Abstract
Public hesitation to engage in advance care planning, consultation with palliative care, and admission to hospice is a significant barrier to improving patient outcomes. In previous study, we derived five empirically supported messaging principles, and in this project we used these messaging principles to design, place, and analyze a 60-second video for the Oregon Coalition for Living Well with Serious Illness. The video was scripted to emphasize a single message that had emerged from the empirical research: that “you should have a say in your care.” We deliberately did not use the term “advance care planning” because our prior focus groups showed that most consumers did not use this term. The marketing plan for the resulting video on Facebook included a paid sponsorship of the full-length video through a Facebook boosted post, and three 10-second ads featuring key lines from the video that tested different elaborations of the central message. Facebook Analytics indicated that the 60-second video was viewed 67,650 times in the week it was promoted. The three 10-second ads, which showed selected moments from the full-length video, were viewed a total of 253,087 times. Of the three 10-second ads, the one emphasizing “What matters to me is being near my family” (65% of clicks) strongly outperformed “I don't want to go out on a machine” and “I've heard so many medical horror stories” (35% for both messages combined) as messages that persuaded viewers to click through the ad to the website. Use of the messaging principles to guide the design and marketing of this Facebook ad shows that (1) the public is interested in serious illness care, (2) that targeted social media can be used to reach a defined public audience, and that (3) this public messaging strategy can be implemented at relatively low cost.
Introduction
Public hesitation to engage in advance care planning, consultation with palliative care, and admission to hospice is a significant barrier to improving patient outcomes. 1 Most consumers have heard of advance care planning, but a minority have engaged in any form of it, and in surveys spanning nearly a decade, the proportion of the public that knows what palliative care is remains unchanged. 2
How our field can engage the public is the type of problem that change makers would characterize as “wicked.”3,4 Wicked problems arise from systems of high complexity where simple linear cause-and-effect relationships fail to describe behavior of the system. Rather than look for a single “solution” to solve a wicked problem, change makers look instead for leverage points where small low-risk experiments that perturb the system can be performed, and the consequences assessed.
In previous study, we identified public engagement as an important leverage point for the future of advance care planning, palliative care, and hospice, and we embarked on a project to develop public messaging with a group of stakeholders* that was based on empirical findings describing consumer perceptions. 5 Our aim was to enable the wide ranging stakeholder group to create messaging that was aligned and persuasive. To conduct this messaging study, we developed with the stakeholders a set of messaging principles that we tested in consumer focus groups. 5 The resulting five messaging principles describe a fresh approach to public messaging 6 intended to introduce consumers to serious illness care. The five messaging principles, which are described in detail elsewhere, are (1) talk up the benefits (rather than try to “convince” with fear), (2) present choices for every step (because you should have a say in your care), (3) use stories (rather than statistics), (4) invite dialogue (that unfolds over time), and (5) invoke a new team (that involves the patient, family, friends, health care professionals, the health care system, social service organizations, faith-based organizations, and other).
In this brief report, we describe the design, marketing, and analytics of a public engagement campaign based on a 60-second video produced for a state-based multistakeholder organization. The University of Washington Human Subjects Institutional Review Board declined to review this project as they considered it consumer marketing rather than research. However, we describe the process here in a research-like format to provide an example of public messaging guided by empirical data.
Methods
Design of a Facebook video
We drew on the five messaging principles to guide the development of a Facebook video sponsored by the Oregon Coalition for Living Well with Serious Illness, which has attracted participation from >50 organizations representing every niche of the serious illness ecosystem in Oregon.
We drew upon existing consumer research to design, script, and produce the video, which was timed to coincide with National Healthcare Decisions Day (April 16, 2021). We defined the target consumer as a 50–70-year old. The advertising was geared to both men and women, although we assumed women were the likeliest audience, as they routinely make health care decisions for spouses, parents, and themselves. To reach this consumer group, we selected Facebook, the social media platform that best enabled us to reach this audience. In designing this video specifically for the Facebook platform, we specified that the maximum length of the video should be no >60 seconds to optimize Facebook views. Our focus group data indicated that “experts” were viewed with caution, so we featured “regular people” as messengers who were ethnically diverse. The ad was scripted to emphasize a single message that had emerged from more than one consumer dataset: that you should have a say in your care. This message responded to a consumer perception in national surveys and multiple focus groups that consumers perceive serious illness decisions as situations where they are forced to make a choice between options that all seem bad. We deliberately did not use the term “advance care planning” as our focus groups 5 demonstrated that consumers did not know what this term means. To emphasize that the messengers in the ad were “people like me,” we shot video on location in Oregon, and specified that the opening scenes would include photos of urban and rural Oregon that locals would recognize. The full 60-second video can be viewed on at (https://www.youtube.com/watch?v=0h48b936rxo).
Social media marketing approach
The marketing plan for placing the video on Facebook included a paid sponsorship of the full-length video through a “Facebook boosted post” (which shows the video to Facebook users in a target audience), and three 10-second ads that were excerpts from the 60-second video featuring key lines from the video that tested different elaborations of the central message. The demographic targeting included adults living in Oregon and Washington states, both genders, individuals of age 30 years or older, and individuals interested in keywords (caregiving, aging, elderly care, and successful aging) related to growing older. The targeting was deliberately broader than the design target because we sought to cast a wide net for this messaging experiment. The paid sponsorship of the video, starting on April 15, 2021, ran for one week.
