when all patients were deceased. All future costs and benefits were discounted at a rate of 5% per annum. Half-cycle corrections were used and all modelling was conducted using TreeAge Pro Suite 2009 (TreeAge software, Williamstown, MA) and Excel Microsoft®.

### Input parameters for the model

#### Clinical data (Appendix 1)

Information on the age-specific CRC prevalence for the general population [2] was adjusted for dialysis and transplant patients by a factor of 1.25 and 2.5, respectively [21, 22]. The CRC stage distributions at diagnosis [6, 8, 11, 23] and stage-specific survival [7, 9, 23, 24] of screened and unscreened individuals, the incidence of cancer relapse [25–27] and the iFOBT test performance [4] for the transplanted and dialysis populations were extrapolated from published data for the general population. Information on all-cause mortality of dialysis patients, probability of receiving transplant and graft loss were extrapolated from the Australia and New Zealand Dialysis and Transplant Registry (ANZDATA) [21]. Utility-based quality-of-life estimates for the various stages of CRC in CKD were based on a recently published study in the Australian CKD population [28]. The treatment and management of stage-specific CRCs was sourced from current recommended clinical practice guidelines [29].

#### Costs data (Appendix 2)

Only direct health care costs were used in the analyses. Unit costs of CRC screening and treatment were obtained from the Australian Refined Diagnosis Related Groups (AR-DRG), the Australian Medicare Benefits Schedule, Cancer Institute of New South Wales and the Australian Institute of Health and Welfare [3, 30–32].

### Sensitivity analysis

Assumptions were tested over a range of plausible values to assess the robustness of the uncertainties in the model's parameters using sensitivity analysis. Using one-way sensitivity analysis, we identified all the influential variables within the model. Probabilistic sensitivity analysis was also undertaken. Instead of using point estimates for parameter values, this approach assigns a distribution to each model parameter and samples from that distribution using Monte Carlo simulation to estimate the expected value of each option. We used log-normal distributions for relative risks, beta distributions for utility weights and gamma distributions for costs, and sampled over 2000 iterations for each variable of interest.

### Model outcomes

The model outcomes included the average and incremental costs and benefits of CRC screening, measured in life years and QALYs, an estimate of the number of deaths averted, the relative risk of mortality reduction and the incremental cost-effectiveness ratio (ICER) of screening compared with no-screening in all three subpopulations. ICERs are calculated according the following formula:

### Ethics

Ethics approval for this study was not required as no patients directly participated. Clinical parameter estimates for the model were sourced from published literature and from de-identified data from existing data registry.

## RESULTS

### Base-case analysis

In the base-case analysis, we assumed that the screening participation rate was 70% (but varied between a conservative estimate of 40–80% in the sensitivity analysis), and the starting age for screening was at 50 years. Table 1 shows the average, and the incremental costs and benefits of screening CRC among dialysis patients waiting for transplants, dialysis patients not on the transplant waiting list and the kidney transplant populations. A total of 2.62 days of life were saved from screening for CRC compared with no screening in the dialysis patients not on the transplant list. Among those waiting for a deceased donor transplant and those who have already received a kidney transplant, a total of 6.9 and 12.0 days of life, respectively, were saved through screening CRC compared with no screening. Compared with no screening, the ICER for screening CRC in CKD were $138 828 per QALY [$122 977 per life year saved (LYS)] among dialysis patients not on the waiting list, $121 973 per QALY ($ 85 095 per LYS) among patients listed on the transplant waiting list and $44 790 per QALY ($25 076 per LYS) in transplanted patients.

Table 1.

