Impact of a Decision Aid on Colorectal Cancer Screening

Colorectal cancer (CRC) is the second-leading cause of cancer-related deaths in the United States and the third most common cancer in men and women according to the Centers for Disease Control and Prevention. Almost all colorectal cancers begin as pre-cancerous polyps (abnormal growths) in the colon or rectum. Pre-cancerous polyps can exist in the colon for years without causing any symptoms before invasive cancer develops. CRC screening can be used to find pre-cancerous polyps so they can be removed before they develop into cancer, as well as to find colorectal cancer early to maximize the potential for treatment to be effective. Regular screening, beginning at age 50, is key to the prevention and early treatment of colorectal cancer. Unfortunately, about one in three adults in the United States between the ages of 50 and 75 years (23 million) is not getting screened as recommended. Studies show that people who are able to select the colorectal test they prefer are more likely to complete screening than people whose screening test is selected for them. Decision aids help patients make informed, value-concordant choices about a particular course of action based on an understanding of potential benefits, risks, probabilities, and scientific uncertainty. Studies show that decision aids for CRC screening enable patients to identify a preferred screening test, reduce decisional conflict around the screening test, and increase interest in screening.

The Impact of A Decision Aid on Colorectal Cancer Screening is an interactive, web-based multimedia tool that uses a series of brief video presentations to provide information on the importance of CRC screening, the different screening tests available (e.g., fecal occult blood testing [FOBT], flexible sigmoidoscopy, colonoscopy), and the relative risks and benefits associated with each screening test. The decision aid was designed to be used by 50- to 75-year-old patients who are due for CRC screening when they arrive for a routine office visit with their primary care provider. In the research reviewed for EBCCP, the decision aid was formatted on DVD for use with portable computer stations to maximize flexibility of delivery in two primary care settings. Currently, the decision aid is publicly available online. The decision aid comprises modules in which professional actors play the role of a Black, Hispanic female moderator and a White, non-Hispanic male physician to convey colorectal cancer screening information via video, animation, and graphics. The web-based decision aid modules include the following:

-- An introduction that discusses the importance of screening, purpose of the decision aid, and instructions for navigating the different modules

-- An overview of the epidemiology of CRC, natural history, rationale for screening, benefits of screening, availability of multiple screening options, and the lack of consensus regarding a best screening method

-- Brief and detailed descriptions of each of the six different screening tests endorsed by the American Cancer Society, US Multi-society Task Force on Colorectal Cancer, and the American College of Radiology

-- Comparisons of each screening method with respect to individual test features, including accuracy, inconvenience, discomfort, recommended frequency of testing, complications, and need for further diagnostic studies if the screening test result is positive

-- A decision-making module in which users are asked to identify a screening preference and prioritize test features influencing their screening test selection

-- Patient vignettes from a racially diverse group describing personal experiences with the tests

Decision aid users can navigate forward or backward through the tool using a mouse and print key information.

-- 20-40 minutes to view the decision aid on CRC screening prior to a regularly scheduled visit with a primary care provider
-- 10 minutes during the visit with the primary care provider to discuss CRC screening and identify a preferred screening test

The Impact of A Decision Aid on Colorectal Cancer Screening targets primary care patients, aged 50-75 years, who are at average risk for colorectal cancer (i.e., with no personal or family history in a first-degree relative of colorectal cancer or polyps, no history of chronic inflammatory bowel disease) and are due for a CRC screening test.

The Impact of A Decision Aid on Colorectal Cancer Screening can be implemented in a medical facility (e.g., primary care office) with a health care educator or patient navigator onsite to answer questions from patients viewing the website or having difficulty navigating the website. Because the decision aid resides on a public website, it can also be self-administered for free anywhere with an internet connection.

