Dissemination of Decision Aids about Hip and Knee Osteoarthritis to Spanish-Speaking Adults through Social Media

Abstract

Background

Older adults and Hispanic individuals are increasingly turning to social media platforms to access health-related information. The purpose of this project was to evaluate a social media campaign to disseminate information from decision aids (DAs) on hip and knee osteoarthritis to Spanish-speaking adults.

Methods

A social media marketing team helped create an 8-mo campaign posted across 3 social media platforms to promote visits to a Web site offering free multilingual DAs for treatment of hip or knee osteoarthritis. Eight videos featuring Spanish-speaking actors discussing DAs were boosted (i.e., money was paid so posts landed on user’s feeds). In this observational study, metrics tracked reach, extended engagement, and number of users accessing the Web site. Videos were qualitatively coded for themes, including pain and benefits of treatment options; biserial correlations assessed relationships between the presence of a code and the metrics. We calculated cost per visitor using the total campaign cost and number of users accessing the Web site.

Results

During the campaign, boosted videos reached 83,937 users. Of the users, 22,777 had extended engagement with the videos, and 1,835 users visited our Web site with access to the DAs. Videos that included themes of pain tended to reach more users (r = 0.88) and have higher engagement (r = 0.70). When videos included the theme of benefits of treatment, more users tended to visit our Web site (r = 0.78). The total campaign cost was $25,950, making the cost per Web site visitor $14.14.

Conclusions

Social media was a useful, inexpensive tool for disseminating health care information on hip and knee osteoarthritis DAs to predominantly Spanish-speaking adults. Further work should identify how exposure to such social media content affects a viewer’s health care attitudes, health care behaviors, and surgical decision making.

Highlights

• An 8-mo social media campaign was able to reach more than 83,000 users and result in almost 2,000 users accessing a Web site with multilingual decision aids.

• Social media may be an inexpensive tool for disseminating health care information such as decision aids to Spanish-speaking adults.

Social media has become a dominant source of health information and is used by millions of Americans to obtain information on a range of health conditions, seek medical advice from the community, share health experiences, and receive emotional support.^1–4 Two populations—Hispanic individuals and older adults—may be the most likely to benefit from the use of social media for health information. Hispanic individuals rank among the top users of social media platforms in the United States. ⁵ Social media provides an “intersection between entertainment, community, and language” that Hispanic individuals seek and that can be leveraged for providing trustworthy health information. ⁶ In fact, more than a dozen studies have used social media as a method to promote health information in Hispanic populations. ⁵ Older adults are also increasingly turning to social media platforms, such as Facebook and Twitter, to access health-related information.^7–10 A survey of older adults found that the Internet is the most preferred source of health information and is often used to obtain information on symptoms, prognosis, and treatment options. ⁸ Health-related social media use has been shown to play a positive role in the health of older adults and may improve quality of life and subjective well-being.^11–13 By allowing individuals to take an active role in their health management and share health information with others, social media can instill a sense of empowerment and meaningfulness.^14,15 Further, social media approaches to empowerment may be effective to promote health equity since social media can remove geographic and physical access barriers. ¹⁶

However, research has shown that older adults and Hispanic individuals are underinformed and underinvolved when making key medical decisions.^17–20 For example, one study found that surgeons’ conversations with older adults about hip and knee osteoarthritis (OA) were often lacking discussion of reasons not to have surgery and of nonsurgical options for treatment. ²¹ Thus, one approach to improving both involvement and knowledge is empowering and preparing patients for these types of conversations. For example, patient decision aids (DAs) can help patients understand that a decision needs to be made; their options, risks, and benefits; and how their own values and preferences should play a role in the decision-making process. Studies have repeatedly shown that that DAs increase patient knowledge, accuracy of risk perceptions, likelihood of receiving preference-concordant care, and active participation in decision making. ²² When DAs have been made available through the Internet, they have been found to improve a person’s knowledge without requiring additional effort compared with viewing the DA in a clinic setting. ²³

Despite the increasing use and clear benefits of social media to improve health and education, few projects, that we are aware of, have used social media as a dissemination platform for DAs,^24,25 and only 1 study that we are aware of has specifically disseminated DAs. ²⁶ Herein, we examine the impact of a social media campaign to reach Spanish-speaking adults about hip or knee OA in the Boston area, raise awareness of shared decision making, and connect them with DAs.

