Results Summary
What was the research about?
Decision aids help people choose between two or more healthcare options based on what is most important to them. Decision aids may include numbers to help show benefits and risks. Researchers don’t know if using numbers makes decision aids more useful or more confusing.
In this study, the research team compared two decision aids that showed risks and benefits of two ways to screen for colorectal cancer, or CRC: colonoscopy or a test for blood in the stool. The team assigned patients by chance to use either a
- Decision aid with numbers to describe the risks and benefits of CRC screening. For example, “Four out of 1,000 people who have a colonoscopy will have to go to the hospital due to heavy bleeding.”
- Decision aid with words instead of numbers to describe risks and benefits. For example, “Very rarely, people having a colonoscopy need to go to the hospital due to heavy bleeding.”
What were the results?
Similar numbers of patients from each group, about 29 percent, got CRC screenings in the six months after they used the decision aids.
Compared with patients who viewed the decision aid that used words, more patients who viewed the decision aid with numbers
- Planned to get a stool blood test
- Felt they were at risk for getting CRC in their lifetimes
- Answered knowledge questions about CRC and screening correctly
The groups didn’t differ in intent to get a colonoscopy, ratings of CRC screening benefits, worry about problems getting CRC screening, or conflict about their screening decisions. Patients’ ability to use numbers didn’t affect the results.
Comparing before and after viewing the decision aids, both groups said they were
- More likely to get some type of CRC screening
- More likely to get a stool blood test
- More convinced of the benefits of CRC screening
- Less worried about problems getting CRC screening
- Less conflicted about their decisions about whether to get screening
In both groups, patients’ knowledge about CRC and screening increased.
Who was in the study?
The study included 688 patients ages 50–75 who were past due for CRC screening. Of these, 61 percent were white, 28 percent were black, 1 percent were Asian, and 10 percent were either multiple races, other races, or did not answer the question. The average age was 59, and 60 percent were women. Patients received care at clinics in two healthcare systems in the Indianapolis area.
What did the research team do?
The research team created the decision aids, which patients could view on a computer. The decision aids had slides with text, pictures, and audio. Patients used either the decision aid that used numbers or the one with words. Patients filled out a survey before and after viewing the decision aid. The team looked at health records six months later to see how many patients had received CRC screening.
What were the limits of the study?
Patients lived in one midwestern state. Results may differ for patients in other areas.
Future research could compare decision aids that do and don’t use numbers for other types of health problems and in other areas.
How can people use the results?
Decision aid designers can use the results of the study to help decide whether to use numbers when presenting information to patients.
Professional Abstract
Objective
To compare the effectiveness of including or not including numerical descriptions of benefits and risks in a colorectal cancer (CRC) screening decision aid on increasing CRC screening
Study Design
| Design Elements | Description |
|---|---|
| Design | Randomized controlled trial |
| Population | 688 patients receiving primary care, ages 50–75, past due for CRC screening |
| Interventions/ Comparators |
|
| Outcomes |
Primary: proportion of patients receiving CRC screening Secondary: intent to get any CRC screening, intent to get stool blood test, intent to get colonoscopy, knowledge of CRC and screening, perceived benefits of and barriers to CRC screening, perceived risk of CRC, decision conflict |
| Timeframe | 6-month follow-up for primary outcomes |
This randomized controlled trial compared the effectiveness of including or not including numerical data in a CRC screening decision aid on increasing the proportion of patients receiving screening among those viewing each decision aid.
The research team randomly assigned patients to one of two groups. In both groups, patients viewed a computer-based decision aid that compared two screening options, colonoscopy and stool blood test, presenting benefits and risks associated with each. For the first group, the decision aid provided this information using numerical descriptions of risk, such as the number out of 1,000 patients who will die from colon cancer or the number having complications following colonoscopy. The second group’s decision aid presented these risks with words only, using terms such as often or some.
The study included 688 primary care patients, ages 50–75, who were past due for CRC screening. Of these, 61% were white, 28% were black, 1% were Asian, and 10% were either multiple races, other races, or did not answer the question. The average age was 59, and 60% were female. Patients received care from clinics in two healthcare systems in the Indianapolis area.
