Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Contemp Clin Trials . Author manuscript; available in PMC: 2026 Apr 11. Published in final edited form as: Contemp Clin Trials. 2025 Dec 16;161:108188. doi: 10.1016/j.cct.2025.108188 Search in PMC Search in PubMed View in NLM Catalog Add to search Sequential choice vs colonoscopy outreach for colorectal cancer screening: Design and rationale of a pragmatic randomized clinical trial Shivan J Mehta Shivan J Mehta a Perelman School of Medicine, University of Pennsylvania, United States of America b Center for Health Care Transformation & Innovation, University of Pennsylvania, United States of America Find articles by Shivan J Mehta a, b, * , Pamela A Shaw Pamela A Shaw a Perelman School of Medicine, University of Pennsylvania, United States of America c Biostatistics Division, Kaiser Permanente Washington Health Research Institute, United States of America Find articles by Pamela A Shaw a, c , Catherine Reitz Catherine Reitz a Perelman School of Medicine, University of Pennsylvania, United States of America b Center for Health Care Transformation & Innovation, University of Pennsylvania, United States of America Find articles by Catherine Reitz a, b , Caitlin Brophy Caitlin Brophy a Perelman School of Medicine, University of Pennsylvania, United States of America b Center for Health Care Transformation & Innovation, University of Pennsylvania, United States of America Find articles by Caitlin Brophy a, b , Evelyn Okorie Evelyn Okorie a Perelman School of Medicine, University of Pennsylvania, United States of America b Center for Health Care Transformation & Innovation, University of Pennsylvania, United States of America Find articles by Evelyn Okorie a, b , Keyirah Williams Keyirah Williams a Perelman School of Medicine, University of Pennsylvania, United States of America b Center for Health Care Transformation & Innovation, University of Pennsylvania, United States of America Find articles by Keyirah Williams a, b , Abraham Segura Abraham Segura a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by Abraham Segura a , Jinming Tao Jinming Tao a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by Jinming Tao a , Christopher K Snider Christopher K Snider b Center for Health Care Transformation & Innovation, University of Pennsylvania, United States of America Find articles by Christopher K Snider b , Colin Wollack Colin Wollack a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by Colin Wollack a , Sadie Friday Sadie Friday a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by Sadie Friday a , Katharine A Rendle Katharine A Rendle a Perelman School of Medicine, University of Pennsylvania, United States of America d Penn Center for Cancer Care Innovation, Abramson Cancer Center, University of Pennsylvania, United States of America Find articles by Katharine A Rendle a, d , Tamar Klaiman Tamar Klaiman a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by Tamar Klaiman a , Karen Glanz Karen Glanz a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by Karen Glanz a , Corinne Rhodes Corinne Rhodes a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by Corinne Rhodes a , David A Asch David A Asch a Perelman School of Medicine, University of Pennsylvania, United States of America Find articles by David A Asch a Author information Article notes Copyright and License information a Perelman School of Medicine, University of Pennsylvania, United States of America b Center for Health Care Transformation & Innovation, University of Pennsylvania, United States of America c Biostatistics Division, Kaiser Permanente Washington Health Research Institute, United States of America d Penn Center for Cancer Care Innovation, Abramson Cancer Center, University of Pennsylvania, United States of America * Corresponding author at: Perelman School of Medicine, University of Pennsylvania, 3535 Market St, 15 th floor, Philadelphia, PA 19104, United States of America. [email protected] (S.J. Mehta). Issue date 2026 Feb. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). PMC Copyright notice PMCID: PMC13068070 NIHMSID: NIHMS2163388 PMID: 41412475 The publisher's version of this article is available at Contemp Clin Trials Abstract Background: Colorectal cancer (CRC) screening rates remain limited, and effective methods for offering the choice of colonoscopy or stool testing through outreach have not been identified. We evaluate the effect of sequential choice compared to colonoscopy outreach on screening completion, and further evaluate behavioral nudges in the electronic health record (EHR). Methods: In this pragmatic randomized clinical trial, patients were randomly allocated in a 1:2:2 ratio to 1) usual care (no outreach), 2) colonoscopy only, or 3) sequential choice of colonoscopy, then fecal immunochemical testing (FIT). Patients in arms 2 and 3 were additionally randomized to receive either (a) usual care, or (b) a visit-based, clinician-directed nudge facilitated by the EHR with follow-up texting to the patient. The primary outcome is CRC screening completion within 3 years by either colonoscopy, 2 negative fecal immunochemical tests (FIT), or 1 positive FIT followed by colonoscopy within one year. Analysis: For the patient-directed analysis, the primary outcome will be evaluated by comparing CRC screening completion among patients randomized to either outreach arm (2 or 3) to the no outreach arm (1). We will also compare completion between the colonoscopy only arm (1) and the sequential choice arm (2). For the visit-based analysis, we will compare CRC screening completion among patients between the usual care arms (2a and 3a) and the nudge arms (2b and 3b). Conclusion: This trial is unique in evaluating the long-term effectiveness of offering sequential choice to colonoscopy alone through a multi-level, centralized outreach and visit-based design. Clinical Trials Identifier: NCT05693649 Keywords: Colorectal cancer screening, Colonoscopy, Fecal immunochemical