Incremental cost-effectiveness ratios CRC screening in the dialysis and transplant populations (screen start age = 50)

Populations Strategies Average cost per patient ($) Incremental cost per patient ($) Average benefits per patient (LYS) Incremental benefits per patient (LYS) Average benefits per patient (QALYs) Incremental benefits per patient (QALYs) ICER ($/LYS) ICER ($/QALY)
Dialysis patients not on waiting list Screening 396 609 883.0 6.1167 0.00718 5.7033 0.0130 122 976 138 828
No screening 395 726 6.1236 5.6904
Listed dialysis patients Screening 520 531 1619 9.3499 0.0190 6.1507 0.0133 85 095 121 973
No screening 518 912 9.3308 6.1374
Transplanted patients Screening 146 236 834 10.2971 0.0333 8.0457 0.0181 25 036 44 790
No screening 145 401 10.2638 8.0275
Populations Strategies Average cost per patient ($) Incremental cost per patient ($) Average benefits per patient (LYS) Incremental benefits per patient (LYS) Average benefits per patient (QALYs) Incremental benefits per patient (QALYs) ICER ($/LYS) ICER ($/QALY)
Dialysis patients not on waiting list Screening 396 609 883.0 6.1167 0.00718 5.7033 0.0130 122 976 138 828
No screening 395 726 6.1236 5.6904
Listed dialysis patients Screening 520 531 1619 9.3499 0.0190 6.1507 0.0133 85 095 121 973
No screening 518 912 9.3308 6.1374
Transplanted patients Screening 146 236 834 10.2971 0.0333 8.0457 0.0181 25 036 44 790
No screening 145 401 10.2638 8.0275

### Sensitivity analysis

The model was robust to changes in the following variables over the ranges tested in the sensitivity analysis: test sensitivity, the participation rates of screening and the costs of the screening tool. Figure 2 shows the variability of the ICERs with the influential variables being tested over a range of values in the one-way sensitivity analyses across the three CKD populations. The direction of the arrows indicates the direction in which the variables changed between the lowest and the highest estimates shown on either side of the bar. For example, if the prevalence of disease was varied between 1.25 (base-case) to over 5-fold greater than that of the general population, the ICER for screening (compared with no screening) among the dialysis patients listed for transplantation then reduced from $121 937/QALY gained to less than$85 095/QALY gained. Test specificity (and not test sensitivity) was the most influential variable on the overall cost-effectiveness of screening, and was consistent across the three CKD populations. If the test specificity was varied between 99 and 70%, the ICER then varied between $82 871/QALY to over$260 000/QALY. Other variables that are considered to be influential include the discount rates of costs and benefits, the prevalence of disease, the costs of the screening tool (iFOBT) and the starting age of the screening programme.

FIGURE 2:

One-way sensitivity analyses indicating the potential variation in ICER estimates by altering certain inputted variables for the different subpopulations: (A) dialysis patients not on the waiting list, (B) listed dialysis and (C) transplanted patients.

FIGURE 2:

One-way sensitivity analyses indicating the potential variation in ICER estimates by altering certain inputted variables for the different subpopulations: (A) dialysis patients not on the waiting list, (B) listed dialysis and (C) transplanted patients.

Table 2 shows the effects of age and the incremental costs, benefits and the ICER between screening and no screening for CRC among the listed dialysis patients. The apparent improvement in the ICERs comparing screening CRC and no screening is driven predominately by a reduction in the incremental costs rather than an increase in the incremental gains in life years associated with the varying starting ages of screening. Older dialysis patients are more likely to experience premature deaths from cardiovascular disease, resulting in a reduction in the overall costs of maintenance dialysis in both arms and the subsequent screening costs in the screened arm. For example, the incremental costs comparing screening and no screening varied between $1619 and$410, if the starting age of screening was pushed back from 50 to 70 years of age. On the contrary, the incremental gains in quality-adjusted life years varied between 0.013 and 0.0041 QALYs if the starting age for screening was increased from 50 to 70 years.

Table 2.