Methods

Campaign Creation

The goals were to raise awareness of DAs among Spanish-speaking and Hispanic individuals in the Boston area and to empower individuals to be active participants in medical decisions by emphasizing that they are the expert of their knee and hip experiences, that their voice matters in medical decisions, to encourage them to use DAs and to talk with their doctor about treatment choices.

We hired a full-service marketing agency specializing in social media, TLC Marketing Consultants, who was unaffiliated with our hospital, to create our social media campaign. After discussing the goals of our project, the TLC Marketing team worked to identify the demographics of our intended audience (e.g., men and women aged 30–65+ y, in the Boston area with a focus on Hispanic areas) and platforms (Facebook, Instagram, and Twitter) for our project. Although we expected that the content would be most applicable to older adults (65+ y old), we included younger adults in the intended audience as well expecting that they might have parents or family members who would find the DAs relevant.

They first created a logo and selected complementary fonts and complementary colors for our social media campaign; they created a business Facebook page (https://www.facebook.com/HablaConTuDoctor), a Twitter profile (https://twitter.com/HablaConTuDoc), and an Instagram profile (https://www.instagram.com/hablacontudoctor/) for the campaign, as well as a landing page off of the Health Decision Sciences Center Web site for the campaign (https://mghdecisionsciences.org/habla-con-tu-doctor/). The landing page was then programmed with more information about DAs and a “contact us” option.

We partnered with Healthwise, a DA vendor, and secured an agreement to provide DAs for hip and knee OA in English, Spanish, and Portuguese (as there is a large proportion of Portuguese-speaking individuals in the Boston area) on the Web site for the duration of the campaign. The 6 DAs (2 topics in 3 languages) that were posted on the Web site were accessible in pdf format. The DAs provide information on treatment options including conservative options of low-impact exercise, physical therapy, pain medicines, steroid shots, acupuncture and massage, and dietary supplements, as well as surgery. The DAs did not discuss doing nothing as a possible option. The DAs also included a section for patients to answer a few factual questions about what they learned after reviewing the DA. This was followed by a section containing questions for patients to reflect on what matters most to them. These DAs were developed by Healthwise and were written at a sixth-grade reading level. The DAs were created to be used by patients alone or jointly with their clinician. The Ottawa Hospital Research Institute previously evaluated the DAs according to the International Patient Decision Aid Standard criteria and indicated that the DAs meet 7 of 7 criteria to be defined as a patient DA and 9 of 9 criteria to lower the risk of making a biased decision.^27,28

The TLC Marketing team created 9 mo of social media content (including a 1-mo pilot followed by an 8-mo campaign); each day of the week focused on a theme. Based on our goals, TLC marketing developed about 7 d worth of content each week. They aimed for an 80/20 split of fun content (80%) and health-related content (20%) each week to encourage users to continue to engage with the accounts. ²⁹ The fun content included cute animal pictures, funny memes, sports, and holidays and national days selected to resonate with the intended audience.

Weeks consisted of themed content, examples of which can be seen in Table 1. Note that all material posted on the platforms were in Spanish; however, we are reproducing them here with English translations beside the posted Spanish language content for ease of reader interpretation. Content was reviewed by our team of experts, including an expert in shared decision making, an orthopedic surgeon, 2 primary care physicians (1 who worked at a community health center that serves a large Spanish-speaking population), and a physical therapist. Once the content was approved by our team, the TLC Marketing team translated the materials into Spanish and held focus groups in Spanish with Spanish-speaking individuals on the East Coast of the United States to review the Spanish-language materials for understanding, acceptability, and interest. In addition, TLC worked with the group to secure actors who appropriately “mirrored” our intended audience. These 4 actors then recorded the 8 educational health videos for the campaign. The videos were purposefully brief with the average length being 48 s (standard deviation = 9 s) and ranging from 37 s (November’s video) to 62 s (September’s video). One video was released and boosted on Facebook each month.

Table 1.

