Colorectal Cancer
Conditions
Keywords
Colorectal cancer screening, Shared decision-making, Patient preferences, Risk assessment, Advanced colorectal neoplasia
Brief summary
Shared decision-making (SDM) has been advocated as a strategy for increasing colorectal cancer (CRC) screening rates. Our studies to date suggest that while the use of a novel computer-based decision aid facilitates several components of SDM from both the patient and provider perspective, there is a reluctance among providers to acquiesce to patient preferences for a particular screening strategy when its differs from their own. The overall objective of this study is to assess whether risk stratification for advanced colorectal neoplasia influences clinical decision-making related to screening test selection and adherence within a SDM framework. Eligible subjects will be randomized to either an experimental arm, in which they will be asked to complete a 6-item risk assessment questionnaire known as the Advanced Colorectal Neoplasia Index \[ACNI\] after reviewing a web-based decision aid, or a control arm, in which they will only review the decision aid. Both interventions will take place just before a prearranged office visit with their provider. The primary outcome will be screening test ordered; secondary outcomes will include test completion rates, concordance between test preference and test ordered,, patient satisfaction with decision-making process, screening intentions, 6-month test completion rates and provider satisfaction. Outcomes will be evaluated using computerized tracking systems or validated instruments.
Detailed description
Colorectal cancer (CRC) is the second leading cause of cancer-related death in the United States. Screening by any of at least 6 different methods is a cost-effective yet underutilized strategy for reducing both CRC incidence and mortality. Because these methods differ with respect to risks and benefits and because existing evidence fails to identify a single best strategy, most authoritative groups advocate a shared decision-making (SDM) approach when selecting an appropriate screening strategy. SDM is a sequential, interactive process involving information exchange, values clarification, decision-making and mutual agreement. To facilitate this process, patient-oriented decision aids have been developed to enable patients to identify a preferred strategy based on personal values and empower them to participate in the decision-making process. Our recent studies to date find that although decision aids enable patients to make informed choices, providers are often unwilling to acquiesce to patient preferences when they differ from their own. Since accurate risk assessment is a critical component of effective clinical decision-making, the investigators postulate that risk stratification for the point prevalence of advanced colorectal neoplasia will enable providers to incorporate objective risk-based criteria in their decision-making when considering patient preferences for screening. To that end, the investigators have recently developed and validated the so-called Advanced Colorectal Neoplasia Index \[ACNI\] that stratifies patients into low versus intermediate/high risk categories based on available clinical data, including age, sex, race/ethnicity, smoking history, daily alcohol intake and use of non-steroidal anti-inflammatory drugs. The overall objective of this study is to determine whether risk stratification using the ACNI influences clinical decision-making related to screening test selection and adherence to screening within a SDM framework. Hypothesis: Providers who incorporate risk estimates of ACN in their decision-making when recommending screening tests are more likely to consider patient preferences for options other than colonoscopy than providers lacking this information.
Interventions
BEHAVIORALRisk Assessment
Patients randomized to the experimental arm will be asked a complete the ACNI risk assessment tool after reviewing a web-based colorectal cancer decision aid. The ACNI uses a point based system to stratify patients into low (mean rate of ACN \ 3%) versus intermediate/high (\ 8%) risk groups based on responses to 6 items: age (50-59, 60-69, 70+), sex (male/female), race/ethnicity (non-Hispanic black, other), smoking history (never, \<20 years, 20+ years), daily alcohol intake (\< 2 vs. \>/=2 drinks) and use of non-steroidal anti-inflammatory drugs (ever, never). The index represents a prototype version of the Advanced Colorectal Neoplasia Index (Am J Gastroenterol 2015;110:1062-71).