The incremental costs, benefits and incremental cost-effectiveness ratios comparing screening CRC and no screening among dialysis patients on the transplant waiting list with the varying starting ages

Starting ages Strategies Average costs ($) Incremental costs ($) Average benefits (QALYs) Incremental benefits (QALYs) ICER ($/QALY) 50 No screen 518 911 1619 6.1374 0.0133 121 973 Screen 520 531 6.1507 55 No screen 514 737 1745 5.7528 0.0151 115 622 Screen 516 479 5.7679 60 No screen 515 251 1627 5.3366 0.0133 122 498 Screen 516 878 5.3499 65 No screen 508 610 1341 5.0235 0.0108 123 775 Screen 509 950 5.0434 70 No screen 498 835 410 4.8217 0.0041 100 855 Screen 499 245 4.8258 Starting ages Strategies Average costs ($) Incremental costs ($) Average benefits (QALYs) Incremental benefits (QALYs) ICER ($/QALY)
50 No screen 518 911 1619 6.1374 0.0133 121 973
Screen 520 531 6.1507
55 No screen 514 737 1745 5.7528 0.0151 115 622
Screen 516 479 5.7679
60 No screen 515 251 1627 5.3366 0.0133 122 498
Screen 516 878 5.3499
65 No screen 508 610 1341 5.0235 0.0108 123 775
Screen 509 950 5.0434
70 No screen 498 835 410 4.8217 0.0041 100 855
Screen 499 245 4.8258

### Probabilistic sensitivity analysis

The scatter plots shown in Figure 3 illustrate the incremental costs and health outcomes, and the uncertainties surrounding the mean cost and effect estimates (represented by the orange circles) associated with screening for CRC and no screening among dialysis patients not listed on the transplant waiting list, patients listed on the waiting list and the transplanted patients. The x-axis represents the incremental gains in QALYs, and the y-axis represents the incremental costs of screening over no screening in the three different CKD populations. As would be expected, the cost and effect pairs are located on the northeast (NE) quadrant of the cost-effectiveness plane, with positive incremental costs and effects, indicating that screening is more effective but also more costly than no screening. Compared with no screening, screening for CRC among patients on the transplant waiting list gained on average, a total of 4.9 days of QALYs, whereas screening in the transplanted populations gained on average, a total of 6.9 days of QALYs. For dialysis patients not on the transplant waiting list, a total of 4.7 days of QALYs were gained through screening compared with no screening.

FIGURE 3:

Probabilistic sensitivity analysis for (A) dialysis patients on waiting list, (B) dialysis patients not on waiting list and (C) transplanted patients.

FIGURE 3:

Probabilistic sensitivity analysis for (A) dialysis patients on waiting list, (B) dialysis patients not on waiting list and (C) transplanted patients.

