Estimated impact and value of blood-based colorectal cancer screening at varied adherence compared with stool-based screening

Abstract

Objective

This analysis estimated the outcomes of triennial blood-based colorectal cancer (CRC) screening at various adherence, including perfect adherence, compared with triennial multi-target stool DNA (mt-sDNA) screening at the reported real-world adherence rate.

Methods

The validated CRC-AIM model simulated a US cohort of average-risk individuals receiving triennial screening with mt-sDNA or blood-based test from ages 45 to 75 years. Modeled specificity and sensitivity were based on reported data. Adherence was set at a real-world rate of 65.6% for mt-sDNA and at 65.6%, relative 10% incremental increases from 65.6%, or 100% for the blood-based test. Costs of mt-sDNA and the blood-based test were based on prices for clinically available tests ($508.87 and $895, respectively). Value-based pricing was estimated at a willingness-to-pay threshold of $100,000.

Results

Both tests resulted in life-years gained (LYG), reduced CRC cases, and reduced deaths versus no screening. With adherence for mt-sDNA set at 65.6% and for blood-based test set at 100%, mt-sDNA resulted in 30% more LYG, 52% more averted CRC cases, and 32% more averted CRC deaths. At reported sensitivity and specificity rates, mt-sDNA at 65.6% adherence dominates (is more effective and less costly) the blood-based test at any adherence. There was no price at which triennial screening with the blood-based test at any adherence was cost-effective compared with mt-sDNA at 65.6% adherence.

Conclusions

Triennial screening with mt-sDNA resulted in better clinical outcomes at a lower cost compared with the modeled blood-based test even at perfect adherence, supporting application of blood-based tests only as a secondary screening option.

PLAIN LANGUAGE SUMMARY

Blood-based colorectal cancer screening has lower diagnostic accuracy, lower clinical and health outcomes, and is more expensive than mt-sDNA, even with perfect blood-based screening participation. Although better than no screening at all, blood-based testing is unlikely to exceed performance of stool-based assessment unless a blood-based test is able to meaningfully detect precancerous growths.

Keywords:

• Liquid biopsy
• colorectal neoplasms
• computer simulation
• early detection of cancer
• feces/chemistry

JEL CLASSIFICATION CODES:

• I10
• I1
• I
• C31
• C3
• C

Introduction

Clinical organizations suggest that average risk individuals aged 45–75 years should be screened with one of the many available colorectal cancer (CRC) screening options^1–3. The benefits of CRC screening with multiple modalities has been convincingly demonstrated in the US population^4,5. Adherence is a critical contributor to the clinical and economic impacts of average-risk CRC screening^6–8. Not surprisingly, screening adherence has been less than perfect and varies across test- and interval-specific strategies because of barriers at multiple levels^9–12. Among available recommended non-colonoscopic options, published real-world adherence to triennial multi-target stool DNA (mt-sDNA; Cologuard*) remains higher than any other guideline-endorsed screening strategy^1,2,13. Screening options that require minimal pre-test preparation and are broadly accessible may improve adherence, motivating efforts to develop blood-based screening tests that are accurate and acceptable. However, studies demonstrate mixed results with respect to the uptake and completion of blood-based CRC screening tests in several clinical scenarios^14–18. Blood-based CRC screening tests are not currently endorsed by CRC screening guidelines¹⁹, and their adherence is unknown. Previous simulation studies modeled comparative effectiveness of endorsed screening tests with a blood-based test assuming either perfect adherence to screening^20,21 or modeled assumed blood-test adherence rates^7,22. None of these studies evaluated the economic impact of differing adherence rates for blood-based screening or an estimated price range at which the blood-based tests are cost-effective compared with current stool-based tests.

In addition to adherence, the potential benefit-to-harm ratio associated with any CRC screening strategy is also dependent on test performance (sensitivity and specificity to detect adenomas and CRC)^6,23. Blood-based tests have lower sensitivity to detect advanced neoplasia, particularly precancerous lesions, than mt-sDNA^24,25. The direct impact of adenoma and precancerous lesion detection, serving as an indicator of cancer prevention, has not been demonstrated across blood-based and stool-based CRC screening modalities. Accordingly, the objective of this analysis was to use a validated simulation model to estimate the clinical and economic outcomes of triennial blood-based CRC screening at various adherence rates – including perfect adherence – compared with triennial mt-sDNA screening at the reported real-world adherence rate.