Examples of Social Media Campaign Content

Day	Theme	Post Image	Post Text (in English)
Monday	Memes		“When you were going to go to the doctor, but you had a bad hair day”
Tuesday	Hip & Knee Pain Information		“Tired of that knee pain. Talk to your doctor about arthritis treatment options” If you’re struggling with hip or knee pain, we can help. @ Mass general hospital, you matter. We want the best result for you. Visit the link below for more information. https://mghdecisionsciences.org/habla-con-tu-doctor/ #Hablacontudoctor
Wednesday	Boston Area Sports		Congratulations to the Bruins on their victory last night against the Winnipeg Jets! Goalkeeper Jeremy Swayman had 35 assists and secured his third shutout of the season. #BostonBruins #BostonProud https://www.boston.com/…/03/17/bruins-beat-jets-takeaways/
Thursday	Knee and Hip Educational Videos		@ Mass general hospital, you matter. We want the best result for you. Visit the link below for more information. https://mghdecisionsciences.org/habla-con-tu-doctor/ #Hablacontudoctor #MassachusettsGeneralHospital 45-s video, scripted to state, “Decision aids help patients understand a medical decision that needs to be made. Decision aids also help patients understand the risks involved in the decision and the possible results of the medical procedure. Finally, decision aids help people to consider their personal values in relation to the procedure. Remember, always talk to your doctor! A shared decision is a good decision.”
Friday	Health Tips		Winter is the worst season for infectious diseases. Wash your hands frequently to avoid the spread of disease. For more health tips visit: https://mghdecisionsciences.org/habla-con-tu-doctor/ #MassachusettsGeneralHospital #Hablacontudoctor
Saturday	Empowerment Quotes		“At the end of the day, we can endure much more than we think we can.” #Hablacontudoctor #MassachusettsGeneralHospital
Sunday	Hip & Knee Pain Stretches		Quadriceps stretching Purpose: This stretch targets the quadriceps, the muscles in the front of the thighs. Performing this move can help improve flexibility in the hip flexors and quadriceps. Stretching your hips will allow you to stand straighter and walk more easily. Reps: 3 times each side Hold: 30 seconds each time. To do this stretch-up: 1. Stand against a wall or use a chair or counter to lean. Your feet should be shoulder-wide apart. 2. Bend down one knee so your foot goes up to your glutes. 3. Grab your ankle and gently pull it down your glutes as comfortably as you can. 4. Hold the position for 30 seconds while maintaining a straight line between the shoulder and hip. 5. Return to original position and change legs. 6. Repeat 3 times on each side. For more health tips and information please visit below link.” #Hablacontudoctor #MassachusettsGeneralHospital #Estiralo https://mghdecisionsciences.org/habla-con-tu-doctor/

When all content was finalized, the TLC Marketing team scheduled all posts and the campaign officially launched on June 15, 2022. Content was released on an almost daily basis on each platform, and 1 post each month was boosted on Facebook to encourage users to click the link to our Web page. For most months, the boosted post on Facebook was an educational video.

Campaign Metrics

Throughout the campaign, TLC Marketing provided us with monthly social media reports. For Facebook posts that were “boosted” (i.e., paying a small amount to get the post on people’s feeds), we received information on

• how many users each boosted post reached (i.e., the number of users’ feeds on which the post appeared),

• how many users engaged with that post (e.g., watched any amount of the video),

• how many users engaged with at least 3 s of that post (e.g., watched at least 3 s of the video), and

• how many users clicked the link to the campaign’s landing page (note this number does not reflect individual users, solely the number of times the link was clicked, which may slightly overestimate the actual number of individuals who reached our page). ³⁰

In addition, we received information on the number of fans (on Facebook) or followers (on Twitter and Instagram) who were following our campaign profiles and the engagement with the posts (e.g., likes/comments/reactions).^30–32 It is important to note that our goal in boosting these posts was to ensure as many users reached our landing page as possible. It was not our goal to encourage fans/followers, as we did not have plans to continue the social media channels past the campaign. As such, we chose to boost for maximum reach and impressions to reach the largest number of people possible. From our Web site hosting service, we were able to get a report on the number of unique users who accessed the page.

We calculated an estimate of the cost per Web site visitor who may have seen the DA via the social media campaign by dividing the total campaign cost by the number of Web site users.

Analysis of Campaign Content

We did not report on the metrics from the first month pilot run, as many metrics were incomparable to the later months. For example, there were more than 700 visits to the Web site during that month, mostly due to traffic from our own team as we formatted and revised the page. Also, no videos were completed during this first month, so we boosted 2 image posts, which may have slightly inflated the summed numbers for reach (10,526), engagement (3,623), extended engagement (3,317), and clicks (306). Descriptive statistics were used to describe the metrics over 8 mo of the campaign, and Pearson correlations were used to determine the relationship between the metrics of interest for the 8 mo of the campaign. No missing data were present.