Sponsors
Boston Medical Center
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
HEALTH_SERVICES_RESEARCH
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
50 Years to 75 Years
Healthy volunteers
Yes
Inclusion criteria
* English-speaking average-risk patients 50 to 75 years of age; * Due for CRC screening based on current recommendations (i.e. no prior screening or \> 1year since last fecal occult blood testing \[FOBT\], \> 3 years since last stool DNA test, \> 5 years since last flexible sigmoidoscopy, virtual colonoscopy or double-contrast barium enema \[DCBE\], or \> 10 years since last colonoscopy); * Under the direct care of a staff (attending) primary care provider or physician extender; * Absence of major co-morbidities that preclude CRC screening.
Exclusion criteria
* High-risk condition (personal history of colorectal cancer or polyps, family history of colorectal cancer or polyps involving one or more first degree relatives \< 60 years of age, chronic inflammatory bowel disease); * Presence of alarm gastrointestinal symptoms, including rectal bleeding, recent change in bowel habits, abdominal pain, unexplained weight loss and iron deficiency anemia; * Comorbidities that preclude CRC screening by any method; * Lack of fluency in written and spoken English (since decision aid and personalized risk assessment tool will be in English only due to funding issues).
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Concordance Between Patient Preference and Test Ordered | 3 months | Concordance is a measure of the agreement between the patient's test preference and actual test ordered for standard care vs. risk assessment patients. It is defined as the number of patients who had their preferred test ordered. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Concordance Between Patient Preference and Test Ordered for High vs. Low Risk Patients | 3 months | Concordance between patient preference and test ordered for high versus low risk patients. It is defined as the number of patients who had their preferred test ordered. |
| Screening Test Completion | 6 months | Test completion rates were tracked using BMC's electronic medical record, which captures results for all endoscopic procedures, imaging studies, and stool blood tests. |
| Satisfaction With Decision-making Process (SDMP) | One month | SDMP was assessed on the posttest using the validated 12-item Satisfaction with the Decision-Making Process scale. Individual items are assigned a point value ranging from 1 for ''strongly disagree'' (or ''poor'') to 5 for ''strongly agree'' (or ''excellent''). A cumulative score is then calculated based on the summed response scores for each item (maximum score = 60). Data was missing for 11 patients in the concordant group and 6 patients in the discordant group |
| Screening Intentions | 3 months | Screening intentions were assessed on the posttest. Patients were asked how sure they were that they would complete the screening test that got scheduled Scores ranged from 5 = ''completely'' to 1 = ''not at all sure.'' Data was missing for 11 patients in the concordant group and 6 patients in the discordant group. |
| Provider Satisfaction | Two years | Provider satisfaction was assessed based on responses to a 3-item pretest administered prior to commencement of the study and the same 3-item posttest. The 3 items assessed to the extent to which providers felt that personalized risk assessment would be useful for: (1) selecting an appropriate screening test for their average risk patients \[test selection\]; (2) reduce time to decide on an appropriate screening modality \[save time\]; and (3) make them more receptive to patient preferences and possibly order a screening test other than colonoscopy \[receptive to patient preferences\]. Responses were assigned a point value ranging from 5= strongly agree and 1 = strongly disagree. |
Other
| Measure | Time frame | Description |
|---|---|---|
| Concordance Between Patient Preference for Colonoscopy and Test Ordered for High Versus Low Risk Patients | 3 months | Test-specific concordance between patient preference for colonoscopy and test ordered for high versus low risk patients. It is defined as the number of patients who had their preferred test ordered. |
| Concordance Between Patient Preferences for a Screening Tests Other Than Colonoscopy and Test Ordered for High Versus Low Risk Patients | 3 months | Test-specific concordance between patient preference for a screening test other than colonoscopy (fecal occult blood testing, flexible sigmoidoscopy, double-contrast barium enema, CT colonography and stool DNA) and test ordered for high versus low risk patients. It is defined as the number of patients who had their preferred test ordered. |
| Concordance Between Patient Preferences for Screening Tests Other Than Colonoscopy and Test Ordered | 3 months | Test-specific concordance between patient preference for a screening test other than colonoscopy (fecal occult blood testing, flexible sigmoidoscopy, double-contrast barium enema, CT colonography and stool DNA) and test ordered for standard care versus risk assessment arms. It is defined as the number of patients who had their preferred test ordered. |
| Concordance Between Patient Preference for Colonoscopy and Test Ordered | 3 months | Test-specific concordance between patient preference for colonoscopy and test ordered for standard care versus risk assessment groups. It is defined as the number of patients who had their preferred test ordered. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Standard Care Subjects randomized to the control arm reviewed the web-based decision aid (http://www.colorectalcancerscreening4u.com) just prior to a scheduled visit with their provider. | 168 |
| Risk Assessment Subjects randomized to the experimental arm completed the ACNI risk assessment tool after reviewing the web-based decision aid (http://www.colorectalcancerscreening4u.com) just prior to a scheduled office visit with their provider.
