Cost-Effectiveness of mt-sDNA versus Mailed FIT Outreach for Medicare Advantage Enrollees using the CRC-AIM Microsimulation Model

Background: Despite proven effectiveness in reducing Colorectal Cancer (CRC) cases, screening for CRC remains underutilized. As the use of an outreach program may increase uptake and adherence to screening, we examined the cost-effectiveness of provider-ordered multi-target stool DNA (mt-sDNA) versus an outreach program with a mailed Fecal Immunochemical Test (FIT), in a Medicare Advantage population. Methods: Annual FIT, mailed as part of an outreach program, and triennial mt-sDNA were compared using the validated CRC-AIM microsimulation model over a lifetime horizon. Costs and clinical outcomes of 2 million average-risk individuals, free of diagnosed CRC at age 40, who initiated CRC screening at age 65, were simulated. Test sensitivity and specificity inputs were based on the 2021 United States Preventive Services Task Force modeling study; adherence rates were based on published real-world data for stool-based tests and follow-up colonoscopies (COLs). The cost of the outreach program, screening tests, COLs, complications, and CRC care were included. Discount rates (3%) and the perspective of Medicare Advantage were employed. Outcomes are reported in Life-Years Gained (LYG), Quality-Adjusted Life Years (QALY), Incidence Reduction (IR), mortality reduction (MR) and the incremental cost-effectiveness ratio. Scenario analyses explored alternate adherence rates. Results: In the base case, using real-world adherence rates, mt-sDNA had higher IR, MR and LYG; mt-sDNA was cost-effective at a willingness-to-pay threshold of $50,000 per QALY compared to mailed FIT outreach. In scenario analyses, when assuming 100% adherence for both screening test and follow-up COL, mt-sDNA was dominated by mailed FIT outreach. mt-sDNA remained cost-effective versus outreach with mailed FIT in a scenario where FIT had 20% higher adherence than the base case. Conclusions: Adherence to CRC screening and follow-up COL greatly impacts clinical and cost-effectiveness outcomes. Future analyses should consider evidence-based, health plan-specific data to accurately reflect outcomes in order to aid payers in decision making.g.


Introduction
Colorectal Cancer (CRC) is the second most common cause of cancer deaths [1]. The United States Preventive Services Task Force (USPSTF) recommends screening for CRC among average-risk adults beginning at age 45 [2] as screening has long been established in reducing both the incidence of and death from CRC [3]. While recommended CRC screening use has increased since 2000 [4], rates remain below the National Colorectal Cancer Roundtable goal of screening 80% of eligible adults [5]. Despite studies estimating that increasing screening by 10 percentage points would reduce the number of deaths from CRC by 15 percentage points [6], CRC screening remains underutilized [2,7].
Multiple options for CRC screening are supported by evidence of net benefit, including stool-based tests [2]. Stoolbased tests have multiple advantages: they are non-invasive; they are accessible in resource-constrained settings and they do not rely on traveling to a health centre to perform. Two commonly used stool-based tests include Fecal Immunochemical Test (FIT) and multi-target stool DNA (mt-sDNA). FIT is recommended to be performed annually while mt-sDNA is recommended to be repeated every three years [8] and includes patient navigation with every test [9], which has been found to be a key factor in increasing adherence [10]. While patient navigation is not routinely included with FIT tests, mailed outreach programs have been found to increase use and adherence to FIT among eligible screeners enrolled in Medicare Advantage [11].
CRC screening varies based on the type of insurance coverage, including Medicare with private insurance versus Medicare with no supplementary insurance or Medicare and Medicaid [12]. Medicare Advantage is provided through providesector health insurers and was chosen by nearly 40% of Medicare enrollees in 2020 [13]. Despite the inclusion of preventive health services within the Medicare Advantage program, CRC screening remains underutilized by enrollees. To better understand the health and economic consequences of mt-sDNA versus a FIT mailed outreach program, this study examined the cost-effectiveness of these stool-based screening tests in a Medicare Advantage population due for screening using a microsimulation model.

Overview
This analysis examined the cost-effectiveness of mt-sDNA versus a FIT mailed outreach program from the perspective of Medicare Advantage. The validated CRC-AIM model was used [14,15], and included average-risk US adults aged 65 to 75 (Medicare Advantage eligible years for screening). Individuals were assumed to be due for CRC screening; that is, they were previously screened and it had been one year since their previous negative FIT test, 5 years after a previous normal sigmoidoscopy and/or 10 years after a previous normal colonoscopy. Screening modalities included in the model were either annual FIT, or triennial mt-sDNA until age 75, as per the latest guidelines [2]. Individuals who did not undergo follow-up colonoscopy in the model were assumed to be non-adherent until they become symptomatic. A lifetime horizon was chosen to assess the impact of screening; costs and Quality-Adjusted Life Years (QALYs) were discounted at 3% [16].