test, Nudge, Electronic health record, Behavioral economics 1. Introduction Colorectal cancer (CRC) is the second leading cause of cancer death in the United States (US). Routine screening has reduced the burden of disease, but national screening still falls below goals, with rates at approximately 67 %. [ 1 – 5 ] There are also persistent racial, ethnic and socioeconomic disparities in outcomes, particularly among Black patients. [ 6 , 7 ] Multi-level approaches that decrease clinician and patient barriers are important for durably increasing screening rates, including both direct population-based outreach along with clinician recommendations during visits. [ 8 ] However, these approach have not been adopted widely and have been difficult to sustain over time. New insights from behavioral science have helped recognize that humans use systematic shortcuts in thinking that thwart long-term health goals. [ 9 – 11 ] For example, present-time bias reduces the motivation to engage in behaviors whose outcomes are not immediate; status quo bias makes past choices the more likely future choices; and choice overload paralyzes decision making, even when stakes are high. [ 12 – 14 ] However, there is evidence that we can harness these same biases to encourage healthy behaviors. [ 15 , 16 ] Reminders can overcome present-time bias; shifting from opt-in to opt-out framing can increase participating in cancer screening; and both can simplify decisions and prevent overload. [ 17 – 21 ] Colonoscopy and fecal immunochemical testing (FIT) are evidence-based and top-tier tests for CRC screening. [ 4 , 22 ] FIT achieves a higher response rate in a single outreach, but it requires annual adherence. [ 23 , 24 ] Colonoscopy is more sensitive for identifying polyps and cancer, but the procedure is costly and burdensome for patients. [ 8 ] There is evidence that offering a choice of testing can boost screening rates, but there is the potential for choice overload. [ 25 , 26 ] Sequential choice that offers colonoscopy first, then FIT to those who decline or defer has the potential to simplify choice, and it is recommended by clinical guidelines, but there is limited evidence to support this approach. [ 27 – 29 ] This trial evaluates a multi-level screening intervention that combines direct outreach and nudges in the electronic health record (EHR) during visits to durably increase screening rates. We incorporate behavioral principles such as opt-out framing and reminders while evaluating the comparative effectiveness of sequential choice compared to colonoscopy only. 2. Methods The aims of the 3-year trial are to evaluate the rates of CRC screening completion between patients with 1) no outreach, 2) colonoscopy only, or 3) sequential choice (colonoscopy, then FIT); and to compare the effect of EHR nudges to clinicians and patients during office visits. The trial was approved by the University of Pennsylvania Institutional Review Board and a waiver of consent was granted. [ 30 ] It was prospectively registered on clinicaltrials.gov ( NCT05693649 ). 2.1. Setting The trial is being conducted over approximately 3.5 years (June 2023 and January 2027) in 24 primary care clinics within a large, urban, academic health system in the Philadelphia region. 2.2. Eligibility and enrollment Using automated data extraction from the EHR database (Epic Clarity), we identified eligible patients at participating primary care practices within 30 miles of a health system endoscopy site. [ 31 ] We included patients aged 50–72 who, at the time of enrollment, were seen by primary care at least once in the past two years, were of average risk of developing CRC, and who were not up-to-date with CRC screening per guidelines at the time. [ 32 ] Patients who were not average risk were excluded, including those with a personal or significant family history of CRC, colonic polyps, Hereditary Nonpolyposis Colorectal Cancer Syndrome (HNPCC/Lynch Syndrome), other gastrointestinal cancer, gastrointestinal bleeding, or inflammatory bowel disease ( Supplement Table 1 ). Additionally, patients who may not benefit from screening were excluded, including those with dementia, metastatic cancer, para/quadriplegia, total colectomy, those receiving hospice or palliative care, or those with an elevated chance of mortality within 3 years ( Supplement Table 2 ). Finally, patients were also excluded if they were uninsured, had an active order for FIT (within the last 60 days) or multitarget stool DNA testing (MT-sDNA), or a positive stool test (FIT or MT-sDNA) result in the last 5 years. We developed a mortality risk indicator to exclude patients who may not benefit from screening. Existing risk indicators included variables such as insurance type that may differentially exclude patients who are already underrepresented in screening but still might benefit. [ 33 , 34 ] Our model uses predictive factors related to end-of-life risk, from domains similar to those represented in the Epic End of Life (EOL) Care Index, with the Medicaid predictor removed, and additional medication, diagnosis, and utilization variables added ( Supplement Table 3 ). This 3-year mortality prediction model was trained on a cohort of 42,451 patients eligible for CRC screening as of 12/1/17, 12/1/18, and 12/1/19 (three years, used to mimic the length of the intervention) using weighted logistic regression with L1 regularization and 5-fold cross validation, which was implemented via LogisticRegressionCV from the scikit-learn library (v1.0.2) in Python. [ 35 ] Of this cohort, 779 (1.8 %) patients were deceased within 3 years. The AUC of this prediction model was 0.844 in the held-out test set, performing similarly to the Epic EOL model (0.838). Using a threshold of 0.9, the model’s accuracy was 97.2 % with a positive predictive value of 26.8 %. Upon completion of enrollment, 625 patients of 19,969 eligible patients (3.1 %) were excluded because of an increased risk for mortality within three years. 