Methods

CRC-AIM model

Microsimulation models are a means of estimating outcomes, and the impact of interventions on outcomes, in a simulated cohort of individuals as they transition from one health state to another. The Colorectal Cancer and Adenoma Incidence and Mortality Microsimulation Model (CRC-AIM) has been previously described, calibrated to CRC incidence in the US²⁶, cross-model validated to three CRC Cancer Intervention and Surveillance Modeling Network (CISNET) models, and externally validated to the United Kingdom Flexible Sigmoidoscopy Clinical trial^6,27.

The model contains CRC natural history and screening components. The natural history component models the progression of CRC from the adenoma-carcinoma pathway in an unscreened population from birth until age 100 years or death. It matches the US incidence of CRC²⁷. The natural history dynamics of CRC-AIM include functions that estimate the risk of adenoma development by sex and age, location of the adenoma, adenoma growth rates, the probability of an adenoma to transition to preclinical cancer, the potential transition to symptomatic CRC, CRC growth, and CRC stage-specific survival²⁷. The screening component of CRC-AIM contains assumptions related to CRC screening, namely, the performance characteristics of the CRC screening test (i.e. sensitivity, specificity, and associated complications), the frequency of the test, and patient adherence to the frequency of the test⁶.

Outcomes in the current analysis were simulated using CRC-AIM version 2.0.3 in Python 3.8 for a US cohort of 2 million average-risk individuals that received screening with mt-sDNA or blood-based test every 3 years between the ages of 45 and 75 years.

Model assumptions

The assumed performance characteristics (specificity and sensitivity) for mt-sDNA were the same as those used by the US Preventive Services Task Force (USPSTF); characteristics for the Shield^** blood-based test were from results of a patient registrational study^23,25. Specificity, sensitivity for CRC, and sensitivity for large (≥10 mm) adenomas were 0.91, 0.94, and 0.42, respectively, for mt-sDNA and 0.90, 0.83, and 0.13, respectively, for the blood-based test (Supplemental Table S1).

Assumed adherence in the model for mt-sDNA was the published real-world adherence of 65.6%¹³. Since real-world adherence to a blood-based CRC screening test is currently unknown, several adherence scenarios were used for the blood-based test, starting at the published adherence rate for mt-sDNA (65.6%), increasing by relative 10% increments from 65.6% (10% higher = actual 72.2%, 20% higher = actual 78.7%, 30% higher = actual 85.3%, or 40% higher = actual 91.8%) and ending at perfect (100%) adherence.

A diagnostic follow-up colonoscopy is needed after any positive, non-colonoscopy CRC screening test to search for and remove any precancerous adenomas or CRC^1,2. The assumed performance characteristics for colonoscopy were the same as those used by the USPSTF (Supplemental Table S1)²³. Several complications from colonoscopy were modeled that were identical to those used by the USPSTF, where the risk of complications increases by age²³.

In accordance with methods used by the USPSTF, adherence to a follow-up colonoscopy after a positive test was assumed to be 100% for both mt-sDNA and the blood-based test²³. A sensitivity analysis was performed where a conservative published real-world follow-up colonoscopy rate of 71.5%²⁸ was used for mt-sDNA and follow-up colonoscopy was assumed to be 100% for the blood-based test. Adherence to ongoing surveillance colonoscopy after discovery of advanced neoplasia was assumed to be 100% for both stool-based and blood-based tests.

