Currently, healthcare systems are striving to engage patients as partners in order to make healthcare more patient-centered. There is strong evidence that shared decision-making (SDM) improves both patient satisfaction and health outcomes. In order for patients to actively participate in SDM, they need to be informed regarding the pros and cons of available treatment options. Despite the availability of information from multiple sources (e.g., clinicians, patient education materials, social media, friends and family), many patients still do not have this type of information available to them.
We believe a new type of decision aid that safely puts the power of a large clinical data repository into the hands of patients could help meet the information needs for SDM. Patient decision aids improve people's knowledge of options, create accurate perceptions of benefits and harms, reduce difficulty in decision-making, and increase participation in the decision-making process. Current patient decision aids have limitations including: 1) need for experts to summarize the medical evidence; 2) labor required to produce frequent updates; 3) limitations and contradictions of the medical evidence. In addition, few decision aids provide stories about patients in similar situations.
To address these limitations, we envision an extendable and engaging patient decision aid called VeteransLikeMe (VLMe).
1. Analyze the SDM needs of Veterans with atrial fibrillation (AF).
2. Create a mockup design for visual displays and vignettes to communicate outcomes of interest to Veterans with AF.
3. Conduct an uncontrolled open trial of the mockup VLMe to assess acceptability, feasibility and safety.
Currently, there are two different avenues of interest for this research: the patient perspective regarding their condition and the development of the tool; and the clinic visits and the post-clinic visit interviews are being analyzed for congruence. The Methods for both are described below.
Patient Facing Methods: 25 patients participated in semi-structured interviews immediately after a cardiology clinic visit for treatment of atrial fibrillation/flutter. The interviews included questions regarding the patient's history of atrial fibrillation/flutter, their discussion of diagnostic and treatment options with their physician, and the factors they weighed in the decision-making. Participants were also shown a paper mock-up of the "Veterans Like Me" interface and were asked specific questions regarding design aspects of the system. Question related to the tool design included: which attributes of the cases presented make these people "like you"?, what other attributes would you like to able to filter cases on? when comparing potential treatments which outcomes would you like to see? What do you think of this interface/ tool and how would you envision using it? And do you have specific ideas on the interface design? . We analyzed the transcripts of these interviews focusing on themes relevant to the design of the tool.
Clinic Visit and Interpretation Methods: Consultations for treatment of a-fib were audio-recorded. Visits were transcribed and deductive content analysis was conducted based on self-determination theory to abstract themes of patient autonomy, competence and relatedness. Additional coding will be conducted to identify shared decision making behaviors.
Patient Facing Results: Overall participants were interested in using the tool, expressed enthusiasm for the ability to see both aggregated outcomes as well as individual's stories, and believed that the tool would assist them in their understanding and decision-making. In our analysis we found many information needs that, with redesign, the tool could address. These included the natural history of the disorder, common treatments, and treatment trajectories. When examining drivers of patient's decisions we found that mortality risk, side effects and the invasiveness of treatments were foremost in patients' minds. Our analysis also suggested several factors that patients wanted to control when viewing "patients like me", these included attributes that might be clinically related to the disorder (e.g. bodyweight, exercise history and alcohol use) as well as information for which there is no known association between the behavior and the disorder in question (e.g. agent orange exposure). Patients also expressed interest in a range of outcomes that were not considered in the original design including changes in patient reported symptoms with different treatments, medication side effects and interactions amongst medications.
Clinic Visit and Interpretation Results: Nine consultations of patients with four different cardiologists were recorded. Patients were male, averaged 62 years, and a-fib diagnoses ranged from one to 13 years (SD = 5.2). Three cardiologists were male and one was female. Preliminary content analyses related to self-determination theory revealed four themes informing treatment decisions. (1) Patients' experience and narratives of illness (competence): My period of staying out of a-fib is shortening over time; (2) Patients' level of self-confidence in their ability to practice treatment related health behaviors (competence): I am good at pill taking. It's automatic; (3) Assertion of life style and preferences (autonomy): I like exercising and want to start again; Coming down to get jump-started every couple of weeks is inconvenient; and (4) Elicitation of doctor's advice (relatedness): Knowing what you know, what would you do? SDM coding will be conducted next.
Summary of Key Findings:
Veterans have significant gap in anticoagulant knowledge;
Veterans have desire to learn from Veterans Like themselves;
Veterans would like to participate in the definition of "like me" and would like to see patient-reported outcomes such as functional status;
Veterans Like Me helps patients with decision making and majority of the patients find the stories helpful as well.
The overarching goal of this project is to enhance SDM between Veterans and healthcare providers. We hypothesize that we can empower patients to participate in SDM by providing graphical information regarding outcomes of interest along with concrete stories about individuals like themselves. VLMe will retrieve relevant cases from a clinical repository and use interactive graphic displays to communicate summary statistics and automatically composed patient stories to facilitate patient engagement. Our vision is to use "big clinical data" to help inform patients.
While we still hold to the vision described above, more work is needed. Future work will evaluate use of the Veterans Like Me tool use in terms of decisional conflict and patient satisfaction. Deciding how the tool will be used during clinic flow will be particularly important . Probably put the that reversibility of therapy affects the usage pattern of the decision aid- for a fib it might be used over time to monitor symptoms, as well as over time due to changes in therapy cause by the natural history of the disorder.
Ultimately, Veteran care can be improved by the development of VLMe tool that can be used to inform clinical decisions and support education.
- Zeng-Treitler Q, Gibson B, Hill B, Butler J, Christensen C, Redd D, Shao Y, Bray B. The effect of simulated narratives that leverage EMR data on shared decision-making: a pilot study. BMC research notes. 2016 Jul 22; 9:359.
- Butler J, Gibson BS, Bray B, Ellington L, Zeng Q. Veterans like Me: Development of a Patient-Facing Decision Support Tool for Veterans Seeking Treatment for Atrial Fibrillation. In T. Bastiaens, Ed. Proceedings of E-Learn: World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education 2014 (pp. 299-302). Chesapeake, VA: Association for the Advancement of Computing in Education; 2015 Oct 20. Available from: https://www.learntechlib.org/p/148794.