Risk Assessment: Patients randomized to the experimental arm will be asked a complete the ACNI risk assessment tool after reviewing a web-based colorectal cancer decision aid The ACNI uses a point based system to stratify patients into low (mean rate of ACN \
3%)versus intermediate/high (\
8%) risk groups based on responses to 6 items: age 50-59,60-69, 70+), sex (male/female), race/ethnicity (non-Hispanic black, other), smoking history (never, \<20 years, \>20 years), daily alcohol intake (\< 2 vs. \>/=2 drinks) and use of non-steroidal anti-inflammatory drugs (ever, never). | 173 |
| Total | 341 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Deemed ineligible after study visit | 3 | 6 |
| Overall Study | Missing all outcome data | 1 | 1 |
Baseline characteristics
| Characteristic | Total | Risk Assessment | Standard Care |
|---|---|---|---|
| Age, Continuous | 56.4 years STANDARD_DEVIATION 6.3 | 56.8 years STANDARD_DEVIATION 6.8 | 56.0 years STANDARD_DEVIATION 5.9 |
| Desired role in decision-making Mostly patient | 111 Participants | 53 Participants | 58 Participants |
| Desired role in decision-making Mostly provider | 54 Participants | 32 Participants | 22 Participants |
| Desired role in decision-making Shared | 176 Participants | 88 Participants | 88 Participants |
| Education Greater than high school | 129 Participants | 59 Participants | 70 Participants |
| Education High school or less | 212 Participants | 114 Participants | 98 Participants |
| Insurance Free care | 41 Participants | 20 Participants | 21 Participants |
| Insurance Medicaid | 86 Participants | 49 Participants | 37 Participants |
| Insurance Medicare | 100 Participants | 54 Participants | 46 Participants |
| Insurance None | 4 Participants | 1 Participants | 3 Participants |
| Insurance Other | 17 Participants | 11 Participants | 6 Participants |
| Insurance Private/HMO | 93 Participants | 38 Participants | 55 Participants |
| Marital status Married/living with partner | 96 Participants | 47 Participants | 49 Participants |
| Marital status Refused | 1 Participants | 1 Participants | 0 Participants |
| Marital status Single/separated/divorced/widowed | 244 Participants | 125 Participants | 119 Participants |
| Race/Ethnicity, Customized Ethnicity : Hispanic | 23 Participants | 10 Participants | 13 Participants |
| Race/Ethnicity, Customized Ethnicity : Non-Hispanic | 317 Participants | 163 Participants | 154 Participants |
| Race/Ethnicity, Customized Ethnicity : Refused | 1 Participants | 0 Participants | 1 Participants |
| Race/Ethnicity, Customized Race : Black | 208 Participants | 104 Participants | 104 Participants |
| Race/Ethnicity, Customized Race : Other | 58 Participants | 27 Participants | 31 Participants |
| Race/Ethnicity, Customized Race : White | 75 Participants | 42 Participants | 33 Participants |
| Sex: Female, Male Female | 180 Participants | 92 Participants | 88 Participants |
| Sex: Female, Male Male | 161 Participants | 81 Participants | 80 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 168 | 0 / 173 |
| other Total, other adverse events | 0 / 168 | 0 / 173 |
| serious Total, serious adverse events | 0 / 168 | 0 / 173 |
Outcome results
Primary
Concordance Between Patient Preference and Test Ordered
Concordance is a measure of the agreement between the patient's test preference and actual test ordered for standard care vs. risk assessment patients. It is defined as the number of patients who had their preferred test ordered.