The costs used in this analysis are expressed in US dollars. Cost inputs were $508.87 for mt-sDNA as reported in the Centers for Medicare & Medicaid Services (CMS) 2022 clinical laboratory fee schedule and $895 for the blood-based test as reported by the manufacturer^29,30. Cost inputs for colonoscopy (with and without polypectomy), colonoscopy complications, and CRC medical costs were derived from published literature and are shown in Supplemental Table S2. The analysis was carried out from the payer perspective, wherein costs were valued based on reimbursement rates of commercial and Medicare beneficiaries, excluding any copayments and/or cost-sharing. Costs were inflated to December 2022 values. Utility inputs derived from published literature are also shown in Supplemental Table S2.

Outcomes and analysis

A lifetime horizon was chosen to assess the impact of screening per 1,000 individuals. Clinical outcomes of the analysis were life-years gained (LYG), number of CRC cases and deaths averted by screening, LYG per colonoscopy (represents the benefit-to-burden ratio), and LYG from adenoma or CRC detection compared with no screening. To determine the breakdown of LYG based on adenoma and CRC detection, each simulated individual with extended life-years was further evaluated. If the individual did not develop cancer as a result of screening, the additional life-years were linked to adenoma detection. Alternatively, if cancer was detected, the additional life-years were attributed to CRC detection. The economic value of the triennial blood-based test compared with triennial mt-sDNA was estimated by the incremental cost-effectiveness ratio (ICER), additional costs per case and death averted, and value-based pricing (VBP), which is the maximum price at which a willingness-to-pay (WTP) threshold is reached. The WTP for this analysis was set at $100,000 with secondary analyses set at a WTP of $50,000 and $150,000 per quality-adjusted life-years (QALYs). Details of VBP calculations are described in the Supplemental Methods. QALYs were calculated using the same general framework used by the USPSTF. All costs and QALYs were discounted at 3%³¹.

Since there is a lack of data supporting the interval of blood-based CRC screening, a sensitivity analysis was conducted to estimate clinical outcomes and VBP of the blood-based test received at an interval of every 1 or 2 years. In addition, one-way deterministic sensitivity analyses where adherence, costs, and utilities were varied ±10% of their base case values were conducted. Finally, to show the robustness of economic outcomes of this analysis, probabilistic sensitivity analysis where all parameters were varied simultaneously over 500 iterations were conducted, and the ICERs were calculated accordingly.

Results

Estimated clinical outcomes

mt-sDNA resulted in greater LYG and reduced CRC cases and deaths compared with the blood-based test (Table 1). With adherence for mt-sDNA set at 65.6% and for blood-based test set at 100%, mt-sDNA resulted in 30% more LYG, 52% more averted CRC cases, and 32% more averted CRC deaths (Figure 1). Compared with a triennial blood-based test at adherence rates ranging from 65.5% to relatively 40% higher, triennial mt-sDNA resulted in 31–41% more LYG and 12–14% more LYG per colonoscopy (the benefit-to-burden ratio; Figure 1a), as well as 57–72% more averted CRC cases and 35–45% more averted CRC deaths (Figure 1b). Adenoma detection resulted in 83% of the LYG with mt-sDNA and 64–67% of the LYG with the blood-based test in the various adherence scenarios (Figure 2).

Figure 1. (a) LYG and LYG per COL (benefit-to-burden ratio) and (b) CRC cases and deaths averted per 1,000 individuals. Results are for triennial mt-sDNA at published real-world adherence (65.6%) and a triennial blood-based test at adherence equal to mt-sDNA real-world adherence, increasing adherence relative to mt-sDNA real-world adherence, or at perfect (100%) adherence. Percentages over each column indicate the percent higher outcome with mt-sDNA relative to the blood-based test. Abbreviations: CRC, colorectal cancer; COL, colonoscopy; LYG, life-years gained; mt-sDNA, multi-target stool DNA.

Figure 2. Percentage of LYG per 1,000 individuals resulting from detection of adenomas or CRC cases. Results are for triennial mt-sDNA at published real-world adherence (65.6%) and a triennial blood-based test at adherence equal to mt-sDNA real-world adherence, increasing adherence relative to mt-sDNA real-world adherence, or at perfect (100%) adherence. Abbreviations: CRC, colorectal cancer; LYG, life-years gained; mt-sDNA, multi-target stool DNA.