Materials required for implementation include:
-- Decision Aid User's Guide
-- Web-based decision aid

A randomized controlled trial evaluated the effects of the decision aid on colorectal cancer screening among average-risk patients due for CRC screening and scheduled for a regular office visit with their primary care provider at one of two urban outpatient health care sites. Of the 13,518 patients identified as potentially eligible for CRC screening because of their age, 825 patients were enrolled into the study and randomized. Eight-four percent of the randomized patients were less than 65 years old, 59% were female, 62% were Black, 34% were White, and 2% were Asian. Five percent of patients randomized were Hispanic, and 95% were non-Hispanic. Seventy-seven percent of those randomized had at least a high school degree; 35% were married or living with a partner; and 67% were covered by Medicare, Medicaid, or the Massachusetts Commonwealth Care (formerly "Free Care") program. Eighty-seven percent of patients randomized had no prior FOBT. In regard to a preferred approach to selecting a screening option, 27% preferred a patient-dominant decision-making approach, and 53% preferred a shared decision-making approach.

Several methods were used to recruit patients. Most of the patients (n=796) were recruited through a monthly audit of electronic medical records at both sites that identified patients due for a CRC screening 2-4 weeks ahead of a regularly scheduled visit. These patients were contacted directly by telephone if clearance for CRC screening was provided by the patients' primary care physician. Patients expressing interest in the study were provided with a brief overview, evaluated for eligibility, and invited to participate. Patients were ineligible to participate if they had met any one of the following conditions: (1) a prior CRC screening by any method other than FOBT; (2) a personal history of CRC or polyps, a family history of CRC or polyps involving one or more first-degree relatives, or chronic inflammatory bowel disease; (3) lack of fluency in written and spoken English; or (4) medical comorbidities that precluded CRC screening by any recommended method as determined by the patient's primary care provider.

Study patients arrived 1 hour before their regularly scheduled visit. After informed consent was obtained, patients were administered a pre-test and were randomized to one of three conditions—two interventions and a usual care control—stratified by health care provider. One intervention condition used the DVD version of the decision aid alone, and the other used the same version of the decision aid plus the Your Disease Risk (YDR) a risk assessment tool (also on DVD in the study) that provides personalized feedback about CRC risk. The investigators hypothesized that adding the YDR tool might result in higher rates of screening completion than using the decision aid alone based on prior evidence suggesting that individualized risk communication increases completion rates of CRC screening. Patients in the usual care control condition reviewed a modified, online version of "9 Ways to Stay Healthy and Prevent Disease", which described lifestyle changes other than CRC screening to minimize the risk of developing preventable diseases. Patients completed their assigned interactive computer session and then immediately met with their primary care provider to discuss screening and identify a preferred CRC screening test strategy. All patients handed their primary care provider a written notification that they were participating in the study to ensure that CRC screening was discussed, but no information was provided regarding study group assignment. After the visit, patients completed a post-test.

The primary outcomes targeted were ordered CRC screening test rates 1, 3, 6, and 12 months after the clinic visit and completed CRC screening test rates 6 and 12 months after the clinic visit. These outcomes were tracked using electronic clinical data reporting systems (electronic medical records), which captured results for all endoscopic procedures, imaging studies, and FOBTs completed at the two study sites.

Additional outcomes targeted included satisfaction with the decision-making process, CRC screening knowledge, and CRC screening intention. These outcomes were assessed using pre- and post-tests, 10-minute, paper-based tests administered using a structured interview format. The pretest included 28 questions to assess knowledge, beliefs, attitudes, and behaviors related to CRC screening, as well as the level of desire for participating in decision making related to CRC screening. The post-test administered after the physician visit reassessed CRC screening knowledge, beliefs, and attitudes but also assessed whether CRC screening was discussed with the primary care provider, whether a screening test was chosen, patient satisfaction with the decision-making process, and CRC screening intentions.

Specifically, CRC screening knowledge was assessed at pre- and post-test using 12 "true/false/don't know" items that asked about CRC risk factors, the rationale and goals of CRC screening, and age at which screening should begin. The knowledge questionnaire was created for the study using content derived from key messages endorsed by the National Colorectal Cancer Roundtable and the Massachusetts Department of Public Health for public education. An example of one of the items is "CRC is the number 1 cause of cancer death among nonsmokers", with the correct response being "true". A cumulative knowledge test score is calculated by summing correct responses. Scores range from 0 to 12, with higher scores indicating greater knowledge.

