Colorectal Cancer
Conditions
Keywords
cancer screening, decision making
Brief summary
Systematic efforts to improve colorectal cancer screening use in the VA Healthcare System have resulted in an increase in not only appropriate, but also inappropriate use of screening. The purpose of this study is to test a new, more patient-centered approach to colon cancer screening. In older individuals (ages 70 to 75) who are due for screening, the investigators will provide education on the benefits and harms of colon cancer screening. But instead of simply providing these patients with average information about these benefits, the investigators will give them information that takes into account their personal characteristics (e.g., age, gender), overall health, and screening history and therefore applies to them more personally. The investigators will also work with the health system to create time and space for patient and doctor to discuss whether screening is the right choice for each individual. This way, patients can make an informed choice about what is right for them, with the help of their doctor. In the future, the results of this study will help patients make more informed screening decisions, especially when the benefits of screening may be uncertain for them personally.
Detailed description
Colorectal cancer (CRC) screening is a widely recommended, evidence-based preventive service that has traditionally been underused. Over the last decade, organized efforts by the Veterans Health Administration (VHA) to increase population screening among Veterans have been successful. But these population-centered efforts have increased screening utilization in a way that is not always concordant with screening benefit, particularly among older Veterans, those with comorbid illness, and those who have previously been screened. As patients get older, acquire health problems, and undergo negative screening tests, the benefit of screening decreases and the potential harm of screening increases. Yet, existing population-centered efforts fail to adequately inform these patients about this changing balance in benefit and harm, often yielding screening utilization that is discordant with benefit. The purpose of this study is to test a more Veteran-centered approach to screening in these individuals, one that encourages informed, personalized screening decisions based on individual values, preferences, and health status. The 3-part intervention consists of: (1) a decision aid to help Veterans make informed screening decisions; (2) education for providers on how the benefits of screening vary between patients; and (3) modification of clinical reminder systems to allow Veterans to make informed decisions about screening. The intervention will be tested in a pragmatic cluster-randomized controlled trial (cluster = provider) at two sites in the VA Ann Arbor Healthcare System. The primary outcome will be whether screening was ordered at the clinic visit. The investigators will also assess the appropriateness of screening orders (i.e., whether screening is ordered in concordance with screening benefit), conceptual understanding of screening, elements of informed decision-making addressed in the screening discussion, and screening utilization at 6 months. Note: In March 2023, during preparation of the final manuscript for submission for publication, the study team noted that one subject in the intervention arm had undergone colorectal cancer screening immediately prior to the study visit (but after assessment for study eligibility), making the subject ineligible (protocol violation). Study results were re-analyzed accordingly and updated on clinicaltrials.gov (analyzed N=431 rather than N=432). Additionally, a data entry error was noted on clinicaltrials.gov for the secondary outcome of screening utilization (control N=96 rather than control N=95).
Interventions
BEHAVIORALDecision Aid
Printed booklet comprising educational information about benefits and harms of screening, individualized estimates of benefits and harms of screening, and values clarification exercise
BEHAVIORALProvider Education
Providers of patients in both arms of the study will be given an educational module about recent data on the benefits and harms of screening how these data fit in with existing population-centered guidelines.
BEHAVIORALPerformance Measure Modification
The clinical reminder system will be modified so to facilitate documentation of informed decision making about CRC screening, including specific reasons for not screening. Additionally, providers who indicate a specific exception for not screening (using the modified clinical reminder) will be considered as satisfying the requirements and will not be penalized in terms of performance pay, and will be removed from feedback reports that encourage population screening.
BEHAVIORALSimple Informational Booklet
A simple informational booklet explaining colorectal cancer screening and current screening recommendations.
Sponsors
University of Colorado, Denver
Memorial Sloan Kettering Cancer Center
Erasmus Medical Center
VA Office of Research and Development
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
HEALTH_SERVICES_RESEARCH
Masking
NONE
Intervention model description
A cluster-randomized (by provider) pragmatic trial of the patient-centered intervention versus pragmatic control.