Time frame: 3 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Standard Care | Concordance Between Patient Preference and Test Ordered | 148 Participants |
| Risk Assessment | Concordance Between Patient Preference and Test Ordered | 147 Participants |
p-value: 0.4Chi-squared
Secondary
Concordance Between Patient Preference and Test Ordered for High vs. Low Risk Patients
Concordance between patient preference and test ordered for high versus low risk patients. It is defined as the number of patients who had their preferred test ordered.
Time frame: 3 months
Population: Patients with cumulative ACNI scores of 5 to 12 were classified as intermediate/high risk (hereafter referred to as high risk), with a mean ACN rate of 8.3% (95% CI, 7.1% - 29.6%). Patients with cumulative scores of less than 5 were classified as low risk, with a mean ACN rate of 3.1% (95% confidence interval \[CI\], 2.4% - 24.1%).
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Standard Care | Concordance Between Patient Preference and Test Ordered for High vs. Low Risk Patients | 86 Participants |
| Risk Assessment | Concordance Between Patient Preference and Test Ordered for High vs. Low Risk Patients | 61 Participants |
p-value: 0.51Chi-squared
Secondary
Provider Satisfaction
Provider satisfaction was assessed based on responses to a 3-item pretest administered prior to commencement of the study and the same 3-item posttest. The 3 items assessed to the extent to which providers felt that personalized risk assessment would be useful for: (1) selecting an appropriate screening test for their average risk patients \[test selection\]; (2) reduce time to decide on an appropriate screening modality \[save time\]; and (3) make them more receptive to patient preferences and possibly order a screening test other than colonoscopy \[receptive to patient preferences\]. Responses were assigned a point value ranging from 5= strongly agree and 1 = strongly disagree.
Time frame: Two years
Population: The difference in the number of providers reflect provider attrition during the 2-year study period.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Standard Care | Provider Satisfaction | Test selection | 4.0 units on a scale | Standard Deviation 0.7 |
| Standard Care | Provider Satisfaction | Save time | 3.2 units on a scale | Standard Deviation 1 |
| Standard Care | Provider Satisfaction | Receptive to patient preferences | 3.8 units on a scale | Standard Deviation 0.7 |
| Risk Assessment | Provider Satisfaction | Test selection | 3.9 units on a scale | Standard Deviation 0.9 |
| Risk Assessment | Provider Satisfaction | Save time | 3.6 units on a scale | Standard Deviation 1 |
| Risk Assessment | Provider Satisfaction | Receptive to patient preferences | 3.8 units on a scale | Standard Deviation 1 |
p-value: >0.05Wilcoxon (Mann-Whitney)
Secondary
Satisfaction With Decision-making Process (SDMP)
SDMP was assessed on the posttest using the validated 12-item Satisfaction with the Decision-Making Process scale. Individual items are assigned a point value ranging from 1 for ''strongly disagree'' (or ''poor'') to 5 for ''strongly agree'' (or ''excellent''). A cumulative score is then calculated based on the summed response scores for each item (maximum score = 60). Data was missing for 11 patients in the concordant group and 6 patients in the discordant group
Time frame: One month
Population: The subgroup of patients who had their preferred test ordered, regardless of study arm or risk-category. The subgroup of patients who had a non-preferred test ordered, regardless of study arm or risk-category.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Standard Care | Satisfaction With Decision-making Process (SDMP) | 52.0 units on a scale | Standard Deviation 5.3 |
| Risk Assessment | Satisfaction With Decision-making Process (SDMP) | 48.9 units on a scale | Standard Deviation 6.1 |
p-value: <0.001Wilcoxon (Mann-Whitney)
Secondary
Screening Intentions
Screening intentions were assessed on the posttest. Patients were asked how sure they were that they would complete the screening test that got scheduled Scores ranged from 5 = ''completely'' to 1 = ''not at all sure.'' Data was missing for 11 patients in the concordant group and 6 patients in the discordant group.
