Article Text

Download PDFPDF

Why don't end-of-life conversations go viral? A review of videos on YouTube
  1. Imogen A Mitchell1,2,
  2. Anne L R Schuster2,
  3. Thomas Lynch3,
  4. Katherine Clegg Smith4,
  5. John F P Bridges2 and
  6. Rebecca A Aslakson3
  1. 1The Commonwealth Fund, Boston, Massachusetts, USA
  2. 2Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, Massachusetts, USA
  3. 3Department of Anesthesiology and Critical Care Medicine and Palliative Care Program, Johns Hopkins Hospital, Baltimore, Massachusetts, USA
  4. 4Department of Health, Behavior and Society, Johns Hopkins Bloomberg School of Public Health, Baltimore, Massachusetts, USA
  1. Correspondence to Dr Rebecca A Aslakson, Department of Anaesthesiology and Critical Care Medicine and Palliative Care Programme, The Johns Hopkins Hospital, 600 North Wolfe Street, Meyer 297A, Baltimore, MD 21287-7295, USA; raslaks1{at}jhmi.edu

Abstract

Objective To identify videos on YouTube concerning advance care planning (ACP) and synthesise existing video content and style elements.

Methods Informed by stakeholder engagement, two researchers searched YouTube for ACP videos using predefined search terms and snowballing techniques. Videos identified were reviewed and deemed ineligible for analysis if they: targeted healthcare professionals; contained irrelevant content; focused on viewers under the age of 18; were longer than 7 min in duration; received fewer than 150 views; were in a language other than English; or were a duplicate version. For each video, two investigators independently extracted general information as well as video content and stylistic characteristics.

Results The YouTube search identified 23 100 videos with 213 retrieved for assessment and 42 meeting eligibility criteria. The majority of videos had been posted to YouTube since 2010 and produced by organisations in the USA (71%). Viewership ranged from 171 to 10 642. Most videos used a documentary style and featured healthcare providers (60%) rather than patients (19%) or families (45%). A minority of videos (29%) used upbeat or hopeful music. The videos frequently focused on completing legal medical documents (86%).

Conclusions None of the ACP videos on YouTube went viral and a relatively small number of them contained elements endorsed by stakeholders. In emphasising the completion of legal medical documents, videos may have failed to support more meaningful ACP. Further research is needed to understand the features of videos that will engage patients and the wider community with ACP and palliative and end-of-life care conversations.

  • Communication
  • Clinical decisions
  • Terminal care

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Advance care planning (ACP) offers the potential opportunity to improve palliative and end-of-life care, and has been reconceptualised as a series of steps that may equip patients and their surrogates for ‘in-the-moment’ decision-making. These steps include choosing a surrogate decision-maker, identifying and communicating patients’ values, and establishing the degree to which surrogate decision-makers’ have flexibility in making decisions.1 Historically, the ACP process has focused on individuals making and documenting decisions to guide future clinical care should they lose decision-making capacity.2 Despite national campaigns, such as those in Canada, the UK and the USA that aim to encourage better palliative and end-of-life care,3–5 only a minority of people document their advance treatment wishes.6 Studies suggest that even when these documents are completed, they often fail to inform better clinical decisions.7–9

As the steps of ACP ask patients to consider and make preference-based decisions, it is possible that patients could benefit from decision support interventions.10 ,11 A recent ACP systematic review identified formally developed and evaluated instrument-based ACP support tools (Aslakson RA, et al, 2015, unpublished data), including videos,12 ,13 workbooks14 and interactive websites.15 ,16 Another avenue of support includes facilitated conversations with a trained healthcare professional.17 ,18 While most initiatives have originated from within the healthcare arena, ACP extends beyond this boundary and recent work advocates for community-based initiatives to help advance ACP.7 ,19

One potential avenue for broader ACP engagement might be through social media platforms. These platforms are internet tools that encourage user participation and allow individuals to create, share, and exchange ideas and information.20 These platforms could serve as potential tools for the dissemination of health information as given evidence indicates that adults are increasingly depending on the internet for health and medical information.21–24