Figure 4 shows the results of the probabilistic sensitivity analyses in the form of the cost-effectiveness acceptability curve (CEAC). A CEAC shows the probability that an intervention is cost-effective compared with the alternative, given the observed data, for a range of maximum monetary values that a decision-maker might be willing to pay for a particular unit change in outcome [33]. Using our example in Figure 4, the CEAC indicates that there is a 90% likelihood that the cost-effectiveness of screening CRC using iFOBT compared with no screening in kidney transplant recipients, is less than $60 000 per QALY gained, i.e. given a willingness to pay of$60 000/QALY, the probability that screening CRC is cost-effective is 0.75. On the other hand, if the willingness to pay threshold is $60 000/QALY, screening for CRC in either group of dialysis patients is not cost-effective. Screening for CRC is cost-effective among dialysis patients on the transplant waiting list and not on the transplant waiting list if the willingness to pay threshold is at least$120 000/QALY and $130 000/QALY, respectively. FIGURE 4: Cost effectiveness acceptability curve. FIGURE 4: Cost effectiveness acceptability curve. ## DISCUSSION Using the current available evidence from the general and the CKD populations, our modelled analyses suggest that screening for CRC using annual iFOBT saves lives from cancer across the various stages of CKD: dialysis patients not listed on the transplant waiting list, dialysis patients on the waiting list and the transplanted patients. However, the incremental cost-effectiveness thresholds of screening compared with no screening varied considerably between dialysis and transplanted patients. Assuming screening starts at age 50, the cost-effectiveness of screening CRC varied from$45 000 per QALY gained in kidney transplant recipients to almost $140 000 per QALY gained among patients on dialysis deemed not suitable for transplantation. The decision to recommend cancer screening in CKD patients is complicated, owing to differences in patient characteristics, patient preferences and also the test properties of the cancer-screening tool in relation to the non-CKD setting. The incremental gains in life years and QALYs due to screening in patients with CKD, particularly among those on dialysis and not listed on the transplant waiting list, are much lower than expected in the general population for a number of reasons. Many people on dialysis will die prematurely from cardiovascular-related events before developing cancer. The benefits gained through screening are driven mainly by a reduction in cancer deaths through treatment of less-advanced screen-detected disease [6, 8, 11, 23], and the absolute risk of dying from competing cardiovascular events among dialysis patients, who are deemed not suitable for transplantation, is at least 5- to 10-fold greater than that of those in the general population [34, 35]. However, for those who have already received a kidney transplant, screening may be worthwhile, largely because of their relatively preserved survival and the heightened risk of cancer owing to on-going immunosuppression. Although our modelled results showed that screening among dialysis patients listed on the transplant waiting list achieved a greater gain in QALYs compared with those not on the list, data of mortality benefits supporting the role of early detection for bowel cancer among dialysis patients whilst waiting for a transplant is lacking and clinical practice guidelines generally do not specify the screening modality and the screening frequency for cancer among those on the transplant waiting list [36]. Besides the differences in life expectancy and prognoses of patients with CKD, the diagnostic accuracies of the screening tool are also likely to be different than in people without CKD. Patients with CKD are more likely to experience gastrointestinal bleeding from dysfunctional platelets secondary to uraemia, use of anti-coagulation during dialysis and antiplatelet agents for primary and secondary cardiovascular risk prevention, which then may affect the overall test specificity of the screening tests [37–39]. We have demonstrated from our decision analytical modelling that test specificity is the most influential variable in the screening algorithm for CRC in CKD. If the test specificity was assumed to be 50%, the overall ICER for screening CRC in the listed dialysis population compared with no screening was over$260 000/QALYs. However, if the false-positive rates are negligible with the test specificity assumed to be >99%, the overall ICER for screening CRC in patients on dialysis waiting for a kidney transplant was reduced to less than \$80 000/QALYs. These findings were similar in the transplanted population.

Estimates of the test performance characteristics for CRC screening are well defined in the general population, but there is limited information about the test accuracy of iFOBT screening in the CKD population. A single diagnostic test study evaluating the positive predictive value (PPV) of iFOBT for clinically important colonic lesions in the pre-dialysis population reported a favourable PPV estimate of 30.2%, and the predictive value appeared to increase as the severity of CKD increased. However, a single estimate of test performance such as PPV is insufficient for decision-making. Estimation of other characteristics such as test specificity, sensitivity and negative predictive values were not obtained due to the lack of follow-up to detect interval cancers and advanced neoplasia [40]. A similar study was performed on maintenance dialysis patients using guaiac faecal occult blood tests. The overall positivity rate was at least 20% [41]. Gastritis, arteriovenous malformations and colonic polyps were the most common lesions on diagnostic gastroscopy and colonoscopy after positive screening tests. Similarly, the lack of adequate follow-up time precluded an accurate estimation of the test performance characteristics of the screening tests [41].

In addition to the physical and psychological harms associated with screening CRC in CKD, including anxiety created from knowledge of positive screen results, over diagnosis is a major issue in CRC screening [42]. In our current model, we assumed that the natural history of CRC in CKD follows the typical adenoma-dysplasia and carcinoma sequence as observed in the general population. Previous observational studies and randomized, controlled trials of screening in the general population suggested that the removal of adenomatous polyps using colonoscopic polypectomies reduces the cumulative incidence of CRC by at least 66% [43, 44]. Among those with CKD, the clinical significances of adenomatous and non-adenomatous lesions are unclear. It is likely that long-term immunosuppressant use may accelerate the progression of pre-malignant adenomatous lesions to cancer, but so far, no published evidence has shown convincing and definitive causative effects of immunosuppressant use and cancer risks of the gastrointestinal tracts. Removing polyps in patients with CKD increases the costs and risks (particularly bleeding) of colonoscopy, potentially without genuine clinical benefit. Instead of affecting cancer mortality, such interventions may only benefit an intermediate outcome which does not have the same prognostic significance, particularly in patients with co-existing co-morbidities and reduced expected life expectancies.