Patient satisfaction was assessed at post-test only using the 12-item Satisfaction with the Decision-Making Process (SDMP) Scale. Patients rated each item on a 5-point scale ranging from 1 ("strongly disagree" or "poor") to 5 ("strongly agree" or "excellent"). An example of one of the items is "How would you rate your doctor's interest in you and your choice of colorectal screening test?" Individual item ratings are summed across all 12 items to generate a cumulative score. Scores range from 1 to 60, with higher scores indicating greater satisfaction.

CRC screening intentions were assessed at post-test only by asking patients how sure they were on a 5-point scale from 1 ("not at all sure") to 5 ("completely sure") that they would schedule a CRC screening and that they would complete the screening.

At pre-test, there were no statistically significant differences among patients in the three conditions for age, gender, ethnicity, race, marital status, education, insurance coverage, prior FOBT, or decision-making preference.

• A higher percentage of patients in the decision aid only intervention group than in the control group had a CRC screening test ordered at the 1-month (69.1% vs. 60.5%, p<.035), 3-month (71.8% vs. 62.3%, p=.019), 6-month (77.0% vs. 65.2%, p=.002), and 12-month (80.7% vs. 71.4%, p=.011) follow-ups.

• A higher percentage of patients in the decision aid only intervention group than in the decision aid plus YDR intervention group had a CRC screening test ordered at the 1-month (69.1% vs. 60.4%, p<.031), 6-month (77.0% vs. 67.1%, p<.010), and 12-month (80.7% vs. 73.6%, p=.048) follow-ups. The percentage of patients with CRC screening test orders did not differ significantly between the two groups at the 3-month follow-up (71.8% vs. 64.3%, n.s.).

• A higher percentage of patients in the decision aid only intervention group than in the control group completed a CRC screening test at the 6-month (34.2% vs. 26.4%, p=.049) and 12-month (43.1% vs. 34.8%, p=.046) follow-ups. The percentage of patients completing a CRC screening test did not differ significantly between the two intervention groups at the 6- and 12-month follow-ups.

• Scores on patient satisfaction with the decision-making process (SDMP Scale scores) were higher for patients in each of the intervention groups than for patients in the control group (50.7 for the decision aid only group and 50.5 for the decision aid plus YDR group vs. 46.7 for the control group, p<.001). Satisfaction scores did not differ significantly between the two intervention groups.

• Among patients preferring a patient-dominant approach in the decision-making process, scores on satisfaction with the decision-making process were higher for patients in each of the two intervention groups than for patients in the control group (50.2 for the decision aid only group and 49.4 for the decision aid plus YDR group vs. 46.0 for the control group, p=.01). Among patients preferring patient-dominant decision making, satisfaction scores did not differ significantly between the two intervention groups.

• Among patients preferring a shared patient-doctor approach in the decision-making process, scores on satisfaction with the decision-making process were higher for patients in each of the two intervention groups than for patients in the control group (50.6 for the decision aid only group and 50.8 for the decision aid plus YDR group vs. 46.6 for the control group, p<.001). Among patients preferring shared decision making, satisfaction scores did not differ significantly between the two intervention groups.

• CRC screening knowledge test scores at post-test were higher for patients in each of the intervention groups than for patients in the control group (10.9 for the decision aid only group and 10.7 for the decision aid plus YDR group vs. 8.6 for the control group, p<.001). Knowledge test scores at post-test did not differ significantly between the two intervention groups. Baseline knowledge test scores did not differ significantly among the three study groups (7.9 for the decision aid only group, 7.7 for the decision aid plus YDR group, and 7.5 for the control group).

• Scores on intention to schedule a CRC screening test at post-test were higher for patients in each of the intervention groups than for patients in the control group (4.4 for the decision aid only group and 4.3 for the decision aid plus YDR group vs. 3.9 for the control group, p<.001). Scores on intention to schedule a CRC screening test did not differ significantly between the two intervention groups.

• Scores on intention to complete a CRC screening test at post-test were higher for patients in each of the intervention groups than for patients in the control group (4.3 for the decision aid only group and 4.4 for the decision aid plus YDR group vs. 4.0 for the control group, p<.001). Scores on intention to complete a CRC screening test did not differ significantly between the two intervention groups.