Table 1. Estimated health outcomes per 1,000 individuals screened for CRC from age 45–75 years.

Test	Screening Adherence	Follow-up COLs	Surveillance COLs	COLs for Symptoms	Total COLs	LYG	LYG per COL^a	CRC Incidence Cases	CRC Mortality Cases
No screening	–	–	–	80	80	–	–	80	37
mt-sDNA	65.6%	767	778	13	1558	306	0.196	28	10
Blood test	65.6% (0% higher)	725	495	25	1245	218	0.175	50	18
	10% higher	748	511	24	1284	223	0.174	49	18
	20% higher	769	528	23	1320	227	0.172	48	17
	30% higher	787	535	23	1345	231	0.172	48	17
	40% higher	798	542	22	1362	234	0.172	47	17
	100%	824	559	21	1404	236	0.168	46	16

Results are for triennial mt-sDNA at published real-world adherence (65.6%) and a triennial blood-based test at adherence equal to mt-sDNA real-world adherence, increasing adherence relative to mt-sDNA real-world adherence, or at perfect (100%) adherence.

Abbreviations. COL, colonoscopy; CRC, colorectal cancer; LYG, life-years gained; mt-sDNA, multi-target stool DNA.

^a Represents the benefit-to-burden ratio.

Estimated economic value-based price

The ICERs were negative in all scenarios, indicating that the mt-sDNA test dominated (was more effective and less costly) blood-based testing at the list price of $895 at any adherence (Supplemental Figure S1). At a WTP threshold of $50,000, $100,000, or $150,000, there is no price at which the triennial blood-based test can be used cost-effectively instead of mt-sDNA (Table 2 and Supplemental Table S3). Among all scenarios, when comparing the additional cost per case and death averted between mt-sDNA and blood-based tests, the range was $185,877–$206,510 more with blood-based tests for cases averted and $277,162–$295,469 more for deaths averted.

Table 2. Estimated economic outcomes (discounted costs and QALYs).

Test	Screening Adherence	Total Costs^a^,b	QALY	Δ Cost	Δ QALY	ICER^c	Expected Total Cost to be Cost-Effective	Non-Screening Cost^d	Incremental Screening Price^e	Expected Times Screening Offered	VBP^f
mt-sDNA	65.6%	$7,165	16.453	REF	REF	–	–	–	–	–	–
Blood test	65.6% (0% higher)	$10,435	16.421	$3,271	−0.0324	−$100,907	$3,923.03	$6,302	−$2,379	4.6	−$515
	10% higher	$10,573	16.422	$3,408	−0.0310	−$109,905	$4,063.49	$6,292	−$2,228	4.8	−$466
	20% higher	$10,688	16.424	$3,523	−0.0298	−$118,070	$4,180.53	$6,281	−$2,101	4.9	−$427
	30% higher	$10,825	16.425	$3,661	−0.0286	−$128,064	$4,306.15	$6,301	−$1,995	5.1	−$395
	40% higher	$10,913	16.426	$3,749	−0.0278	−$134,617	$4,379.76	$6,285	−$1,905	5.2	−$368
	100%	$11,068	16.426	$3,904	−0.0275	−$141,768	$4,411.04	$6,318	−$1,907	5.3	−$359

Results are for triennial mt-sDNA at published real-world adherence (65.6%) and triennial blood-based test at adherence equal to mt-sDNA real-world adherence, increasing adherence relative to mt-sDNA real-world adherence, or at perfect (100%) adherence. WTP threshold is set at $100,000 per QALY.

Abbreviations. COL, colonoscopy; CRC, colorectal cancer; ICER, incremental cost-effectiveness ratio; mt-sDNA, multi-target stool DNA; QALY, quality-adjusted life year; REF, reference; VBP, value-based pricing; WTP, willingness-to-pay.

^a Total costs include initial screening, follow-up colonoscopies and associated complications, and CRC medical costs.

^b 3% annual discount for costs.

^c For all scenarios the ICER is negative, implying that blood-based tests with any adherence are dominated by mt-sDNA.

^d Non-screening costs are equivalent to total costs without the screening costs.