The three 10-second ads, which ran from April 15 to May 15, 2021, featured different narrator-messengers talking about what was important to them in their medical care. The three messengers talked about the following topics that mattered to them: (1) “Being with my family,” (2) “I don't want to go out on a machine,” and (3) “I've heard so many medical horror stories lately.” The call to action at the end of the ad was designed as a “click-through” link on the accompanying Facebook post text that took users to a newly launched website for the Oregon Coalition for Living Well with Serious Illness (https://oregonlivingwell.org/), which pointed users to curated resources (PREPARE [https://prepareforyourcare.org], The Conversation Project [https://theconversationproject.org], and the National Patient Advocate Foundation care plan [https://www.npaf.org/patients-and-caregivers/care-planning/])
Analytics
We assessed the campaign using data about user behaviors reported by Facebook Analytics. The data available included views (what we report in this article is the number of Facebook users who watched the video all the way through), “click-throughs” (the number of times a user who watched a 10-second ad or the 60-second video clicked on the accompanying link), and cost per click-through (the cost of the campaign in dollars and cents divided by the number of click-throughs). (Note: Facebook retired this tool June 30, 2021, and has replaced it with Facebook Business Suite and Ads Manager)
Results
Consumer reach of the video and ads
The 60-second video was viewed 67,650 times in the week it was promoted. The video viewers were 76% female, and their age distribution was as follows: 14% were 30–44 years old, 19% were 45–54 years old, 31% were 55–64 years old, and 36% were age 65 years or higher. The geographic distribution of the viewers was 58% from Washington state, 42% from Oregon.
The three 10-second ads were viewed a total of 253,087 times. In setting up this campaign, we allowed a Facebook algorithm to identify the 10-second ad that was followed by the most click-throughs, and increase the frequency of that particular 10-second ad. The Facebook algorithm, which is built using factors that are not visible to advertisers, identified “What matters to me is being near my family” as the most desirable message, and it was viewed 170,224 times, with 2571 click-throughs to the website. The message “I don't want to go out on a machine” was viewed 61,774 times, with 993 click-throughs to the website. The message “I've heard so many medical horror stories lately” was viewed 21,089 times, with 407 click-throughs to the website. For all three 10-second ads, viewers were 65% female, with an age distribution as follows: 15% were 30–44 years old, 21% were 45–54 years old, 26% were 55–64 years old, and 38% were age ≥65 years old. The geographic distribution of the ad viewers was 68% from Washington, 32% from Oregon.
Compared with Facebook averages, viewer engagement measured by click-throughs was extremely cost-effective. At the time of this analysis, the Facebook average click-through rate was ∼0.90% and the average cost per click was ∼$1.72. This Oregon Coalition ad campaign had a much higher average click-through rate at 1.57% and a much lower cost per click at $0.74. The cost for sponsoring the full-length video came out to $0.02 per click.
Discussion
To our knowledge, this is the first report of an empirically designed public engagement intervention for serious illness care that includes social media analytics. Overall, use of the messaging principles to guide the design and marketing of this Facebook ad shows that (1) the public is interested in serious illness care, (2) that targeted social media can be used to reach a defined public audience, and that (3) this public messaging strategy can be implemented at relatively low cost. Like a larger randomized study of three different video messages about COVID-19 targeted to Black and Latinx communities, 7 our experience emphasizes the importance of using a tested message, and messengers who resonate with the target audience; in the randomized study, participants were recruited by a survey company and shown videos after recruitment to a study. The data we report here show how a tested message performs in the real world of social media.
Interestingly, the social media analytics comparing views from the three 10-second ads provide empirical proof of something that other studies describe: that consumers are not attracted to dying as an introduction to serious illness care.2,8–10 The Facebook algorithm selects the best performing ads and begins showing them more frequently, so these data on views do not represent a consumer choice based on viewing all three ads. However, according to the Facebook algorithm, the introductory message “What matters to me is being near my family” (65% of clicks) strongly outperformed “I don't want to go out on a machine” and “I've heard so many medical horror stories” (35% for both messages combined) as messages that persuade a consumer to click through the ad to the website.
The major limitation of this project is that it was not designed to track behaviors beyond the call to action, which was for a user to click through the video to the Oregon Coalition website. Ideally, public engagement will result in an engaged consumer taking an action that is associated with an improved outcome. However, the reality of public engagement work is that we are dealing with a problem that is wicked: we would not expect a Facebook ad to result in a completed advance directive—other follow-up steps would need to be linked to that first click-through. We appreciate that advance care planning is controversial as a stand-alone intervention to improve patient outcomes,11,12 and yet if we take the long view of public engagement as essential for social change, our field must find ways to encourage and normalize conversations about care for serious illness more broadly among consumers.
This report demonstrates that a video and ad campaign designed using empirically tested messaging principles could be used as a first step in a new paradigm for public engagement in advance care planning. In just two states with a total of 11.8 million people (Washington, 7.6 million residents, Oregon 4.2 million residents), we delivered a hopeful message about serious illness care to >250,000 people at a modest cost. This same approach could be used to design public campaigns for palliative care, or hospice, or any other service needed by people living with serious illness.
Footnotes
Acknowledgment
We thank Peggy Maguire who founded and leads the Oregon Coalition for Living Well With Serious Illness.
Funding Information
This study was supported by Cambia Health Foundation and The John A Hartford Foundation.
Author Disclosure Statement
No competing financial interests exist.