YouTube, a video-sharing website, is one such platform allowing individuals and non-healthcare organisations to upload, view and share videos. In 2014 (7 years after YouTube's 2007 creation), approximately 100 h of video are uploaded to YouTube every minute and more than one billion users visit the website every month.25 Given the volume of videos available on YouTube and its large-scale reach, it is possible that individuals may seek out ACP information from this website.26 ,27 Moreover, YouTube tracks the number of times a video is viewed or shared. These measures can indicate a video's broad appeal and influence, and a video may be considered ‘going viral’ if it is viewed more than five million times in a 3–7 days period.28

The aim of this study was to identify videos on YouTube concerning ACP, and synthesise video content and style elements to help inform the development of a video that could support ACP and in-the-moment decision-making for patients and families preparing for high-risk surgery. This study is relevant as it provides a collective snapshot of current approaches to promote ACP. Moreover, the study could serve as a potential methodological model for scientifically exploring content on social media platforms such as YouTube.

Methods

As part of an environmental scan to better understand ACP decision support tools,29 this study undertook a two-stage approach: (1) a 1-day stakeholder summit and (2) a review of related videos on YouTube.

Stakeholder summit

The summit goal was to engage stakeholders to identify key video elements for viewability and content relevant to ACP. Fourteen ACP stakeholders (three surgeons, two patient family advocates, two health services researchers, four palliative care clinicians and/or researchers, one medical film-maker, one patient-centered outcomes researcher and one patient safety and quality researcher) met to review a representative sample of seven ACP videos selected from published studies,30 websites31 or YouTube.4 Using design-thinking techniques,32 stakeholders identified important video elements to be: (1) patient-centered videos that feature personal stories delivered by ‘real’ patients and families; (2) a decision framework that emphasises and elicits ‘how people want to live’ rather than the treatments they want or do not want and (3) ‘upbeat’ music (where upbeat music was highly linked to increased engagement).

Review of YouTube

Eligibility criteria for selecting ACP videos

Criteria from a previously conducted ACP systematic review (Aslakson RA, et al, 2015, unpublished data), a recently published qualitative study,33 and stakeholder-identified characteristics seeded inclusion/exclusion eligibility criteria. This ACP review of YouTube aimed to identify decision support tools that could promote perioperative ACP. Exclusion criteria included videos that: targeted healthcare workers; discussed irrelevant content; focused on viewers under the age of 18; were longer than 7 min; had fewer than 150 views; were in a language other than English; or were a duplicate version.

Given that the list of YouTube results could change rapidly, the results needed to be obtained, viewed and downloaded in a relatively short period of time (ie, less than 24 h). To make this feasible while still ensuring a robust dataset, the authors chose to exclude videos that had fewer than 150 views.

Information source and search

On 22 January 2014, YouTube was searched concurrently for all unique, relevant videos using the following predefined list of terms: ‘Advance Care Directive’ OR ‘ACP’ OR ‘Advance Care Plan’ OR ‘Advance Health Care Planning’ OR ‘Advance Medical Planning’ OR ‘Advanced Care Planning’ OR ‘Advanced Care Plan’ OR ‘Advanced Care Plans’ OR ‘End of Life Care’ OR ‘End of Life Care Planning’ OR ‘End of Life Care Plan’ OR ‘End of Life Care Plans’ OR ‘End of Life Plans’ OR ‘End of Life Planning’. In order to reduce bias based on the influence of previously conducted YouTube searches, this search was performed using Duckduckgo (http://www.duckduckgo.com), a search engine that has previously allowed researchers to anonymously search the internet.34 Additional videos were identified through a modified snowballing technique, where researchers reviewed relevant videos suggested by YouTube that had not been identified through the original search terms.

Data collection process

Search results were sorted according to the number of viewings. Two research team members jointly reviewed videos to determine their eligibility for inclusion. If there were disagreements, a third researcher was engaged to resolve disagreements. Videos that met the eligibility criteria were downloaded from YouTube to ensure long-term accessibility and analysis. A structured video extraction tool was developed around the important video elements identified by the stakeholders. Using this tool, two trained research team members independently reviewed and extracted data for each video. Discrepancies were resolved through discussion and consensus.