Our study has several strengths. To our knowledge, this is the first modelled analysis, using prospectively collected utility-based quality-of-life data from patients with CKD who valued the quality of life of having cancers [28], to assess the direct health care costs and benefits of CRC screening in patients on dialysis and with kidney transplants. Using probabilistic sensitivity analyses, we have taken into consideration the joint uncertainties surrounding the distribution of individual parameter estimates. Our study does have some potential limitations. First, in the absence of trial-based data in the CKD population, the test performance characteristics of the screening tool, the stage distribution of disease and the cancer-specific mortality risks in the screened and unscreened arms were sourced from the general population. Therefore, estimates used in the current model may over-estimate the benefits experienced by our patients, especially those with more severe co-morbid diseases. Given the lack of cancer-specific mortality data among the dialysis patients not on the transplant waiting list, we have extrapolated mortality information from all dialysis patients. We may have overestimated the gains in life expectancy associated with screening among the dialysis populations deemed not suitable on the waiting list. We have also not taken into consideration patient preferences and perspectives in our modelling. Understanding patients' preferences in shared-decision making for intervention such as cancer screening is particularly relevant for our CKD population. In the absence of trial-based evidence of harms and benefits of early detection for cancer in CKD, patients' preferences for the outcomes of treatment are central to determining the optimal screening strategies. Recently published surveys and reviews reported that CKD patients may be aware of the heightened risk of cancer after transplantation and on dialysis, but early detection of the disease through screening is not a major health priority because of their pre-occupation with other more imminent issues such as their graft function [45, 46]. Finally, we had not allowed co-morbidities such as cardiovascular and diabetes with cancer in the model, which may potentially affect the overall mortality benefits through screening in the dialysis and transplant populations.

## CONCLUSIONS

Although screening using iFOBT in kidney transplant patients and those on dialysis appeared to save lives from cancers, the benefits, harms and costs of screening in patients with CKD are subject to variability of the test specificity of screening iFOBT, life expectancy of the patients, patients' preferences for screening and the prevalence of cancer in CKD. Future studies should be conducted to address these major evidence gaps to ensure screening CRC is effective, efficient and safe in patients with CKD.

## AUTHORS’ CONTRIBUTION

GW designed and performed the research, analysed the data analysis and wrote the manuscript. ML designed and performed the research, analysed the data, and contributed to the writing of the paper. KH advised on performance of the research and data analysis and contributed to writing of paper. DH designed and performed the research and also analysed data. JRC advised on the performance of the research and revised the manuscript. MB advised on the performance of the research and revised the manuscript. RT advised on the statistical analyses of the research and revised the manuscript. AT advised on the performance of the research and revised the manuscript. JCC participated in data analysis and writing of the paper.

## CONFLICT OF INTEREST STATEMENT

None declared.

(See related article by Kiberd. Colorectal cancer screening in kidney disease patients: working Backwards. Nephrol Dial Transplant 2013; 28: 774–777.)