Inputs
Screening strategies performance parameters were based on the 2021 USPSTF modeling study (Supplementary Table  1) [2]. Cost inputs were based on Medicare fee for service and were inflated to November 2021 using the Medical Care Services component of the Consumer Price Index (Supplementary Table 2) [17]. FIT does not include the cost of an outreach program, thus, the cost of outreach was taken from a randomized controlled trial, which consisted of a mailed postcard and call, followed by a mailed FIT kit, and then a reminder phone call if the kit was not returned [18]. As patient navigation is included with every mt-sDNA, no additional cost for outreach was necessary. Utility inputs for patients not experiencing colonoscopy complications or utility loss due to complications were included at baseline, adjusted for age, and based on general US population utility values (Supplementary Table 3) [19]. Utility loss for all colonoscopies and complications from colonoscopy were included as utility decrements per event [20]. Utility loss for CRC was stratified by the level of CRC care and stage of CRC; these utility decrements were applied on a per Volume

Base case analysis
The base case analysis used reported real-world adherence rates for both adherence to initial screening and follow-up colonoscopy after a positive stool-based screening test (Supplementary Table 4). Surveillance and symptom-related colonoscopies had an assumed adherence of 100% and were not varied in sensitivity analyses. Fixed adherence rates are assumed over time.
The primary outcome was clinical benefit reported as incidence reduction, mortality reduction and Life Years Gained (LYG) per 1,000 patients. Secondary outcomes included total costs (screening costs, the cost of CRC-related direct medical costs, and costs associated with colonoscopy complications), QALYs, LYG and the Incremental Cost-Effectiveness Ratio (ICER) of mt-sDNA versus FIT + outreach. Willingness to Pay (WTP) was based on commonly accepted thresholds in the United States (US) [16].

Scenario analyses
In order to explore the impact of adherence on outcomes, three alternate scenarios were considered: • Scenario 1: 100% adherence to both the stool-based screening test and follow-up colonoscopy • Scenario 2: reported real-world adherence for stool-based screening only with 100% adherence to follow-up colonoscopy • Scenario 3: for FIT only, an assumption of 20% higher realworld reported adherence rates for both the screening test and follow-up colonoscopy

Additional analyses
In order to determine the impact of the inputs on results, a probabilistic sensitivity analysis, where all parameters were varied + / -10% of their base case values simultaneously, was undertaken. Threshold analyses were also undertaken to explore the robustness of the adherence inputs on the ICER. In the first threshold analysis, stool-based screening test adherence was varied from 0-100% while adherence to follow-up colonoscopy was held constant at 100%. In the second threshold analysis, adherence to the screening test was held constant at the baseline reported real-world adherence rates, while adherence to follow-up colonoscopy was varied from 0-100%.

Results
In the base case analysis under reported real-world adherence rates for both stool-based screening tests and follow-up colonoscopies, use of mt-sDNA for screening resulted in a greater reduction in CRC incidence and mortality compared to a FIT + outreach program (Figure 1). Screening with mt-sDNA resulted in an incidence reduction of 25% and a mortality reduction of 32%, compared to 10% and 13%, respectively with FIT + outreach. Compared to FIT + outreach, screening with mt-sDNA resulted in greater LYG per 1,000 patients and lower CRC costs (Table 1).

Scenario analyses
The highest reduction in incidence of and mortality from CRC were observed with perfect adherence to stool-based screening and follow-up colonoscopy (Scenario 1). In this scenario, FIT + outreach resulted in higher reduction in incidence and mortality from CRC compared to mt-sDNA ( Table 2). The reduction in incidence and mortality from CRC, along with total QALYs, was lower in magnitude in Scenario 2 compared with Scenario 1 for both mt-sDNA and FIT + outreach when using reported real-world adherence for stool-based screening and perfect (100%) adherence for follow-up colonoscopy. mt-sDNA, however, still resulted in greater reduction in incidence and mortality from CRC compared to FIT (Scenario 2). When increasing the reported real-world adherence of FIT + outreach by 20% for both the stool-based test and follow-up colonoscopy (Scenario 3), mt-sDNA still resulted in greater reduction in incidence of and mortality from CRC as compared to FIT + outreach. Similar patterns were observed for life years gained, with the greatest life years gained observed for FIT + outreach under a perfect adherence scenario (Scenario 1), while mt-sDNA resulted in incrementally more life years gained under Scenario 2 and Scenario 3. Total lifetime costs were highest for both strategies under Scenario 3 and were lowest for mt-sDNA for Scenario 2 and for FIT + outreach under Scenario 1 (   For the base case analysis, while total costs were higher when screening with mt-sDNA, due to increased QALY gains, mt-sDNA was cost-effective versus FIT + outreach with an ICER of $25,065 / QALY. Under both Scenarios 2 and 3, mt-sDNA was cost-effective versus FIT + outreach at a WTP threshold of $50,000 / QALY (Supplementary Figure 1