The educational video scripts were developed by the marketing team with input from the research team. However, themes in these videos were not the focus of development at the outset. Therefore, 2 trained study staff independently qualitatively coded the educational videos using an inductive approach to identify themes that appeared in the scripts. A first pass of all scripts was made by the first author to create a set of possible themes. This was followed by the second pass by the second author. The 2 coders met and discussed themes, adding, removing, or combining themes as needed. Then, the 2 coders independently coded for the presence or absence of each theme in each script. Kappa values and percentage agreement were assessed to ensure >70% agreement before discrepancies were resolved through discussion. If values for a theme were below the 70% threshold, both coders met to discuss the theme in greater detail before recoding all scripts for that theme again. A total of 3 rounds of coding were completed to reach perfect agreement between coders. Biserial correlations were used to identify if any of our themes were related to our outcomes of interest using the 8 observations from the 8 videos in the campaign. Given the small sample size (N = 8 videos), probability values were not calculated for any correlations.

Results

Overall, our boosted video posts on Facebook reached a total of 83,937 users, which was an average of 10,492 users per month. In total, 26,181 users engaged with our boosted posts (an average of 3,272 users a month), and 22,777 had extended engagement with our boosted posts (an average of 2,847 users a month). As shown in Figure 1, data indicated similar reach, engagement, and extended engagement from August 2022 to December 2022, with an increase in these metrics in January 203 and February 2023 followed by a drop back down in March 2023.

Figure 1.

Reach, engagement, and extended engagement over time.

Overall, 3,327 users clicked from Facebook to our Web site (an average of 415 a month), with our Web site reporting 1,835 users (an average of 229 per month). Figure 2 depicts the number of clicks and Web site users. Fifty-five percent of the clicks recorded by Facebook resulted in users accessing the Web page. The highest number of clicks and Web site users happened in December 2022 and January 2023.

Figure 2.

Clicks and Web site users over time.

The total campaign cost paid to the social media marketing company was $25,950. This included costs for their time and effort creating, focus group testing, programming, and managing the content for the duration of the campaign as well as the boosting costs on Facebook. Boosting costs varied slightly for each month and content piece but totaled $1,740.31 of the campaign budget. On average, each video cost $193.56 (s = $16.60) to be boosted for 21 to 24 d each month, and the 2 image posts costs $199 combined to be boosted in the first month, each being boosted for 18 d. Assuming the 1,835 Web site users who may have downloaded a DA, with a total campaign cost of $25,950, the cost per Web site visitor who may have seen a DA was estimated to be $14.14 for this campaign.

Several of the outcomes of interest were highly correlated (see Table 2). Reach, engagement, and extended engagement were highly correlated (rs > 0.9), as were clicks and Web site users (r = 0.78).

Table 2.

Correlations between Outcomes of Interest

Outcome	Reach	Engagement	Extended Engagement	Clicks	Web Site
Reach	1.00
Engagement	0.95	1.00
Extended engagement	0.95	1.00	1.00
Clicks	0.08	−0.05	−0.06	1.00
Web site	0.25	0.12	0.11	0.78	1.00

Given the sample size (N = 8 videos), probability values were not calculated for the above correlations.

From the videos, 11 themes emerged. Mentions of a decision or choice needing to be made was the most common theme in the video, with 88% (7/8) of videos including this theme. Of the videos, 75% (6/8) discussed treatment options and 75% (6/8) discussed DAs. Sixty-two percent (5/8) discussed the need to be informed, and 62% (5/8) discussed outcomes or recovery. Fifty percent (4/8) discussed risks, and 50% (4/8) discussed values or preferences. Thirty-eight percent (3/8) discussed support. The remaining codes of pain, involvement of family or friends, and benefits of treatment each only occurred in 25% (2/8) of videos. For a breakdown of themes that appeared (light blue) or were absent (dark blue) in each month’s video, see Figure 3. For an example of 2 videos, their scripts, and their themes, see Table 3.

Figure 3.

Image indicating which themes were present or absent in each video.

Table 3.