Extracted information included video elements (ie, personal stories, decision framework and musical style) and general information such as name of the producer, video duration and country of origin. Other information included the number of times a video was viewed, shared on other social media platforms such as Facebook, and/or endorsed by a viewer where they either ‘liked’ or ‘disliked’ it. ACP-related content was extracted, including video content characteristics such as choosing a surrogate, clarifying values and beliefs, having conversations with surrogate/family/physician, and completing legal medical documents such as an advance directive. Additionally, researchers extracted information about film style and design (ie, documentary, animation, news story), setting (ie, healthcare or otherwise), stakeholder representation, demographics of individuals in the videos (ie, gender, perceived age group, and perceived racial or ethnic group), and use of pictures and/or graphics.

Statistical analysis

Multivariate linear regression analyses were conducted in an attempt to determine the association between individual factors (shares, duration of video, patient representation, healthcare provider representation, healthcare setting, inclusion of music) and number of video views. Researchers hypothesised that more shares, shorter videos and patient representation would be statistically significantly associated with more views. All reported p values were two-sided, with p<0.05 considered statistically significant. Data were analysed using STATA software, V.12 (StataCorp LP, College Station, Texas, USA).

Results

The YouTube search (figure 1) identified 23 100 videos and an additional 10 videos were identified through snowballing techniques. After sorting the videos by view count, 22 897 were excluded because they had fewer than 150 views. Of the remaining 213 videos, 42 fulfilled the eligibility criteria and were included in the final analysis.

Figure 1

Selection results from YouTube search of advance care planning videos.

General information

Table 1 provides descriptive details about the videos as well as whether or not each video addressed key characteristics that stakeholders identified as important. The majority of videos were posted during or after 2011 (32/42, 79%) and most were sourced from the USA (30/42, 71%), with none from the UK. A wide range of groups produced the videos, but the majority were produced by organisations or individuals affiliated with healthcare or healthcare research (30/42, 71%). The median number of views was 456 (IQR 313–816) with two videos having more than 10 000 views each. Half of the videos (21/42, 50%) had received likes (ranging from 1 to 126 likes) or dislikes (ranging from 1 to 2 dislikes). Eleven (26%) videos had been shared with the number of shares ranging from 1 to 27 times and a mean of eight shares.

Table 1

General information and characteristics about ACP videos on YouTube

Forty per cent (17/42) of the videos incorporated personal stories from patients and/or their families. About a quarter of the videos (12/42, 29%) used upbeat or hopeful music, while half of the videos did not include any music at all (23/42, 55%). Nine of the videos (21%) framed the conversation around how a person wants to live. While a number addressed two of these three characteristics, only two videos (2/42, 5%) included all three.

Video content

The majority of videos (36/42, 86%) addressed the completion of ACP documents such as advance directives, durable power of attorney for healthcare forms, and Medical Orders for Life Sustaining Treatment or Physician Orders for Life Sustaining Treatment forms. To complete such ACP documents, videos referred to and/or discussed associated supportive activities. For instance, most videos (34/42, 81%) emphasised having discussions with family, surrogates and physicians to make decisions about and/or inform them of their healthcare wishes. Furthermore, nearly two-thirds (27/42, 64%) of videos spoke of considering personal values and goals when determining advance treatment wishes, and many videos referred to the importance of nominating a surrogate in the event that the patient might become incapacitated (25/42, 60%).

Very few videos (2/42, 5%) presented numerical data as part of the content and the two that did used that numerical data to motivate the need for individuals to engage with ACP. Only two videos (5%) encouraged individuals to consider and specify how much leeway they would give to their surrogate decision-maker, especially in the context of facing unexpected medical decisions that seemed to be at odds with the individual's expressed advance treatment wishes, but would be consistent with the individual's values and goals.

Video style

Most frequently, videos used a narrative documentary style (27/42, 64%), with the predominant stakeholder represented as being a healthcare worker (25/42, 60%) in a hospital setting (20/42, 48%). All but one video included male and female individuals (41/42, 97%). The videos did not feature individuals from a diversity of racial or ethnic backgrounds, and the majority of videos only included people who were perceived to be white (27/42, 64%). Only six videos appeared to be targeted at a particular age group of which five focused on older individuals (5/42, 12%).

Patients were among the least represented stakeholder group (8/42, 19%), but family members did appear in 19 (45%) of the videos. The eight videos that included patients also included family members where they described their stories related to ACP decision-making and goal setting. Some videos that portrayed patients used a combination of one-on-one patient interviews, patient-family interviews, and pictures or images of families spending time together overlaid with the speaker's voice.