2.3. Randomization In each of four batches, 5000 eligible patients were randomly selected from the approximately 20,000 eligible patients identified. Patients were individually randomized, stratified by practice, to three arms related to outreach in a 1:2:2 ratio to no outreach (arm 1), colonoscopy only (arm 2), or sequential choice of colonoscopy then FIT (arm 3). In the outreach arms 2 and 3, patients were additionally randomized to receive either (a) no additional intervention, or (b) a visit-based, clinician-directed nudge facilitated by the EHR and follow-up text messaging ( Fig. 1 ). Upon completion of enrollment, 19,344 patients were randomized as follows: arm 1 (no outreach): 3877; arm 2a (colonoscopy only, no nudge): 3865; arm 2b (colonoscopy + visit nudge): 3870; arm 3a (sequential choice): 3867; arm 3b (sequential choice + visit nudge): 3865. See Table 1 for participant characteristics by study arm. Randomization was implemented using block randomization with stratification using Way to Health (W2H), an NIH-funded software platform that automates research and outreach for patient engagement. [ 36 ] Fig. 1. Open in a new tab Trial design. Table 1. Participant characteristics. Control Colonoscopy Only Colonoscopy + Visit Based Sequential Choice Sequential Choice + Visit Total N 3877 3865 3870 3867 3865 19,344 Age 1 Mean (SD) 58.5 (6.6) 58.6 (6.7) 58.4 (6.5) 58.5 (6.6) 58.6 (6.6) 58.5 (6.6) Sex Female 2176 (56.1) 2221 (57.5) 2204 (57.0) 2211 (57.2) 2143 (55.5) 10,955 (56.6) Male 1701 (43.9) 1644 (42.5) 1666 (43.0) 1656 (42.8) 1722 (44.6) 8389 (43.4) Race 2 Asian 178 (4.6) 173 (4.5) 184 (4.8) 191 (4.9) 203 (5.3) 929 (4.8) Black 1205 (31.1) 1232 (31.9) 1223 (31.6) 1241 (32.1) 1176 (30.4) 6077 (31.4) White 2093 (54.0) 2082 (53.9) 2103 (54.3) 2080 (53.8) 2075 (53.7) 10,433 (53.9) Other 161 (4.2) 159 (4.1) 151 (3.9) 141 (3.7) 179 (4.6) 791 (4.1) Unknown/Refused 240 (6.2) 219 (5.7) 209 (5.4) 214 (5.5) 232 (6.0) 1114 (5.8) Ethnicity Hispanic 127 (3.3) 3632 (94.0) 3649 (94.3) 3642 (94.2) 3628 (93.9) 18,203 (94.1) Not Hispanic 3652 (94.2) 120 (3.1) 133 (3.4) 105 (2.7) 132 (3.4) 617 (3.2) Unknown/Refused 98 (2.5) 113 (2.9) 88 (2.3) 120 (3.1) 105 (2.7) 524 (2.7) Insurance Commercial 2719 (70.1) 2681 (69.4) 2736 (70.7) 2702 (69.9) 2675 (69.2) 13,513 (69.9) Medicaid 414 (10.7) 403 (10.4) 400 (10.3) 427 (11.1) 406 (10.5) 2051 (10.6) Medicare 744 (19.2) 781 (20.2) 734 (19.0) 737 (19.1) 784 (20.3) 3780 (19.5) EMR Portal Status Active 3434 (88.6) 3370 (87.2) 3389 (87.6) 3406 (88.1) 3390 (87.7) 16,989 (87.8) Not Active 443 (11.4) 495 (12.8) 481 (12.4) 461 (11.9) 475 (12.3) 2355 (12.2) Median Household Income 3 Median income (IQR) $93,588 (55,147–118,207) $92,593 (55,979–118,443) $93,588 (54,332–117,331) $93,103 (54,083–118,207) $93,588 (55,979–118,207) $93,588 (55,979–118,207) Previously Completed CRC Screening 4 Any Screening Method 1077 (27.8) 1081 (28.0) 1073 (27.7) 1103 (28.5) 1126 (29.1) 5460 (28.2) Open in a new tab 1 Age at time of enrollment. 2 Race is patient self-reported in the EMR; Asian includes Pacific Islander, Other includes American Indian/Alaska Native or self-identified as Other. 3 Income based on median household income by zip code at time of enrollment using the American Community Survey, 2017–2022. 4 Includes the number of unique patients screened by any method prior to trial enrollment, including: FIT, Colonoscopy, Sigmoidoscopy, Cologuard, or CT Colonography. 2.4. Interventions All patients randomized to an outreach arm (arms 2 and 3) receive centralized direct outreach to complete colonoscopy. Regional physician leads sign bulk order colonoscopy for eligible patients, while the primary care provider (PCP) is listed as the authorizing provider. Patients receive outreach via the electronic patient portal (if active status) or mailed letter informing them their PCP has ordered the procedure, including the colonoscopy order (mailed only) and providing the phone number for direct access to colonoscopy. Outreach is conducted in batches of approximately 4000 patients every 2 months to minimize the burden on clinical services. Additionally, at the time of initial outreach, all patients in arms 2 and 3 receive a text message from the PCP or practice indicating that their primary care provider has already ordered a colonoscopy, with scheduling information. Text messaging via the Way to Health platform is bidirectional, offering automated replies to common colonoscopy related questions. If text messaging is not available, the patient receives similar messaging using Automated Voice Recording (AVR). Patients who are not up-to-date on screened based on EHR data (Epic Health Maintenance), do not schedule their colonoscopy within 2 months from initial outreach, and have not been ordered for multitarget stool DNA (MT-sDNA) receive an additional mailed and text message reminder. Those in arm 2 (colonoscopy only) receive a mailed reminder about colonoscopy scheduling. In arm 3 (sequential choice), patients receive a mailed FIT kit with instructions, a pre-paid return envelope and a letter describing the stool test as an annual alternative to colonoscopy. The FIT order is signed in bulk by the regional physician lead, with results going back to the PCP as the authorizing provider. Patients in arms 2 and 3 who have still not completed FIT, or either scheduled or completed colonoscopy at 4 months, receive a final text message reflecting their study arm assignment recommendation. This centralized, direct patient outreach is repeated annually for patients who are not up-to-date on CRC screening, in the same 4 batches of the prior study year. Patients randomized to a visit-based nudge arm (2B or 3B) who attend a visit with their PCP additionally receive a visit-based, clinician-directed nudge and a follow-up text 3 days post-visit. The nudges include an EHR static highlighted notification (Epic storyboard banner)( Fig. 2 ) notifying clinical staff (medical assistants, PCPs, and nursing staff) that a colonoscopy and/or