^e Incremental screening price is the difference between the non-screening costs and the expected cost to be cost-effective.

^f At a WTP threshold of $100,000 per QALY.

Sensitivity analyses

Triennial mt-sDNA at published real-world adherence rates for screening and follow-up colonoscopy yielded better health outcomes and lower cost per patient screened when compared with blood-based testing with perfect adherence for screening and follow-up colonoscopy rates (data not shown). Triennial mt-sDNA at published real-world adherence resulted in higher LYG, higher LYG per colonoscopy (benefit-to-burden ratio), more averted CRC cases, and more averted CRC deaths than a blood-based test received every 2 years in all adherence scenarios (Supplemental Table S4). Only a blood-based test received every year and at 40% greater adherence than mt-sDNA or 100% adherence resulted in 2 and 7 higher LYG, respectively, than triennial mt-sDNA but at a trade-off of approximately 50% more (762 and 851, respectively) colonoscopies per 1,000 individuals (Supplemental Table S4). Depending on the adherence rate, there was an incremental cost of $145,492–$172,647 and $265,028–$281,458 for preventing an additional CRC case or death, respectively, with annual or biennial blood-based tests compared with triennial mt-sDNA.

At a WTP threshold of $100,000, a VBP of $21–$67 is estimated for the blood-based test if the frequency is every year and if adherence is at least relatively 20% higher than triennial mt-sDNA (Supplemental Table S5).

One-way sensitivity analyses found that adherence to screening and follow-up colonoscopy, along with the cost of blood test, had the greatest impact on the ICER (Supplemental Figure S2). The probabilistic sensitivity analysis found that the blood-based test was dominated (more costly and less effective) by mt-sDNA in 100% of the runs (Supplemental Figure S3).

Discussion

In this microsimulation study using the validated CRC-AIM modeling platform, triennial screening with either mt-sDNA or a blood-based test resulted in clinical benefits compared with no screening. When estimated outcomes from the two screening strategies were compared using published real-world adherence for mt-sDNA and perfect adherence for the blood-based test, mt-sDNA resulted in more LYG, more averted CRC cases and deaths, and lower aggregate medical expenditures per patient screened. A higher proportion of the LYG with mt-sDNA was attributed to effective CRC prevention (adenoma detection), as opposed to CRC detection, compared with the blood-based test. Thus, although a blood-based test results in positive health outcomes compared with no screening, it should only be considered as a screening option if a patient refuses or is unwilling to have a colonoscopy or stool-based test.

A blood-based CRC screening test may be appealing to some patients who are unable or unwilling to complete other effective endorsed strategies such as colonoscopy or stool-based tests. However, there are limited direct data comparing the adherence of stool-based and blood-based modalities in similar populations or in the same setting. The lack of payer coverage or recommended screening intervals for blood-based tests is attributed to the fact that most adherence studies focus on patients who opt for blood-based tests as a secondary choice after refusing the guideline-recommended screening methods. For example, a recent randomized trial by Coronado et al. found that offering a blood-based screening test increased CRC screening from 13% to 30% among adults who had declined prior CRC screening³². To account for the lack of adherence data for blood-based tests, the current analysis used a range of adherence starting with the published real-world adherence of 65.6% for mt-sDNA¹³ for direct comparison and increasing up to perfect (100%) adherence for the blood-based test. The model estimated that a blood-based test at 3-year intervals with perfect adherence produced inferior CRC outcomes and was more costly than mt-sDNA. Even when substantially reducing the price of the blood-based test to be equal to that of mt-sDNA, screening with a blood-based test was dominated (higher cost with lower effectiveness) compared with mt-sDNA. A previous microsimulation study estimated that perfect adherence to annual screening with a blood-based test would achieve similar numbers of CRC cases and deaths averted to triennial mt-sDNA but with a burden of 50% more colonoscopies and associated serious gastrointestinal events⁷.