## REFERENCES

1
National Cancer Institute
Surveillance Epidemiology and End Results Stat Fact Sheets: Colon and Rectum Cancer

2
Australian Government Australian Institute of Health and Welfare (AIHW)
Cancer Incidence Data Cubes
2005

http://www.aihw.gov.au/cancer/data/datacubes/dex.cfm (6 December 2010, date last accessed)
3
O'Leary
BA
Olynyk
JK
Neville
AM
, et al.  .
Cost-effectiveness of colorectal cancer screening: comparison of community-based flexible sigmoidoscopy with fecal occult blood testing and colonoscopy
J Gastroenterol Hepatol
,
2004
, vol.
19
(pg.
38
-
47
)
4
Launoy
GD
Bertrand
HJ
Berchi
C
, et al.  .
Evaluation of an immunochemical fecal occult blood test with automated reading in screening for colorectal cancer in a general average-risk population
Int J Cancer
,
2005
, vol.
115
(pg.
493
-
496
)
5
Zhu
MM
Xu
XT
Nie
F
, et al.  .
Comparison of immunochemical and guaiac-based fecal occult blood test in screening and surveillance for advanced colorectal neoplasms: a meta-analysis
J Dig Dis
,
2010
, vol.
11
(pg.
148
-
160
)
6
Hardcastle
JD
Chamberlain
JO
Robinson
MHE
, et al.  .
Randomised controlled trial of faecal-occult-blood screening for colorectal cancer
Lancet
,
1996
, vol.
348
(pg.
1472
-
1477
)
7
Hewitson
P
Glasziou
P
Watson
E
, et al.  .
Cochrane systematic review of colorectal cancer screening using the fecal occult blood test (Hemoccult): an update
Am J Gastroenterol
,
2008
, vol.
103
(pg.
1541
-
1549
)
8
Kronborg
O
Fenger
C
Olsen
J
, et al.  .
Randomised study of screening for colorectal cancer with faecal-occult-blood test
Lancet
,
1996
, vol.
348
(pg.
1467
-
1471
)
9
Mandel
JS
Bond
JH
Church
TR
, et al.  .
Reducing mortality from colorectal cancer by screening for fecal occult blood
N Engl J Med
,
1993
, vol.
328
(pg.
1365
-
1371
)
10
Quintero
E
Castells
A
Bujanda
L
, et al.  .
Colonoscopy versus fecal immunochemical testing in colorectal-cancer screening
N Engl J Med
,
2012
, vol.
366
(pg.
697
-
706
)
11
Lindholm
E
Brevinge
H
Haglind
E
Survival benefit in a randomized clinical trial of faecal occult blood screening for colorectal cancer
Br J Surg
,
2008
, vol.
95
(pg.
1029
-
1036
)
12
Almog
R
The public prefers fecal occult blood test over colonoscopy for colorectal cancer screening
Eur J Cancer Prev
,
2008
, vol.
17
(pg.
430
-
437
)
13
Hawley
ST
Volk
RJ
Krishnamurthy
P
, et al.  .
Preferences for colorectal cancer screening among racially/ethnically diverse primary care patients
Med Care
,
2008
, vol.
46
(pg.
S10
-
S6
)
14
Saeed
F
Agrawal
N
Greenberg
E
, et al.  .
Lower gastrointestinal bleeding in chronic hemodialysis patients
Int J Nephrol
,
2011