Eligibility
Sex/Gender
ALL
Age
70 Years to 75 Years
Healthy volunteers
No
Inclusion criteria
* Aged 70-75 years * Due for screening according to the 2008 USPSTF colorectal cancer screening guideline * Scheduled for a non-urgent primary care visit at the Ann Arbor VA Medical Center or Toledo VA Community-Based Outpatient Clinic
Exclusion criteria
* Increased risk for colorectal cancer (and therefore not candidates for average-risk screening) * Limited life expectancy (e.g., enrolled in hospice or diagnosed with metastatic cancer), or for whom the provider previously documented an intention not to pursue screening. * Scheduled for an appointment where stimulating a discussion about screening is likely to be inappropriate: urgent appointment (for acute complaints), follow-up visit after hospitalization * Have a condition that would impair his/her ability to participate in the study: dementia or other cognitive impairment, visual impairment, non-English speaking * Assigned to an ineligible primary care provider (i.e., the provider did not consent to the study) * Have Medical Guardian who makes decisions for the patient about his/her care
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Number of Participants With CRC Screening Ordered | 2 weeks | The primary dependent variable in the analysis was whether screening was ordered within two weeks after the clinic visit (dichotomous). Screening orders were determined by manual record review of electronic health records. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Concordance Between Screening Orders and Screening Benefit | 2 weeks | Defined as the degree to which screening orders align with expected screening benefit, such that individuals with low screening benefit receive screening orders at a lower rate than those with high screening benefit. We hypothesized that Veterans randomized to the intervention (decision aid) would receive screening orders that were more concordant with screening benefit than those randomized to the control. The expected benefit of screening (reduction in CRC incidence) was calculated using the MISCAN-Colon model. For a given patient, this value was a function of age, gender, health status, and prior screening history. The regression analysis included screening orders as the dependent variable, and, study arm, expected benefit, and an interaction term between study arm and expected benefit as the independent variables. The p-value reported is for the interaction term. |
| Number of Participants With CRC Screening Utilized | 6 months | Screening test completion was collected through manual review of electronic medical records. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Decision Aid Patients of primary care providers randomly assigned to the intervention which included 1) a decision aid; 2) modified performance measure/reminder; and, 3) provider education.
1. A printed booklet comprising educational information about benefits and harms of screening, individualized estimates of benefits and harms of screening, and a values clarification exercise.
2. The clinical reminder system was modified to facilitate documentation of informed decision making about CRC screening, including specific reasons for not screening. Providers who indicated an exception for not screening a patient were considered as satisfying the requirements and were not penalized in terms of performance pay. Additionally, the patient was removed from provider feedback reports that encourage population screening.
3. Providers were given information about recent data on the benefits and harms of screening & how these data fit in with existing population-centered guidelines. | 258 |
| No Decision Aid Patients of primary care providers randomly assigned to the pragmatic control which included 1) a simple booklet in place of the decision aid; 2) modified performance measure/reminder; and, 3) provider education.
1. A simple informational booklet explaining colorectal cancer (CRC) screening and current screening recommendations.
2. The clinical reminder system was modified to facilitate documentation of informed decision making about CRC screening, including specific reasons for not screening. Providers who indicated a specific exception for not screening a patient were considered as satisfying the requirements and were not penalized in terms of performance pay. Additionally, the patient was removed from provider feedback reports that encourage population screening.
3. Providers were given information about recent data on the benefits and harms of screening & how these data fit in with existing population-centered guidelines. | 173 |
| Total | 431 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Failed to attend study visit | 2 | 0 |
| Overall Study | Failed to sign HIPAA | 0 | 2 |
| Overall Study | Protocol Violation | 1 | 0 |
Baseline characteristics
| Characteristic | Decision Aid | Total | No Decision Aid |
|---|---|---|---|
| Age, Continuous | 71.5 years STANDARD_DEVIATION 1.6 | 71.5 years STANDARD_DEVIATION 1.7 | 71.7 years STANDARD_DEVIATION 1.7 |
| Education 1-3 years of college | 80 Participants | 127 Participants | 47 Participants |
| Education Did not complete high school | 13 Participants | 21 Participants | 8 Participants |
| Education Graduated college or 4 or more years of college | 32 Participants | 55 Participants | 23 Participants |