Two videos (5%) (Speak Up and Begin the Conversation) used animated text and graphics to convey content, which clearly differentiated them from the other videos. Furthermore, while Speak Up did not include patient stories, it did depict people ‘speaking up’ and having conversations by incorporating pictures of real people sitting together who appeared to be talking with one another.

Statistical analysis

Using multivariate regression analysis, factors such as duration of the video, inclusion of a patient, use of music, inclusion of a healthcare worker, and filmed in a healthcare setting had no statistically significant relationship with the number of views. Multivariate analysis did reveal that for each additional time a video was shared, the video was viewed an additional 347 times (p<0.001).

Discussion

Statement of principal findings

On the basis of stakeholder input as well as recommended steps for ACP, it is possible that none of the 42 videos identified here may encourage patients to initiate ACP or provide them with the necessary information or support to make in-the-moment decisions relating to palliative or end-of-life care.1 Most videos focused on the traditional approach to ACP by directing people to complete an advance directive or any other legal medical document. Given that national and international community campaigns exist to promote ACP,3–5 ,19 it is possible that ACP videos would garner more attention in terms of viewership if they were to be closer aligned to such campaigns. Only a relatively small number of videos contained either style elements (upbeat music, dynamic voices, personal stories delivered by real people as opposed to actors) or content elements (framed around living and quality of life specifically) as endorsed by our stakeholders (ie, ACP Conversations; Why Families Should Make an Advance Care Plan). These videos may provide valuable insights for developing new videos that generate community-wide interest in ACP.

Results in context

Recent work advocates for community engagement in ACP7 ,19 and YouTube videos could be one avenue to support such a goal. The results of our statistical analysis highlight the potential importance of generating shares in order to increase viewership. Our results did not identify content or style elements that increased views. From the literature, video style and content that are predominantly associated with virality include funny, cute, disgusting or emotionally arousing videos.35

These categories are challenging to adopt in the context of ACP and may help explain the low viewership, although it is worth noting that some videos (ie, Patients Get a New Perspective; Healthbeat—End-of-life Care) originally aired as news stories and may have been watched by thousands of individuals outside of YouTube without the researchers’ knowledge. In addition, ACP is closely related to conversations about palliative and end-of-life care, a topic that people often tend to avoid.7 As such, videos may need to use approaches that overcome viewer reluctance to engage with ACP. Perhaps stakeholders’ strong preference for upbeat music indicates an approach that can help allay such fears.

Although videos promoted the completion of advance directives or other legal documents, they frequently failed to meaningfully support a value-based ACP process and/or recommended ACP steps.1 For instance, very few videos presented viewers with questions that would help them identify and clarify their values. Instead, the videos often merely stated that people should consider their values and goals. Similarly, while videos frequently encouraged viewers to choose a surrogate decision-maker, they did not mention what factors to consider when selecting someone who would make an appropriate surrogate decision-maker and/or how much flexibility to grant surrogates in decision-making processes.

Furthermore, it is important to consider the emotional responses that individuals may have in relation to the videos. As the videos were most commonly set in a hospital environment and featured a healthcare worker, the viewer may interpret that these decisions are to be made in a serious medical context, which is often too late for palliative and end-of-life care conversations. Moreover, the prominence of soothing music, which can create specific emotions,36 may give the viewer a sense of concern and sadness that one might feel at the very end of someone's life rather than considering ACP as a part of everyday life.

A handful of videos (Speak Up!; ACP Conversations; Why Families Should Make an Advance Care Plan) exemplified desirable elements. In particular, style characteristics such as upbeat or hopeful music, the use of pictures and text, and patient or patient-family interviews potentially served to reinforce the video content and deliver it in viewer-friendly ways. These approaches may help convey content that (1) demonstrates how ongoing ACP conversations can generate a shared understanding about patients’ medical wishes; (2) reinforces how ACP can fit into everyday life and/or (3) describes how ACP can help prevent patient and family suffering. Furthermore, these videos also included content that potentially enables viewers to take further action. Promoting such outcomes could support the UK's37 and the USA's38 recently articulated guiding principles for the care of dying people.