2011:Article ID 272535, 8 pages, doi:10.4061/2011/272535
15
Zuckerman
GR
Cornette
GL
Clouse
RE
, et al.  .
Upper gastrointestinal bleeding in patients with chronic renal failure
Ann Int Med
,
1985
, vol.
102
(pg.
588
-
592
)
16
Hiremath
S
Holden
RM
Fergusson
D
, et al.  .
Antiplatelet medications in hemodialysis patients: a systematic review of bleeding rates
Clin J Am Soc Nephrol
,
2009
, vol.
4
(pg.
1347
-
1355
)
17
Heher
EC
Thier
SO
Rennke
H
, et al.  .
Adverse renal and metabolic effects associated with oral sodium phosphate bowel preparation
Clin J Am Soc Nephrol
,
2008
, vol.
3
(pg.
1494
-
1503
)
18
Wechsler
A
Schneider
R
Sapojnikov
M
, et al.  .
Bowel cleansing in patients with chronic renal failure—an often overlooked hazard
Nephrol Dial Transplant
,
2006
, vol.
21
(pg.
1133
-
1134
)
19
Wong
G
Howard
K
Craig
JC
, et al.  .
Cost-effectiveness of colorectal cancer screening in renal transplant recipients
Transplantation
,
2008
, vol.
85
(pg.
532
-
541
)
20
Kiberd
BA
Screening for prostate, breast and colorectal cancer in renal transplant recipients
Am J Transplant
,
2003
, vol.
5
(pg.
619
-
625
)
21
Australian and New Zealand Dialysis and Transplant Registry
The 33rd Annual Report
2010
ANZDATA Registry
22
Vajdic
CM
McDonald
SP
McCredie
MRE
, et al.  .
Cancer incidence before and after kidney transplantation
JAMA
,
2006
, vol.
296
(pg.
2823
-
2831
)
23
Mandel
JS
Church
TR
Ederer
F
, et al.  .
Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood
J Natl Cancer Inst
,
1999
, vol.
91
(pg.
434
-
437
)
24
O'Connell
JB
Maggard
MA
Ko
CY
Colon cancer survival rates with the new American Joint Committee on Cancer sixth edition staging
J Natl Cancer Inst
,
2004
, vol.
96
(pg.
1420
-
1425
)
25
Gunderson
LL
Sosin
H
Levitt
S
Extrapelvic colon—areas of failure in a reoperation series: Implications for adjuvant therapy
Int J Radiat Oncol Biol Phys
,
1985
, vol.
11
(pg.
731
-
741
)
26
Lee
YTM
Local and regional recurrence of carcinoma of the colon and rectum: I. Tumour-host factors and adjuvant therapy
Surg Oncol
,
1995
, vol.
4
(pg.
283
-
293
)
27
Sargent
DJ
Wieand
HS
Haller
DG
, et al.  .
Disease-free survival versus overall survival as a primary end point for adjuvant colon cancer studies: individual patient data from 20 898 patients on 18 randomized trials
J Clin Oncol
,
2005
, vol.
23
(pg.
8664
-
8670
)
28
Wong
G
Howard
K
Chapman
J
, et al.  .
How do people with chronic kidney disease value cancer-related quality of life?
Nephrology
,
2012
, vol.
17
(pg.
32
-
41
)
29
Australian Government National Health and Medical Research Council
Clinical Practice Guidelines for the Prevention, Early Detection and Management of Colorectal Cancer
2005

30
Australian Government Australian Institute of Health and Welfare (AIHW)
Australian Refined Diagnosis Related Groups Data Cubes from 1998–99 to 2007–08, Version 5.0/ 5.1

31
Australian Government Department of Health and Ageing
Medicare Benefits Schedule Book
,
2006
Canberra ACT
AIHW
32
Cancer Institute NSW
eviQ Cancer Treatments Online – Colorectal Chemotherapy Protocols

33
Fenwick
E
Claxton
K
Sculpher
M
Representing uncertainty: the role of cost-effectiveness acceptability curves
Health Econ
,
2001
, vol.
10
(pg.
779
-
787
)
34
Anavekar
NS
Pfeffer
MA
Cardiovascular risk in chronic kidney disease
Kidney Int
,
2004
, vol.
66

Suppl 92
(pg.
S11
-
S15
)
35
Foley
RN
Parfrey
PS
Samak
MJ
Epidemiology of cardiovascular disease in chronic renal disease
J Am Soc Nephrol
,
1998
, vol.
9