The research team surveyed patients about their knowledge of and attitudes toward CRC screening and screening intent before and immediately after viewing the decision aid. Six months later, the team looked at patients’ electronic health records to see who had received CRC screening.
Patients, patient advocates, and physicians helped create the decision aids and share results.
Results
The two groups did not differ in the proportion who received CRC screening within six months of viewing the decision aid. About 29% of patients in both groups received CRC screening.
Compared with patients who viewed the text-only decision aid, patients who viewed the numerical decision aid had statistically significantly greater increases in
- Intent to get a stool blood test (p=0.007)
- Knowledge of CRC and screening (p=0.007)
- Perceived personal risk of getting CRC in their lifetimes (p<0.001)
The groups did not differ in intent to get a colonoscopy, perceived benefits of and barriers to CRC screening, perceived comparative risk for CRC, and decision conflict. The patients’ level of numeracy did not affect the results.
After viewing the decision aid, compared to beforehand, patients in both groups had significantly (p<0.001)
- Greater intent to get any CRC screening
- Greater intent to get a stool blood test
- Higher ratings of perceived benefits of CRC screening
- Lower ratings of perceived barriers for CRC screening
- Lower ratings of decisional conflict
Limitations
Patients lived in one midwestern state. Results may differ for patients in other areas.
Conclusions and Relevance
Including numerical information in a decision aid did not affect the proportion of patients getting CRC screening, compared with a decision aid without numerical information. Patients showed improvement in most other outcomes after viewing either decision aid.
Future Research Needs
Future research could compare decision aids with and without numerical descriptions of risks and benefits for other types of healthcare decisions.
Final Research Report
View this project's final research report.
More to Explore...
Related PCORI Dissemination and Implementation Project
Journal Citations
Results of This Project
Related Journal Citations
Peer-Review Summary
Peer review of PCORI-funded research helps make sure the report presents complete, balanced, and useful information about the research. It also assesses how the project addressed PCORI’s Methodology Standards. During peer review, experts read a draft report of the research and provide comments about the report. These experts may include a scientist focused on the research topic, a specialist in research methods, a patient or caregiver, and a healthcare professional. These reviewers cannot have conflicts of interest with the study.
The peer reviewers point out where the draft report may need revision. For example, they may suggest ways to improve descriptions of the conduct of the study or to clarify the connection between results and conclusions. Sometimes, awardees revise their draft reports twice or more to address all of the reviewers’ comments.
Peer reviewers commented, and the researchers made changes or provided responses. The comments and responses included the following:
- The reviewers asked for an explanation of the study design. In particular, they wanted to know why the public deliberation portion, which was qualitative, followed the quantitative randomized trial instead of preceding the trial to inform its design. The researchers explained that in this case there was enough previous research to inform the design of the randomized trial, and the results of the trial were then used to inform the public deliberation materials like presentations and the briefing book.
- The reviewers suggested that there may have been different study outcomes based on prior screening experience and encouraged the researchers to describe the sample participants’ screening history at least, if not able to stratify results by screening history. The researchers explained they did not have sufficiently reliable data on participants’ previous screening history to report these results or to stratify by screening history.
- The reviewers expressed concern that no one in the public deliberation panel had opted against colorectal cancer (CRC) screening, which made it difficult to assess whether the presentations to the panel provided a balanced view of screening. The researchers explained that while they strove to maximize the diversity of the panel members in race, age, and educational attainment, they could not assure diversity of screening attitudes. The researchers added a discussion to the report about the lack of participants in the discussion panel who had not undergone CRC screening. The researchers noted that the panel did discuss the risks of screening and why people might choose not to screen.
- The reviewers also noted that the panel was not broadly representative because only a small portion of people contacted responded to the invitation. The researchers agreed that the participants in the panel were not a random sample but noted that in selecting a small subset of the group that responded to the original invitation, care was taken to meet diversity goals. However, the researchers acknowledged selection bias in this process as a study limitation.