Examples of 2 Video Scripts with Associated Themes

Month	Script	Themes
December	Patients considering a medical decision should consider using a decision aid to make their choice. Decision aids can help you: • Summarize the decision to be made • Provide information about options • Scheme of risks and benefits • Provide feedback on results • Consider your preferences • Clarify your personal values And remember, always talk to your doctor! A shared decision is a good decision. —Actor 4, female	Decision to be made, need to be informed, treatment options, risks, decision aid, values or preferences, benefits
February	Thinking of having hip or knee surgery due to arthritis pain? As you consider the options, try using a decision aid. It can help you understand your options, compare risks and benefits, and express your personal preferences and values to decide what’s best for you. Decision aids are available in many different languages. Talk to your doctor! A shared decision is a good decision. —Actor 3, female	Decision to be made, need to be informed, treatment options, risks, pain, decision aid, values or preferences, benefits

Table 4 shows the relationship between our themes and our outcomes of interest. We observed several interesting relationships. When videos included the theme of pain, they tended to receive greater reach and engagement. When videos included the theme of benefits, we tended to see greater users at our Web site. We also note some general patterns of interest. Mentions of a decision or choice to be made, outcomes or recovery, support, or family and friend involvement were all negatively related to all outcomes. Mentions of options, decision aids, being informed, risks, values, pain, hip or knee replacement, or benefits were all positively related to all outcomes.

Table 4.

Correlations between Themes and Outcomes of Interest

Theme	Reach	Engagement	Extended Engagement	Clicks	Web Site
Decision to be made	−0.32	−0.06	−0.05	−0.38	−0.24
Need to be informed	0.21	0.41	0.40	0.11	0.32
Treatment options	0.41	0.35	0.33	0.58	0.43
Risks	0.34	0.48	0.47	0.32	0.45
Outcomes or recovery	−0.25	−0.30	−0.29	−0.23	−0.55
Pain	0.88	0.70	0.70	0.20	0.36
Decision aids	0.36	0.34	0.32	0.53	0.24
Family or friend involvement	−0.01	−0.21	−0.21	−0.04	−0.08
Support	−0.15	−0.27	−0.26	−0.26	−0.22
Values/preferences	0.24	0.33	0.33	0.42	0.20
Benefits	0.43	0.43	0.42	0.51	0.78

Given the sample size (N = 8 videos), probability values were not calculated for the above correlations.

Discussion

This project created and operated a social media campaign intending to reach Spanish-speaking and Hispanic adults interested in hip and knee OA. Over 8 mo, we reached more than 83,000 users, more than 22,000 had extended engagement, and almost 2,000 users accessed the Web site with the DAs. Through analysis of themes in the videos, we found that videos were more likely to reach users and have users engage with the videos when they included the theme of pain. Further, users were more likely to visit our Web page with the DAs when videos included the theme of benefits people may expect from treatment.

Although prior work using social media to disseminate health information is limited, we can attempt to compare our rates of user interaction with other similar studies. One quasi-experimental campaign by Gough and colleagues ²⁴ used Twitter to disseminate messages about sun care and cancer prevention. The campaign ran over 5 mo and paid £10 a post; the study reported more than 417,000 tweet impressions (similar to our reach metric), more than 11,000 engagements (similar to our engagement metric), and more than 1,200 retweets (which require the user to take an action, so we can tentatively compare these to our clicks to the Web page). Another study by Ng et al. ²⁵ used Facebook over 3 mo to disseminate clinical messages to caregivers of patients with tracheostomies or laryngectomies. This campaign reported reaching more than 433,000 users, with more than 6,000 clicks on Facebook.

There are not published benchmarks or standards in the health literature to measure success with the social media campaign. It is notable that our campaign had significant engagement (more than 26,000 over 8 mo compared with their more than 11,000 over 5 mo) and interaction (more than 3,300 clicks over 8 mo compared with their more than 1,200 retweets over 5 mo). Further, the proportion of users reached by the post who clicked on it was about 0.4%.

Through our qualitative work, we found that videos discussing pain garnered more reach and engagement. This may be because these comments provided a compelling hook to our intended audience (e.g., “If you have constant hip pain . . .”). Further, videos relaying benefits garnered greater user interaction with our site. The study by Gough et al. ²⁴ also found that informational messages (as opposed to messages framed to shock or to detail personal stories) were most likely to encourage interacting with the platform. Future campaigns to disseminate DAs should incorporate informative messages that highlight the benefits of the treatment(s) or tests and target resources to boost those messages.