FIT order has been placed on their behalf and encouraging discussion around completion. Patients receive a follow-up text message 3 days after the visit reminding them to follow through on active CRC screening orders (colonoscopy, FIT, or both). Fig. 2. Open in a new tab Clinician-facing nudge, storyboard banner (epic). Participants randomized to arm 1 (usual care) do not receive any direct outreach or visit-based nudges from this trial. Fig. 1 summarizes the trial design. As per usual care across the health system, endoscopy reports and follow-up recommendations are provided to patients by the performing gastroenterologist, and patients completing FIT are also be notified of their results. If a FIT positive patient does not schedule diagnostic colonoscopy follow-up within two months of their result and has not received standard follow-up outreach from the practice, the research team sends a note in the EHR to population health coordinators who will provide additional support. We will obtain baseline data from the EHR on demographics, prior CRC screening, co-morbidities, and mortality risk from automated data extraction and free text searches, as needed. We will use chart review to track outcomes of colonoscopies that are performed with polypectomy, so that we can categorize the finding. Study data are being collected and managed using REDCap electronic data capture tools hosted at the University of Pennsylvania. [ 37 , 38 ] Changes in Health Maintenance (HM) status are assessed via daily Clarity query to identify documentation of screening results and determine continued outreach. All HM status changes and results (internal or external) picked up by this query are chart-reviewed and only counted toward an outcome if documentation is present in the patient chart. 2.5. Mixed methods evaluation Annually, upon completion of all direct outreach in Batch 3, we send a validated survey guided by the Preventive Health Model (PHM), a selfregulation model shown to predict colorectal screening behaviors in a diverse population, to a select group of enrolled participants. [ 39 , 40 ] The original survey was developed and refined by Vernon and colleagues in 1997 and further validated in a larger and more diverse population by Tiro and colleagues in 2005. [ 40 , 41 ] For a target sample size of 100 completed surveys per year (300 total), and assuming a 20 % response rate, we are randomly selecting a minimum of 500 patients each year, stratified by race (Black patients vs all others) and study arms (intervention arms only) to complete the survey. The survey consists of 16 individual items spanning five psychosocial constructs: 1) salience and coherence, 2) social influence, 3) cancer worries, 4) perceived susceptibility, and 5) response efficacy. It will also include questions about CRC screening exposure and preferences. Patients randomly selected to complete the survey receive an invite via text message (Way to Health), and participants receive up to 2 automated reminders at 4 and 9 days. If our sample is lower than expected, remaining participants are purposively selected in batches of 10 to receive a manually generated text message offering a scheduled date and time to complete the survey by phone until an appropriate sample is obtained. Survey administration is repeated in Years 2 and 3, with no repeated participants in subsequent years. We monitor responses by study arm, race, and sex to ensure a representative sample. In year 3 of the trial, a purposive sample of 60 patients will be contacted by phone to conduct qualitative interviews to understand their experience with the specific outreach strategy and the reasons for the effectiveness (or lack of effectiveness) of each intervention. Guided by domains with the PHM, we will also ask semi-structured questions about their perception of the impact and design of outreach. We will also conduct qualitative interviews with 15 clinicians in the visit-based nudge arms to better understand their experiences with the interventions. Interview participants (patients and clinicians) will be consented and given a $25 gift card in appreciation of their time. 3. Analysis 3.1. Outcomes The primary trial outcome is patient completion of CRC screening at 3 years, which can be satisfied by any one of the following: colonoscopy completion at any time, negative FIT completed 2 times, or positive FIT followed by diagnostic colonoscopy within 1 year. The secondary outcomes include CRC screening completion at 3 years by any modality (to include anything satisfying the primary outcome, or completion of sigmoidoscopy, CT colonography, or MT-sDNA); the proportion of time that the patient is adherent to CRC screening process completion (during the 3 years of the study from the date of randomization); and the detection of adenomas, sessile serrated polyps, and advanced neoplasia (adenomas greater than 1 cm, or with high-grade dysplasia or villous histology, or invasive CRC). We will also evaluate the choice of screening test (colonoscopy, FIT, or other modalities) and potential screening harms such as colonoscopy perforation or post-polypectomy bleeding. 