12 Suppl
(pg.
S16
-
S23
)
36
Batabyal. ea
Clinical practice guidelines on waitlisting for kidney transplantation: consistent and equitable?
Transplantation
,
2012
, vol.
94
(pg.
703
-
713
)
37
Sawhney
MS
McDougall
H
Nelson
DB
, et al.  .
Fecal occult blood test in patients on low-dose aspirin, warfarin, clopidogrel, or non-steroidal anti-inflammatory drugs
Dig Dis Sci
,
2010
, vol.
55
(pg.
1637
-
1642
)
38
Levi
Z
Rozen
P
Hazazi
R
, et al.  .
Sensitivity, but not specificity, of a quantitative immunochemical fecal occult blood test for neoplasia is slightly increased by the use of low-dose aspirin, NSAIDs, and anticoagulants
Am J Gastroenterol
,
2009
, vol.
104
(pg.
933
-
938
)
39
Brenner
H
Tao
S
Haug
U
Low-dose aspirin use and performance of immunochemical fecal occult blood tests
JAMA
,
2010
, vol.
304
(pg.
2513
-
2520
)
40
Bini
EJ
Kinkhabwala
A
Goldfarb
DS
Predictive value of a positive fecal occult blood test increases as the severity of CKD worsens
Am J Kidney Dis
,
2006
, vol.
48
(pg.
580
-
586
)
41
Akmal
M
The prevalence and significance of occult blood loss in patients with predialysis advanced chronic renal failure (CRF), or receiving dialytic therapy
Clin Nephrol
,
1994
, vol.
42
(pg.
198
-
202
)
42
Robinson
MHE
Hardcastle
JD
Moss
SM
, et al.  .
The risks of screening: data from the Nottingham randomised controlled trial of faecal occult blood screening for colorectal cancer
Gut
,
1999
, vol.
45
(pg.
588
-
592
)
43
Winawer
SJ
Zauber
AG
Ho
MN
, et al.  .
Prevention of colorectal cancer by colonoscopic polypectomy
N Engl J Med
,
1993
, vol.
329
(pg.
1977
-
1981
)
44
Quintero
E
Colonoscopy versus fecal immunochemical testing in colorectal-cancer screening
N Engl J Med
,
2012
, vol.
366
(pg.
697
-
706
)
45
Williams
NC
Tong
A
Howard
K
, et al.  .
Knowledge, beliefs and attitudes of kidney transplant recipients regarding their risk of cancer
Nephrology
,
2012
, vol.
17
(pg.
300
-
306
)
46
Howell
M
Tong
A
Wong
G
, et al.  .
Important outcomes for kidney transplant recipients: a nominal group and qualitative study
Am J Kidney Dis
,
2012
, vol.
60
(pg.
186
-
196
)
47
Australian Government Australian Institute of Health and Welfare (AIHW)
National Bowel Cancer Screening Program: annual monitoring report 2009 data supplement 2010
2010

http://www.aihw.gov.au/publications/dex.cfm/title/12136 (14 December 2010, date last accessed)
48
Bauer
FW
Robbins
SL
An autopsy study of cancer patients
JAMA
,
1972
, vol.
221
(pg.
1471
-
1474
)
49
Burton
EC
Troxclair
DA
Newman
WP
Autopsy diagnoses of malignant neoplasms
JAMA
,
1998
, vol.
280
(pg.
1245
-
1248
)
50
Papaconstantinou
HT
Sklow
B
Hanaway
MJ
, et al.  .
Characteristics and survival patterns of solid organ transplant patients developing & lt;i& gt;De Novo& lt;/i& gt; colon and rectal cancer
Dis Colon Rectum
,
2004
, vol.
47
(pg.
1898
-
1903
)
51
Buell
JF
Papaconstantinou
HT
Skalow
B
, et al.  .
De novo colorectal cancer: five-year survival is markedly lower in transplant recipients compared with the general population
Transplant Proc
,
2005
, vol.
37
(pg.
960
-
961
)
52
Lüning
T
Keemers-Gels
M
Barendregt
W
, et al.  .
Colonoscopic perforations: a review of 30 366 patients
Surg Endosc
,
2007
, vol.
21
(pg.
994
-
997
)
53
Mai
C-M
Wen
C-C
Wen
S-H
, et al.  .
Iatrogenic colonic perforation by colonoscopy: a fatal complication for patients with a high anesthetic risk
Int J Colorectal Dis
,
2010
, vol.
25
(pg.
449
-
454
)
54
Jansman
FGA
Postma
MJ
van Hartskamp
D
, et al.  .
Cost–benefit analysis of capecitabine versus 5-fluorouracil/leucovorin in the treatment of colorectal cancer in the Netherlands
Clin Ther
,
2004
, vol.
26
(pg.
579
-
589
)
55
Cass
A