Additional analyses
The probabilistic analysis found that 99.8% (499/500) of the iterations resulted in an ICER of no more than $50K/QALY for mt-sDNA versus FIT; 100% of the runs resulted in ICERs less than $100K/QALY (Supplementary Figure 2). The threshold analysis found that when varying the stool-based test adherence (while keeping the follow-up colonoscopy adherence rate fixed at 100%), adherence to both stool-based tests would need to be at least 75% for FIT to dominate mt-sDNA ( Figure 2). When varying follow-up colonoscopy adherence rates (keeping adherence to stool-based screening rates fixed at baseline values), mt-sDNA is cost-effective against FIT at all levels of adherence to FIT (

Discussion
Ensuring eligible patients remain up to date with screening is essential to reducing the burden of CRC. Our results found that under reported real-world adherence rates, screening with mt-sDNA resulted in greater incidence and mortality reduction compared to FIT + outreach, in a Medicare Advantage population. While total costs were higher when screening with mt-sDNA due to increased QALY gains, use of mt-sDNA was found to be cost-effective versus FIT + outreach at a WTP threshold of $50,000 / QALY. Similar results were observed with scenario analyses exploring perfect adherence for follow-up colonoscopy as well as when increasing the adherence to FIT + outreach by 20%. These results highlight the need for robust cost-effectiveness analyses, including the use of real-world adherence rates, when determining the value of a test. This is the first study to explore the cost-effectiveness of stool-based tests in a Medicare Advantage population. Medicare Advantage has been found to deliver significantly better quality of care, with better health outcomes [21]. One study found that Medicare Advantage beneficiaries were more likely to receive mammography screening compared with a Medicare fee for service population [22]. Medicare Advantage may be considered to be more efficient than Medicare fee-for-service, as it offers improved disease management, care coordination and supplemental benefits, all factors that may impact the uptake of CRC screening. Despite these differences, this analysis extends the generalizability of the results observed with mt-sDNA in other insured populations [23,24].
Multiple factors affect adherence to CRC screening. The top barriers to screening identified among respondents who had previously used stool-based screening options were lack of provider recommendations and lack of knowledge [25]. In order to address these barriers and increase adherence to screening, different outreach strategies have been proposed; these organized outreach programs have been shown to be effective to various degrees [26]. The results of this cost-effectiveness analysis are based on a mailed outreach program for FIT that included eligible Medicare Advantage patients receiving a letter containing information about CRC screening, along with a FIT kit where adherence to screening with outreach was found to be 29.0% [11]. This adherence is lower than the 69.8% reported among Medicare Advantage patients using mt-sDNA, which may be due to the fully integrated patient navigation for completion of the stool-based test which is integrated with every mt-sDNA test, including the universal outreach available in over 240 languages [9]. While a previous study in an integrated healthcare delivery system (Kaiser Permanente Northern California) found that the initiation of an organized CRC screening program significantly increased the up to date screening status from 39.0% to 82.7% over 15 years (p<0.01) [27], a claims analysis of multiple health plans found that only 23.4% of patients were adherent to FIT at year 2, and 10.6% were adherent at year 3 [28]. Outreach strategies to increase adherence to screening, including follow-up colonoscopy, need to be explored further, to determine the barriers and facilitators of screening across various populations.
Results of this study should be interpreted in the context of the assumptions. This analysis was based on the use of the CRC-AIM microsimulation model, which has previously been validated [29]. In addition, reported population-specific real-world data was used to inform adherence for both stool-based screening tests (mt-sDNA and FIT), along with the respective adherence to follow-up colonoscopies. However, as Medicare Advantagespecific adherence data are not available for the fecal occult blood test, it was excluded from this analysis. Further, only noninvasive stool tests were explored in this analysis. This analysis also only considered cross-sectional adherence rates; the impact of intermittent or longitudinal adherence on the cost-effectiveness of screening strategies was not explored. While the inputs of this analysis were based on a limited number of studies, these were tested through scenario and sensitivity analyses and results remained robust.

Conclusions
Adherence to initial stool-based CRC screening tests and follow-up colonoscopies impact both clinical and costeffectiveness outcomes. The choice of test and patient outreach can favourably impact patient adherence, thus further contributing towards the goal of attaining 80% CRC screening adherence. Future analyses should consider population-specific real-world evidence, including those living in supportive care facilities and those with chronic conditions, as well as comparisons with colonoscopies, in order to aid payer decision making.

Funding Source
This study was funded by Exact Sciences Corporation. Volume