3.2. Sample size and power For the patient-directed analysis (Aim 1), after completion of the 3-year time frame for all participants, we will compare CRC screening completion for participants in each of the outreach arms, colonoscopy outreach (arm 2) and sequential outreach (arm 3), separately to the no outreach arm (arm 1); and we will also compare completion between colonoscopy (arm 2) and sequential outreach (arm 3) using intention-to-treat protocol. This includes all patients randomized, such that no evidence of completed CRC screening will be recorded as incomplete and not “missing” regardless of follow-up in the health system. Mortality will be analyzed as a screen failure but is expected to be rare and balanced by study arm. We estimate base CRC screening completion rate of 27 % in arm 1 (no outreach), which is based on estimates of response from a prior study. [ 29 ] With 20,000 patients randomly allocated in a 1:2:2 ratio, we will have 85 % power to detect a 3-percentage point increase in response rate of arms 2 and 3 relative to arm 1, and comparing arm 3 to arm 2, accounting for a total of three pairwise comparisons with a two-sided Type 1 error rate of 0.017 (Bonferroni correction of 0.05/3 = 0.017); we will also have 96 % power to detect a 3-percentage point increase in response rate for arm 3 compared to arm 2 (assuming 30 % completion in arm 2 and 33 % completion in arm 3). For the visit-based analysis (Aim 2), we will compare the primary outcome of CRC screening completion among patients in the direct outreach arms (who have at least 1 visit with the PCP) between the usual care arm (arms 2 A + 3 A) and the visit-based nudge arm (Arms 2B + 3B) that receives clinician directed nudges during the visit and follow-up text messages. Patients in the direct outreach arms (arms 2 and 3) will be randomized in a 1:1 ratio to either receive (A) usual care/no visit-based nudge, or (B) visit-based nudge plus follow-up text. Based on retrospective analysis in this population, we estimate approximately 80 % of patients in each arm will have at least 1 visit during the course of the trial, which translates to 12,800 patients in arms 2 and 3 that will be included in the visit-based analysis (about 6400 in each arm). As this group receives direct outreach, we anticipate a 31.5 % (the average of the estimated completion rate among all patients between Arm 2 and Arm 3) response across the intervention arms for usual care (no visit-based nudges). We will have 80 % power to detect a 2.3 percentage point increase in response rates of arm B relative to arm A. For our pre-specified subgroup analysis, we estimate 25 % of the study population (5000 patients) will be Black with a baseline completion rate of 23 %, so we will have 80 % power to detect a 5.5 percentage point increase in arms 2 and 3 compared to arm 1, and 80 % power to detect a 4.7 percentage point increase in arm 3 compared to arm 2 (assuming 28.5 % completion in arm 2). For the visit-based nudge comparison and a baseline completion rate of 30.9 % (the average of the estimated completion rate among Black patients between Arm 2 and Arm 3), we estimate that least 80 % of black patients will have at least one visit during the trial (3200 patients), and that we will have 80 % power to detect a 4.7 percentage point increase in arm B compared to arm A (estimate 4000 in arms 2 and 3 in the visit-based analysis). Power analyses for the difference in proportions were completed in Stata/SE 16.0. While the effects of the individual interventions are seemingly modest, the additional power from the large sample size will help to determine effectiveness of different components in a multi-level intervention, as well as effect on a subgroup of Black patients who have disparities in screening outcomes. Additionally, given the relatively low cost of these interventions, even a small improvement in response could be meaningful when scaled to a large population. 3.3. Analytic plan Analysis will be conducted after completion of the 3-year time frame for all participants according to the pre-specified plan and using intention-to-treat protocol. Mortality will be analyzed as a screen failure but is expected to be rare and balanced by arm. For Aim 1, we will fit a generalized linear model (GLM) for the binomial family with a logit link, robust standard errors, and fixed effects for treatment arm and clinic to compare colonoscopy outreach to no outreach, sequential outreach to no outreach, and sequential outreach to colonoscopy outreach, separately for the primary outcome. [ 42 ] We will perform a similar analysis for the detection of adenomas, sessile serrated polyps or advanced adenomas and a zero-inflated Poisson regression model with robust standard errors for the proportion of time covered by screening tests (number of days screened). [ 43 ] We will also include socioeconomic and clinical variables in a secondary multivariable model, including age, sex, race, insurance type, income (zip code level), and prior CRC screening, along with interaction terms with the clinician directed EHR nudge assignment. This will allow us to evaluate the effect of patient outreach in the context of visit-based nudges. We will perform a subgroup analysis of Black patients for each of the primary comparisons and conduct additional subgroup analyses for non-rare subgroups in an exploratory fashion, including the same socioeconomic and clinical variables described above for all. For Aim 2 (visit-based nudge comparison), we will again fit a generalized linear model (GLM) for the binomial family with a logit link, robust standard errors, fixed effect for clinic, and study arm at stage 1 (no outreach, colonoscopy outreach, sequential choice outreach) and stage 2 (nudge, no nudge) to compare visit-based nudges to usual care for the primary outcome. Individuals who attended at least 1 PCP visit will be included in this analysis. As a sensitivity analysis we may consider the potential interaction between the terms for the arm at stage 1 and at stage 2. This will allow us to evaluate the effect of visit-based nudges in the context of colonoscopy and sequential choice outreach. We will perform a similar analysis for the detection of adenomas, sessile serrated polyps or advanced adenomas and a zero-inflated Poisson regression model with robust standard errors for the proportion of time screened (days). We will also include socioeconomic and clinical variables in an exploratory multivariable model, including age, sex, race, insurance type, income (zip code level), and prior CRC screening, along with interaction terms with the patient outreach choice assignment. We will also conduct a sensitivity analysis to examine whether those that showed up for a visit in each arm are similar to those that were originally randomized to that nudge arm, and conduct a reweighted analysis in order to be representative of those originally randomized. We will perform a subgroup analysis of Black patients for each of the primary comparisons and conduct additional subgroup analyses for non-rare subgroups in an exploratory fashion, including the same socioeconomic and clinical variables described above for all. To evaluate the performance of the 3-year mortality index algorithm, we will perform a 3-year descriptive retrospective evaluation of CRC ordering rates, CRC screening completion rates, number of emergency department visits, number of hospitalizations, and mortality among all patients excluded from participation by the mortality index algorithm. 