Strengths and weaknesses of the study

This study converses with a recently published review26 and an editorial27 about the role of social media in discussing issues related to palliative and end-of-life care, and offers an empirical contribution to this emerging field of research. It is the first known methodical review of YouTube videos related to ACP. In drawing from a recent ‘systematic review’ of YouTube video reviews, we endeavoured to address some listed concerns.34 We attempted to be transparent about our video selection process and the need to disclose the manner in which we coded videos. Using Duckduckgo as a search engine mitigated potential bias from previously conducted YouTube searches. While no formal methodology currently exists for systematically reviewing YouTube content, this study demonstrates an approach to overcoming issues unique to conducting such research and could potentially be utilised as a model for future searches.

There are limitations to this study. First, relevant videos may not have been captured in the search process (ie, videos outside of YouTube). As an example, no videos developed in the UK were recovered despite the existence of national ACP campaigns.22 Second, eligibility criteria may have limited the generalisability of study findings. The third limitation pertains to the dynamic nature of YouTube, which compared to more traditional databases, such as PubMed, uses an algorithm to relevance-rank search results which changes over time as existing videos may be deleted and/or new videos may be added. Fourth, consistently categorising subjective characteristics (ie, category of music) was at times difficult. Lastly, excluding videos with fewer than 150 views yielded a total number of videos that could be downloaded in one specific day, but it may have excluded potentially relevant videos and, in particular, recently posted ones.

Results of the study: possible mechanisms and implications for clinicians or policymakers

YouTube may provide a forum for sharing health information within and outside a healthcare context. A 2006 survey of patients with cancer in the USA suggests that the internet can inform and empower patients in relation to medical decision-making.39 From our study findings, ACP-related videos found on social media platforms, such as YouTube, could prove to be an innovative and valuable way of promoting better ACP among the general public. Ideally, these videos would empower patients, families, surrogate decision-makers, and clinicians to be more involved in meaningful ACP, and better prepared for palliative and end-of-life care decision-making.

As no previous studies have formally evaluated the effectiveness of these videos to support ACP, clinicians may be wary of using them as a resource. The unregulated nature of YouTube and the possibility of videos misinforming patients may cause clinicians and policymakers concern about referring to the site as a potential information source.40 Yet research demonstrates that the public frequently looks to the internet for healthcare information. Perhaps the more practical response is to ensure that the content found there is both research-driven and engaging.

Unanswered questions and future research

Although YouTube is used by people around the world and serves as a potential mechanism for global engagement, ACP videos on the website have yet to be widely embraced. If there is to be a significant change in general public understanding of palliative and end-of-life care, social media could play an important part. In addition, it is unknown whether videos available through social media can both engage communities in ACP and provide sufficient information and guidance to support deliberate palliative and end-of-life care decision-making. Future research could evaluate the value of strongly emotive ACP videos as well as the role that well-known organisations play in establishing viewers’ trust in videos or directing people to high-quality videos.

Conclusion

ACP and palliative and end-of-life care conversations on YouTube have not gone viral and few videos on the website contained the content and style elements endorsed by our stakeholders; inclusion of these elements could potentially improve efficacy, appeal and community interest. Existing YouTube videos about ACP tend to promote the product of ACP, an advance directive, rather than informing about the process of palliative and end-of-life care decision-making. To inform the development of a new ACP video support tool suitable for clinical use and meaningful for patients and family members, further research involving detailed stakeholder engagement is needed in an attempt to understand the features of videos that are most salient to them. In doing so, it may be possible to develop new videos that not only contain the appropriate ACP content, but also engage a larger proportion of the population.

References

View Abstract

Footnotes

  • Twitter Follow Imogen Mitchell at @IA_Mitchell

  • Contributors RAA is the guarantor for the overall content of this paper. JFPB, KCS and RAA conceived the study. JFPB, IAM, ALRS, TL and RAA planned the conduct of this study. IAM, ALRS and JFPB conducted the work and have drafted the manuscript. All authors have reviewed it critically, edited content, and have given approval of the final version to be submitted. ALRS submitted the manuscript.

  • Competing interests This work was partially supported through a Patient-Centered Outcomes Research Institute (PCORI) Communication and Dissemination Award (CD-12-11-4362). A Harkness Fellowship (2013–2014) from The Commonwealth Fund supported IM.

  • Provenance and peer review Not commissioned; externally peer reviewed.