Time frame: 3 months
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Standard Care | Screening Intentions | 4.6 units on a scale | Standard Deviation 0.7 |
| Risk Assessment | Screening Intentions | 4.0 units on a scale | Standard Deviation 1.1 |
p-value: <0.001Wilcoxon (Mann-Whitney)
Secondary
Screening Test Completion
Test completion rates were tracked using BMC's electronic medical record, which captures results for all endoscopic procedures, imaging studies, and stool blood tests.
Time frame: 6 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Standard Care | Screening Test Completion | 109 Participants |
| Risk Assessment | Screening Test Completion | 7 Participants |
p-value: 0.004Chi-squared
Other Pre-specified
Concordance Between Patient Preference for Colonoscopy and Test Ordered
Test-specific concordance between patient preference for colonoscopy and test ordered for standard care versus risk assessment groups. It is defined as the number of patients who had their preferred test ordered.
Time frame: 3 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Standard Care | Concordance Between Patient Preference for Colonoscopy and Test Ordered | 125 Participants |
| Risk Assessment | Concordance Between Patient Preference for Colonoscopy and Test Ordered | 114 Participants |
p-value: 0.51Chi-squared
Other Pre-specified
Concordance Between Patient Preference for Colonoscopy and Test Ordered for High Versus Low Risk Patients
Test-specific concordance between patient preference for colonoscopy and test ordered for high versus low risk patients. It is defined as the number of patients who had their preferred test ordered.
Time frame: 3 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Standard Care | Concordance Between Patient Preference for Colonoscopy and Test Ordered for High Versus Low Risk Patients | 69 Participants |
| Risk Assessment | Concordance Between Patient Preference for Colonoscopy and Test Ordered for High Versus Low Risk Patients | 45 Participants |
p-value: 1Fisher Exact
Other Pre-specified
Concordance Between Patient Preferences for a Screening Tests Other Than Colonoscopy and Test Ordered for High Versus Low Risk Patients
Test-specific concordance between patient preference for a screening test other than colonoscopy (fecal occult blood testing, flexible sigmoidoscopy, double-contrast barium enema, CT colonography and stool DNA) and test ordered for high versus low risk patients. It is defined as the number of patients who had their preferred test ordered.
Time frame: 3 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Standard Care | Concordance Between Patient Preferences for a Screening Tests Other Than Colonoscopy and Test Ordered for High Versus Low Risk Patients | 14 Participants |
| Risk Assessment | Concordance Between Patient Preferences for a Screening Tests Other Than Colonoscopy and Test Ordered for High Versus Low Risk Patients | 15 Participants |
p-value: 0.39Fisher Exact
Other Pre-specified
Concordance Between Patient Preferences for Screening Tests Other Than Colonoscopy and Test Ordered
Test-specific concordance between patient preference for a screening test other than colonoscopy (fecal occult blood testing, flexible sigmoidoscopy, double-contrast barium enema, CT colonography and stool DNA) and test ordered for standard care versus risk assessment arms. It is defined as the number of patients who had their preferred test ordered.
Time frame: 3 months
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Standard Care | Concordance Between Patient Preferences for Screening Tests Other Than Colonoscopy and Test Ordered | 23 Participants |
| Risk Assessment | Concordance Between Patient Preferences for Screening Tests Other Than Colonoscopy and Test Ordered | 29 Participants |
p-value: 0.62Fisher Exact