The sensitivity analysis revealed that estimated clinical outcomes with mt-sDNA remained superior to the blood-based test when follow-up colonoscopy rates were reduced to published real-world adherence rates and perfect follow-up colonoscopy rates were used for blood-based testing. The published real-world adherence rate of 71.5% for a follow-up colonoscopy after a positive mt-sDNA was conservative compared with the rate of 86% reported by Pickhardt et al.³³ Since there are no recommended intervals for blood-tests other than the CMS proposal for covering a test every 3 years, we further analyzed the outcomes of a blood-based test if offered every 2 years. This indicates that the lower sensitivity of the blood-based test to detect adenomas compared with mt-sDNA is the key factor behind the relatively poorer clinical outcomes, not the interval or adherence. The low performance is one of the stated reasons clinical organizations do not currently endorse blood-based CRC screening tests^2,3.

The use of low sensitivity tests decreases the chances to prevent CRC since adenomas will go undetected and potentially progress to CRC³⁴, which increases the risk of CRC-related death. Previous analyses indicated that a triennial blood-based test may need 30% sensitivity to detect advanced adenomas and 100% adherence to achieve the LYG of triennial mt-sDNA at real-world adherence²². Lieberman and the American Gastroenterological Association CRC Workshop Panel proposed that the potential benchmark for an effective blood-based test would be an advanced adenoma sensitivity of 40–50%³⁵. Two other CRC screening modeling studies evaluated the outcomes of triennial blood-based testing using the minimum test performance thresholds defined by CMS (CRC sensitivity 74% and specificity 90%)^20,21. Both modeling studies demonstrated that the triennial blood-based test at perfect adherence was more costly and less effective than triennial mt-sDNA or annual FIT at perfect adherence, primarily because of the inability of the blood-based test to detect advanced precancerous lesions^20,21. One model determined that a triennial blood-based test would require >90% CRC sensitivity, 80% advanced precancerous lesion sensitivity, 90% specificity, and cost less than $140 to be as cost-effective as annual FIT²⁰. Thus, at present, blood-based CRC screening should not substitute for currently endorsed screening methods.

Limitations

The results of this study should be interpreted in the context of the assumptions. First, there are other noninvasive and invasive screening modalities that have been endorsed by US guidelines, but only one noninvasive stool-based test and no invasive tests were explored in this analysis. Second, this analysis only considered cross-sectional adherence rates; the impact of intermittent or longitudinal adherence on the cost-effectiveness of screening strategies was not explored. Third, while the inputs of this analysis, such as the cost of a blood-based test, were based on a limited number of resources, the model outcomes were tested through scenario and sensitivity analyses, and the results remained robust. Fourth, the costs of the blood-based test were assumed to be the same for commercial insurers and Medicare. It is possible that costs for blood-based testing may be reduced in the future, leading to greater cost-effectiveness. However, the model estimated that there is no price at which the triennial blood-based test would be considered cost-effective at a WTP threshold of $50,000–$150,000 when compared with triennial mt-sDNA. Importantly, until blood-based CRC screening tests are endorsed by the USPSTF, in contrast to stool-based testing, patients will likely be responsible for cost-sharing both for the initial screening test and for a follow-up colonoscopy after a positive test. Another limitation is that the analysis focused on the performance of a specific blood-based test. Emerging blood-based tests and biomarkers exhibit varying performance. Previous analyses underscored that the key determinant for the efficacy of blood-based tests lies in their capacity to prevent CRC cancers by detecting adenoma and pre-cancerous lesions^22,36. However, it is observed that most of the emerging blood-based tests currently fall short of detecting adenomas beyond serendipitous findings³⁷. The performance and accuracy of a blood-based screening test may vary in real-world settings, compared to reported performance in clinical trials, and there is limited data to support the consistency of test performance for multiple rounds of screening. Finally, a next-generation mt-sDNA test has been found to have greater performance than the mt-sDNA performance used in the current analysis, but those performance values were not used in the current analysis because it is not commercially available³⁸. It is anticipated that incorporating the enhanced performance characteristics of the next-generation mt-sDNA would result in similar or greater benefits versus a blood-based test compared with those demonstrated in the current analysis.