| Education Graduate or professional school | 16 Participants | 34 Participants | 18 Participants |
| Education High school/GED | 80 Participants | 124 Participants | 44 Participants |
| Education Missing | 9 Participants | 17 Participants | 8 Participants |
| Education Vocational, technical, or business training | 28 Participants | 53 Participants | 25 Participants |
| Marital Status Married | 162 Participants | 270 Participants | 108 Participants |
| Marital Status Missing | 8 Participants | 17 Participants | 9 Participants |
| Marital Status Never Married | 17 Participants | 21 Participants | 4 Participants |
| Marital Status Separated/Divorced | 54 Participants | 91 Participants | 37 Participants |
| Marital Status Widowed | 17 Participants | 32 Participants | 15 Participants |
| Mean Benefit from Screening | 11.3 Cancers prevented per 1,000 pts screened STANDARD_DEVIATION 8.1 | 11.3 Cancers prevented per 1,000 pts screened STANDARD_DEVIATION 8.1 | 11.3 Cancers prevented per 1,000 pts screened STANDARD_DEVIATION 8 |
| Prior Screening Colonoscopy | 43 Participants | 75 Participants | 32 Participants |
| Prior Screening FOBT/FIT | 165 Participants | 276 Participants | 111 Participants |
| Prior Screening None | 50 Participants | 80 Participants | 30 Participants |
| Race/Ethnicity, Customized Black or African American | 18 Participants | 26 Participants | 8 Participants |
| Race/Ethnicity, Customized Missing | 11 Participants | 25 Participants | 14 Participants |
| Race/Ethnicity, Customized Native Hawaiian or other Pacific Islander | 0 Participants | 1 Participants | 1 Participants |
| Race/Ethnicity, Customized Native Indian or Alaskan Native | 3 Participants | 5 Participants | 2 Participants |
| Race/Ethnicity, Customized White | 226 Participants | 374 Participants | 148 Participants |
| Region of Enrollment United States | 258 Participants | 431 Participants | 173 Participants |
| Screening Benefit Participants with Screening Benefit 0-4.9 | 69 Participants | 107 Participants | 38 Participants |
| Screening Benefit Participants with Screening Benefit 10-14.9 | 57 Participants | 97 Participants | 40 Participants |
| Screening Benefit Participants with Screening Benefit 15-19.9 | 7 Participants | 9 Participants | 2 Participants |
| Screening Benefit Participants with Screening Benefit 20-24.9 | 8 Participants | 12 Participants | 4 Participants |
| Screening Benefit Participants with Screening Benefit 25-29.9 | 11 Participants | 20 Participants | 9 Participants |
| Screening Benefit Participants with Screening Benefit 30-35.0 | 20 Participants | 33 Participants | 13 Participants |
| Screening Benefit Participants with Screening Benefit 5-9.9 | 86 Participants | 153 Participants | 67 Participants |
| Sex: Female, Male Female | 5 Participants | 7 Participants | 2 Participants |
| Sex: Female, Male Male | 253 Participants | 424 Participants | 171 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 0 | 0 / 0 |
| other Total, other adverse events | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 |
Outcome results
Primary
Number of Participants With CRC Screening Ordered
The primary dependent variable in the analysis was whether screening was ordered within two weeks after the clinic visit (dichotomous). Screening orders were determined by manual record review of electronic health records.
Time frame: 2 weeks
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Decision Aid | Number of Participants With CRC Screening Ordered | 162 Participants |
| No Decision Aid | Number of Participants With CRC Screening Ordered | 114 Participants |
p-value: 0.491Regression, Logistic
Secondary
Concordance Between Screening Orders and Screening Benefit
Defined as the degree to which screening orders align with expected screening benefit, such that individuals with low screening benefit receive screening orders at a lower rate than those with high screening benefit. We hypothesized that Veterans randomized to the intervention (decision aid) would receive screening orders that were more concordant with screening benefit than those randomized to the control. The expected benefit of screening (reduction in CRC incidence) was calculated using the MISCAN-Colon model. For a given patient, this value was a function of age, gender, health status, and prior screening history. The regression analysis included screening orders as the dependent variable, and, study arm, expected benefit, and an interaction term between study arm and expected benefit as the independent variables. The p-value reported is for the interaction term.
Time frame: 2 weeks
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Decision Aid | Concordance Between Screening Orders and Screening Benefit | 162 Participants |
| No Decision Aid | Concordance Between Screening Orders and Screening Benefit | 114 Participants |
p-value: 0.049Regression, Logistic
Secondary
Number of Participants With CRC Screening Utilized
Screening test completion was collected through manual review of electronic medical records.
Time frame: 6 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Decision Aid | Number of Participants With CRC Screening Utilized | 106 Participants |
| No Decision Aid | Number of Participants With CRC Screening Utilized | 96 Participants |