3.4. Mixed methods analysis The annual embedded survey consists of 16 individual items spanning five psychosocial constructs: 1) salience and coherence, 2) social influence, 3) cancer worries, 4) perceived susceptibility, and 5) response efficacy. [ 39 , 40 ] Each construct can be summed to create a score within each construct, and in total related to beliefs about CRC screening. We will calculate mean differences within each of the five constructs, evaluate each construct on its own scale, and overall look for any associations between CRC screening completion at 3 years and screening beliefs. Finally, we’ll compare screening exposure and preference responses to the type of screening completed at 3 years, if any. We will compare outcomes for those offered the survey and those who were not, to see if receiving the survey was associated with cancer screening completion. For analysis of the qualitative interview transcripts, two members of the research team will independently review the transcripts for inductive themes through an iterative process using the constant comparative approach. [ 44 – 46 ] After the study team reviews five sets of transcripts, we will develop a coding scheme that will be applied to subsequent data and modified as the study progresses and new themes emerge. The study team members will assess and reassess our inter-rater reliability. Finally, we will perform a cost analysis using resource utilization to determine the incremental cost of outreach in each arm and the incremental cost of additional patients screened to evaluate resource requirements for scale. We will evaluate the cost from the perspective of the health system, including the time spent by staff that would be needed as part of routine operations when implemented, while not including effort for activities that are only part of the research process. 4. Discussion Durably increasing CRC screening rates equitably among primary care patients requires interventions at multiple levels. By addressing gaps that have been identified from prior research, this trial leverages existing screening infrastructure and layers it with an organized approach, novel use of existing EHR functionality, informed by key concepts from the field of behavioral science. This will be the first large, prospective trial evaluating the long-term effectiveness of offering sequential choice of colonoscopy then FIT compared to colonoscopy through a multi-level, centralized outreach and visit-based nudge design. Importantly, there is a pre-specified subgroup analysis of Black patients, who are known to have disparities in access and outcomes for CRC. Additionally, if validated, the mortality algorithm could be helpful for the identification of an eligible population without the need for extensive chart review. Further, we have designed and embedded this trial in partnership with health system leadership in a pragmatic design, and have integrated many aspects of the intervention in the EHR and with routine care to make it more efficient and adaptable to other clinical settings. There are some limitations. While the health system includes a diverse range of primary care practices across the Philadelphia region, the learnings may not translate to other regions or types of health systems. We are not powered to detect differences in clinical outcomes such as cancer incidence and mortality, and we may not be able to fully capture care that happens outside of our health system, despite the integrated EHR and access to some data from other health systems. In conclusion, this study will provide additional evidence on operationalizing population-level screening programs to reduce CRC burden, and inform future interventions to increase CRC screening. Supplementary Material 1 NIHMS2163388-supplement-1.docx (20.1KB, docx) Role of the funder/sponsor The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Funding This work was supported by the National Institutes of Health (R01 CA258893). Declaration of competing interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Shivan Mehta reports financial support was provided by National Institutes of Health. Shivan Mehta reports a relationship with Guardant Health Inc that includes: consulting or advisory. Katharine Rendle reports a relationship with Merck & Co Inc that includes: consulting or advisory. Katharine Rendle reports a relationship with Pfizer that includes: funding grants. Katharine Rendle reports a relationship with AstraZeneca PLC that includes: funding grants. Katharine Rendle reports a relationship with MJH Life Sciences LLC that includes: speaking and lecture fees. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.cct.2025.108188 . Footnotes CRediT authorship contribution statement Shivan J. Mehta: Writing – original draft, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Pamela A. Shaw: Writing – review & editing, Methodology, Investigation, Formal analysis, Conceptualization. Catherine Reitz: Writing – original draft, Supervision, Project administration, Data curation, Conceptualization. Caitlin Brophy: Writing – review & editing, Project administration, Data curation. Evelyn Okorie: Writing – review & editing, Project administration, Data curation. Keyirah Williams: Writing – review & editing, Project administration, Data curation. Abraham Segura: Writing – review & editing. Jinming Tao: Writing – review & editing, Formal analysis, Data curation. Christopher K. Snider: Writing – review & editing, Methodology, Formal analysis, Data curation. Colin Wollack: Writing – review & editing, Software. Sadie Friday: Writing – review & editing, Software. Katharine A. Rendle: Writing – review & editing, Methodology, Investigation, Conceptualization. Tamar Klaiman: Writing – review & editing, Methodology, Formal analysis. Karen Glanz: Writing – review & editing, Methodology, Investigation. Corinne Rhodes: Writing – review & editing, Supervision, Investigation, Conceptualization. David A. Asch: Writing – review & editing, Investigation, Conceptualization. Data availability No data was used for the research described in the article. References [1]. Doubeni CA, Corley DA, Quinn VP, Jensen CD, Zauber AG, Goodman M, Johnson JR, Mehta SJ, Becerra TA, Zhao WK, Schottinger J, Doria-Rose VP, Levin TR, Weiss NS, Fletcher RH, Effectiveness of screening colonoscopy in reducing the risk of death from right and left colon cancer: a large community-based study, Gut 67 (2) (2018) 291–298. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [2]. Levin TR, Corley DA, Jensen CD, Schottinger JE, Quinn VP, Zauber AG, Lee JK, Zhao WK, Udaltsova N, Ghai NR, Lee AT, Quesenberry CP, Fireman BH, Doubeni CA, Effects of organized colorectal cancer screening on cancer incidence and mortality in a large community-based population, Gastroenterology 155 (5) (2018) 1383–1391.e5. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [3]. Meester RG, Doubeni CA, Zauber AG, Goede SL, Levin TR, Corley DA, Jemal A, Lansdorp-Vogelaar I, Public health impact of achieving 80% colorectal cancer screening rates in the United States by 2018, Cancer 121 (13) (2015) 2281–2285. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [4]. U.P.S.T. Force, Screening for colorectal cancer: US preventive services task force recommendation statement, Jama 325 (19) (2021) 1965–1977. [ DOI ] [ PubMed ] [ Google Scholar ] [5]. Sabatino SA, Thompson TD, Croswell JM, Villarroel MA, Rodriguez JL, Adam EE, Richardson LC, Use of cancer screening tests, United States, 2023, Prev. Chronic Dis 22 (2025) E52. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [6]. Carethers JM, Racial and ethnic disparities in colorectal cancer incidence and mortality, Adv. Cancer Res 151 (2021) 197–229. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [7]. Siegel RL, Giaquinto AN, Jemal A, Cancer statistics, 2024, CA Cancer J. Clin 74 (1) (2024) 12–49. [ DOI ] [ PubMed ] [ Google Scholar ] [8]. Dougherty MK, Brenner AT, Crockett SD, Gupta S, Wheeler SB, Coker-Schwimmer M, Cubillos L, Malo T, Reuland DS, Evaluation of interventions intended to increase colorectal cancer screening rates in the United States: a systematic review and meta-analysis, JAMA Intern. Med 178 (12) (2018) 1645–1658. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [9]. Kahneman D, Tversky A, Prospect theory: an analysis of decision under risk, Econometrica 47 (2) (1979) 263–291. [ Google Scholar ] [10]. Tversky A, Kahneman D, The framing of decisions and the psychology of choice, Science 211 (4481) (1981) 453–458. [ DOI ] [ PubMed ] [ Google Scholar ] [11]. Mehta SJ, Asch DA, How to help gastroenterology patients help themselves: leveraging insights from behavioral economics, Clin. Gastroenterol. Hepatol 12 (5) (2014) 711–714. [ DOI ] [ PubMed ] [ Google Scholar ] [12]. Wang Y, Sloan FA, Present bias and health, J. Risk Uncertain 57 (2) (2018) 177–198. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [13]. Samuelson W, Zeckhauser R, Status quo bias in decision making, J. Risk Uncertain 1 (1) (1988) 7–59. [ Google Scholar ] [14]. Chernev A, Böckenholt U, Goodman J, Choice overload: a conceptual review and meta-analysis, J. Consum. Psychol 25 (2) (2015) 333–358. [ Google Scholar ] [15]. Halpern SD, Ubel PA, Asch DA, Harnessing the power of default options to improve health care, N. Engl. J. Med 357 (13) (2007) 1340–1344. [ DOI ] [ PubMed ] [ Google Scholar ] [16]. Loewenstein G, Brennan T, Volpp KG, Asymmetric paternalism to improve health behaviors, Jama 298 (20) (2007) 2415–2417. [ DOI ] [ PubMed ] [ Google Scholar ] [17]. Richter KP, Catley D, Gajewski BJ, Faseru B, Shireman TI, Zhang C, Scheuermann TS, Mussulman LM, Nazir N, Hutcheson T, Shergina E, Ellerbeck EF, The effects of opt-out vs opt-in tobacco treatment on engagement, cessation, and costs: a randomized clinical trial, JAMA, Intern. Med 183 (4) (2023) 331–339. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [18]. Montoy JCC, Dow WH, Kaplan BC, Patient choice in opt-in, active choice, and opt-out HIV screening: randomized clinical trial, BMJ 352 (2016) h6895. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [19]. Mehta SJ, Day SC, Norris AH, Sung J, Reitz C, Wollack C, Snider CK, Shaw PA, Asch DA, Behavioral interventions to improve population health outreach for hepatitis C screening: randomized clinical trial, Bmj 373 (2021) n1022. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [20]. Mehta SJ, Khan T, Guerra C, Reitz C, McAuliffe T, Volpp KG, Asch DA, Doubeni CA, A randomized controlled trial of opt-in versus opt-out colorectal cancer screening outreach, Am. J. Gastroenterol 113 (12) (2018) 1848–1854. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [21]. Mehta SJ, Rhodes C, Linn KA, Reitz C, McDonald C, Okorie E, Williams K, Resnick D, Arostegui A, McAuliffe T, Wollack C, Snider CK, Peifer MK, Weinstein SP, Behavioral interventions to improve breast cancer screening outreach: two randomized clinical trials, JAMA Intern. Med 184 (7) (2024) 761–768. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [22]. Rex DK, Boland CR, Dominitz JA, Giardiello FM, Johnson DA, Kaltenbach T, Levin TR, Lieberman D, Robertson DJ, Colorectal cancer screening: recommendations for physicians and patients from the U.S. multi-society task force on colorectal cancer, Am. J. Gastroenterol 112 (7) (2017) 1016–1030. [ DOI ] [ PubMed ] [ Google Scholar ] [23]. Singal AG, Gupta S, Skinner CS, Ahn C, Santini NO, Agrawal D, Mayorga CA, Murphy C, Tiro JA, McCallister K, Sanders JM, Bishop WP, Loewen AC, Halm EA, Effect of colonoscopy outreach vs fecal immunochemical test outreach on colorectal cancer screening completion: a randomized clinical trial, Jama 318 (9) (2017) 806–815. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [24]. Liang PS, Wheat CL, Abhat A, Brenner AT, Fagerlin A, Hayward RA, Thomas JP, Vijan S, Inadomi JM, Adherence to competing strategies for colorectal cancer screening over 3 years, Am. J. Gastroenterol 111 (1) (2016) 105–114. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [25]. Inadomi JM, Vijan S, Janz NK, Fagerlin A, Thomas JP, Lin YV, Munoz R, Lau C, Somsouk M, El-Nachef N, Hayward RA, Adherence to colorectal cancer screening: a randomized clinical trial of competing strategies, Arch. Intern. Med 172 (7) (2012) 575–582. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [26]. Mehta SJ, Palat S, McDonald C, Reitz C, Okorie E, Williams K, Tao J, Shaw PA, Glanz K, Asch DA, A randomized trial of choice architecture and mailed colorectal cancer screening outreach in a community health setting, Clin. Gastroenterol. Hepatol 22 (10) (2024) 2117–2124.e2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [27]. Besedeš T, Deck C, Sarangi S, Shor M, Reducing choice overload without reducing choices, Rev. Econ. Stat 97 (4) (2015) 793–802. [ Google Scholar ] [28]. Pilonis ND, Bugajski M, Wieszczy P, Rupinski M, Pisera M, Pawlak E, Regula J, Kaminski MF, Participation in competing strategies for colorectal cancer screening: a randomized health services study (PICCOLINO study), Gastroenterology 160 (4) (2021) 1097–1105. [ DOI ] [ PubMed ] [ Google Scholar ] [29]. Mehta SJ, Induru V, Santos D, Reitz C, McAuliffe T, Orellana C, Volpp KG, Asch DA, Doubeni CA, Effect of sequential or active choice for colorectal cancer screening outreach: a randomized clinical trial, JAMA Netw. Open 2 (8) (2019) e1910305. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [30]. Asch DA, Ziolek TA, Mehta SJ, Misdirections in informed consent - impediments to health care innovation, N. Engl. J. Med 377 (15) (2017) 1412–1414. [ DOI ] [ PubMed ] [ Google Scholar ] [31]. Epic, Epic Systems Corporation, Verona, WI. 2025. [ Google Scholar ] [32]. U.P.S.T. Force, Screening for colorectal cancer: US preventive services task force recommendation statement, Jama 315 (23) (2016) 2564–2575. [ DOI ] [ PubMed ] [ Google Scholar ] [33]. Keller SC, Momplaisir F, Lo Re V, Newcomb C, Liu Q, Ratcliffe SJ, Long JA, Colorectal cancer incidence and screening in US Medicaid patients with and without HIV infection, AIDS Care 26 (6) (2014) 716–722. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [34]. Hitchcock ME, Green BB, Anderson DS, Advancing health equity for medicaid beneficiaries by adding colorectal cancer screening to the centers for medicare and medicaid services adult core set, Gastroenterology 162 (3) (2022) 710–714. [ DOI ] [ PubMed ] [ Google Scholar ] [35]. Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, Blondel M, Prettenhofer P, Weiss R, Dubourg V, Vanderplas J, Passos A, Cournapeau D, Brucher M, Perrot M, Duchesnay E, Scikit-learn: machine learning in Python, J. Mach. Learn. Res 12 (2011) 2825–2830. [ Google Scholar ] [36]. Asch DA, Volpp KG, On the way to health, LDI Issue Brief 17 (9) (2012) 1–4. [ PubMed ] [ Google Scholar ] [37]. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, McLeod L, Delacqua G, Delacqua F, Kirby J, Duda SN, The REDCap consortium: building an international community of software platform partners, J. Biomed. Inform 95 (2019) 103208. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [38]. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG, Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support, J. Biomed. Inform 42 (2) (2009) 377–381. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [39]. Wainwright JV, Mehta SJ, Clifton A, Bocage C, Ogden SN, Cohen S, Rendle KA, Persistent barriers to colorectal cancer screening completion amid centralized outreach: a mixed methods study, Am. J. Health Promot 36 (4) (2022) 697–705. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [40]. Tiro JA, Vernon SW, Hyslop T, Myers RE, Factorial validity and invariance of a survey measuring psychosocial correlates of colorectal cancer screening among African Americans and Caucasians, Cancer Epidemiol. Biomarkers Prev 14 (12) (2005) 2855–2861. [ DOI ] [ PubMed ] [ Google Scholar ] [41]. Vernon SW, Myers RE, Tilley BC, Development and validation of an instrument to measure factors related to colorectal cancer screening adherence, Cancer Epidemiol. Biomarkers Prev 6 (10) (1997) 825–832. [ PubMed ] [ Google Scholar ] [42]. Nelder JA, Wedderburn RWM, Generalized linear models, J. R. Stat. Soc. A 135 (3) (1972) 370–384. [ Google Scholar ] [43]. Murphy CC, Sigel BM, Yang E, Skinner CS, Halm EA, Gupta S, Sanders JM, McCallister K, Singal AG, Adherence to colorectal cancer screening measured as the proportion of time covered, Gastrointest. Endosc 88 (2) (2018) 323–331.e2. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [44]. Pope C, Ziebland S, Mays N, Qualitative research in health care: analysing qualitative data, BMJ, Br. Med. J 320 (7227) (2000) 114. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] [45]. Corbin JM, Strauss A, Grounded theory research: procedures, canons, and evaluative criteria, Qual. Sociol 13 (1) (1990) 3–21. [ Google Scholar ] [46]. Malterud K, Qualitative research: standards, challenges, and guidelines, Lancet 358 (9280) (2001) 483–488. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials 1 NIHMS2163388-supplement-1.docx (20.1KB, docx) Data Availability Statement No data was used for the research described in the article. ACTIONS View on publisher site PDF (473.0 KB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top