Although all DAs will not be appropriate to be delivered online, many common decisions such as hip and knee OA and cancer screening lend themselves to direct-to-patient delivery. Further, the cost per Web site visitor who may have seen the DAs online was somewhat modest (∼$14 per DA) in comparison with estimates from a clinic-based cost analysis study by Ho et al. ³³ regarding prostate cancer DAs ($38–$60 per DA). It stands to reason that relying solely on health care systems or clinicians to distribute DAs to patients has likely hampered reach and the use of DAs and may even be a more expensive delivery process. Active and focused marketing will be needed to connect patients with the tools that are available on the Internet, but this approach may be better able to reach underserved populations than traditional in-clinic DA delivery approaches. This study presents some of the first evidence of the kinds of messages that are effective in using social media to disseminate DAs to an underserved population.

It is worthwhile to note the benefits and drawbacks of hiring a social media marketing team. As can be seen above, most of the cost of the campaign was to pay for the marketing team’s time and effort on the project, with only about 7% of the budget allocated to the actual cost of boosting the videos on Facebook each month. Thus, a budget-friendly option is for social media–savvy researchers to take on the campaign themselves to reduce costs. With free apps such as Meta Business Suite, ³⁴ researchers could schedule the campaign materials to be posted and boosted on both Facebook and Instagram on their own (up to 20 d out), minimizing the time needed to be devoted to the daily management of the campaign. However, there were many areas where the social media marketing team assisted us that many researchers may have difficulty with. First, the marketing team was able to quickly understand the purpose of our project and identify the key demographic and geographic characteristics we would want to designate as our focus when boosting posts. As post boosting allows you to indicate your intended audience (for a slightly higher price), this meant we would pay a bit more but would ultimately ensure we were reaching those individuals we intended to. Second, they provided sage advice on what would keep users engaged with our page. For example, we were not aware that we needed to create 4 times as much “fun” content as “health-related” content to keep users engaged. They were able to quickly create and revise content such as our Monday memes and Wednesday Boston-area sports posts so that our page would remain relevant throughout the campaign. Third, they actively worked to follow/like pages that might increase our visibility. This included finding other Boston-area Latino groups with large followings each month that could bring in additional attention to our page. Fourth, the speed at which they were able to not only create content but also focus group the content with our intended audience was astonishing and would not have been possible at our institute given the requirements of the institutional review board. Finally, social media is a rapidly changing landscape where free products may change or disappear, and content may become inappropriate overnight. Thus, while we highly encourage others to attempt budget-friendly alternatives to hiring such marketing teams, we do see the benefit to outsourcing this effort to the professionals.

Our work contains some limitations. First, while we suspect that many of the individuals who took the time to visit the Web site also downloaded the DAs, we did not have data on downloads. Second, we used only 3 platforms (Facebook, Twitter, and Instagram), leaving out other platforms including TikTok and WhatsApp, which are popular with Hispanic individuals. These apps were considered when developing the project with the marketing consultants; however, limited resources prevented the use of more apps. Third, we boosted posts only on Facebook and thus are unable to identify any cross-platform differences in engagement or interactions with our materials. Fourth, we were not provided with any demographic information about whom our posts reached, which would have been useful in describing the sample of users to whom we were able to disseminate this material. Fifth, we were not able to track Web site visitors to understand how accessing the DA may have affected their health behaviors. Sixth, our sample used to review relationships between metrics and themes was quite small, and future work is needed to replicate the relationships found herein with a large enough sample size to calculate probability values.

Further work is needed to understand how we can best use social media to support patient empowerment toward improved health care outcomes. Work should aim to track use of such DA resources and survey individuals who engaged with social media for feedback and reactions. As social media platforms are rapidly growing and advancing, more work is needed to explore how individuals use and act on evidence-based health information provided by social media. Additional work is needed to identify innovative ways of surveying individuals or recording information regarding behavior change through social media platforms while keeping participant information private and secure. This will likely require identification of creative ways to measure and evaluate engagement with social media and the possible impact on large decision-making processes.

Conclusions

This study found that a social media campaign was able to reach and engage Spanish-speaking populations about hip and knee OA with messages around shared decision making, patient DAs, and patient engagement. Active outreach via social media may be an effective means of disseminating DAs directly to Hispanic individuals and communities.