Conclusions

Access to multiple CRC screening options provides flexibility to accommodate patient preferences and drive greater engagement, consistent with the well-established perspective adopted by groups such as the USPSTF that the “best test is the one that gets done”³⁹. With the recent attention on blood-based CRC screening, appropriate positioning of these tests relative to other existing, endorsed options requires careful consideration. In this modeling analysis using a validated microsimulation platform, triennial CRC screening with a blood-based screening strategy compared favorably to no screening. However, when blood-based screening was compared with the noninvasive, stool-based mt-sDNA test, the estimated clinical and economic impacts were dominated by the latter option. These analyses emphasize the need for both effective early cancer detection and precancer detection to optimize programmatic CRC screening. Further, the analyses show that even when triennial mt-sDNA at a real-world adherence rate was compared with triennial blood-based screening with perfect adherence, the mt-sDNA strategy resulted in favorable CRC incidence, CRC mortality, LYG, colonoscopy utilization, and per-patient cost outcomes. Consequently, with the reported screening sensitivity for precancer detection, blood-based screening is projected to adversely impact the achievable benefits and burdens of CRC screening if these tests are used as a substitute for existing, higher performing test options. Thus, results from this study support application of blood-based tests only as a secondary option for patients who are not completing endorsed screening strategies, suggesting modification of the above statement to reflect that the “best test is the one that gets done appropriately”.

Transparency

Declaration of funding

Financial support for this study was provided through a contract with Exact Sciences Corporation. The funding agreement ensured the authors’ independence in designing the study, interpreting the data, writing, and publishing the report. Exact Sciences Corporation contributed to the study design, data analysis, interpretation of the data, and writing of the report.

Declaration of financial/other relationships

JBK has received research support from Exact Sciences through a contracted services agreement and is an inventor of Mayo Clinic intellectual property, licensed to Exact Sciences, for which he could receive royalties, paid to Mayo Clinic. AMF has been a consultant for AbbVie, Amgen, Centivo, Community Oncology Association, Covered California, EmblemHealth, Exact Sciences, Freedman Health, GRAIL, Harvard University, Health & Wellness Innovations, Health at Scale Technologies, MedZed, Penguin Pay, Risalto, Sempre Health, the State of Minnesota, U.S. Department of Defense, Virginia Center for Health Innovation, Wellth, and Zansors; has received research support from the Agency for Healthcare Research and Quality, Gary and Mary West Health Policy Center, Arnold Ventures, National Pharmaceutical Council, Patient-Centered Outcomes Research Institute, Pharmaceutical Research and Manufacturers of America, the Robert Wood Johnson Foundation, the State of Michigan, and the Centers for Medicare and Medicaid Services. DWE has a professional service agreement with Exact Sciences serving as an independent contractor to provide guidance on study design and analysis. ABO, VV, CE, and PL are employees of Exact Sciences Corporation.

Author contributions

JBK: conceptualization, writing – review and editing. AMF: conceptualization, writing – review and editing. DWE: conceptualization, writing – review and editing. ABO: conceptualization, methodology, supervision, visualization, writing – review and editing. VV: methodology, data curation, formal analysis, investigation, software, visualization of the study, writing – review and editing. CE: methodology, data curation, formal analysis, investigation, software, visualization of the study, writing – review and editing. PJL: conceptualization, methodology, writing – review and editing. All authors approved the final draft for submission.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Previous presentation

Portions of these data were presented at Digestive Disease Week (DDW) Annual Meeting, Chicago, IL, May 6–9, 2023.

Acknowledgements

Medical writing and editorial assistance were provided by Erin P. Scott, PhD, of Maple Health Group, LLC, funded by Exact Sciences Corporation.

Data availability statement

CRC-AIM demonstrates the approach by which existing CRC models can be reproduced from publicly available information and provides a ready opportunity for interested researchers to leverage the model for future collaborative projects or further adaptation and testing. To promote transparency and credibility of this new model, CRC-AIM’s formulas and parameters are available in the supplementary documents of an open-access publication²⁷.

Notes

** Guardant Health, Inc., Redwood City, CA.