Patient–provider communication over social media: perspectives of adolescents with psychiatric illness

Abstract

Background

Social media is an increasingly dominant platform for communication, especially among adolescents. Statements from professional bodies and a growing body of empirical evidence support a role for social media in improving provider–patient interactions. In psychiatry, particular concerns exist about the suitability of this style of communication. Very limited data are available exploring how patients would like to incorporate social media into their communication with their psychiatric providers.

Methods

We conducted a qualitative study with 20 adolescents attending the Yale Psychiatric Hospital Intensive Outpatient Programme. Interviews were analysed using inductive thematic analysis.

Results

Participants highlighted how social media could allow for constant access to a mental health provider, provide a less anxiety‐provoking mode of communication, and allow for them to be monitored in a more on‐going fashion. However, participants also identified many potential risks associated with these applications, including the potential for anxiety if a provider was not able to respond immediately, and a sense that online interactions would be less rich overall.

Discussion

Our findings suggest that adolescents are open to the idea of communicating with mental health providers over social media and are able to describe a number of instances where this could be of value. The risks participants described, as well as concerns raised by existing literature, indicate the need for further work and protocol development in order for social media to be a feasible tool for communication between providers and adolescents with psychiatric illness.

Introduction

It is indisputable that the use of social media is here to stay. As of 2010, 73% of American teenagers and 47% of adults with Internet access used social networking sites, and the numbers are steadily rising.1 Of even greater interest is the amount of time spent communicating via social networking and blogging: 110 billion minutes per day or 22% of Internet usage as of 2012.2 This explosive shift in communication trends has presented the health‐care sector with many challenges and opportunities.3 The use of technology such as telemedicine has a long and encouraging history, with psychiatry being particularly amenable to such innovations.4 Furthermore, it has been established that many young people rely on the Internet as an important source of health‐care information and that this serves to better inform a group that has relative difficulty in establishing relationships with health‐care professionals.5

The use of the Internet may serve as a way to gather information and avoid stigma associated with severe mental illness.6 In addition to simply gathering information, patients may also seek to develop skills for coping with illness, seek emotional support and share and explore personal illness experiences.7 A report by Patel et al.8 highlights the barriers to conventional mental health access in youth and the potential role of population‐based interventions that may be disseminated online. In support of this notion, a study by Ben‐Zeev and colleagues reported that 81% of patients with severe mental illness were interested in mental health services via their mobile devices,9 and of further interest is the work of Bai et al.,10 who report that patients participating in their online ‘virtual psychiatric clinic' where more likely to be younger, first‐time users of mental health services. Such findings underscore the importance of exploring the potential for online tools to reach a sector of the population that is traditionally less likely to access mental health services.11

A number of existing efforts at engaging patients using online media have been described. Online‐moderated cognitive therapy groups have proven efficacy for reducing risk factors for eating disorders.12 A large study published in 1997 highlights not only the efficacy of online group therapy, but the potential for participants to be more comfortable with sensitive disclosures.13 These and other studies are reviewed by Ybarra and Eaton, who conclude that online treatment is likely to become an increasingly prominent part of mental health practice, which they conceptualize as being overlapping and complimentary to traditional, face‐to‐face interactions.11 Pointedly, they highlight the challenge this shift is likely to present to clinicians.

While the number of physicians responding to this demand has increased, there is still a substantial disconnect between what patients appear to want and what health‐care providers are willing or able to provide.14 The Institute of Medicine (IOM) has emphasized the importance of meeting patients ‘where they are’ and when they need it and that incorporating effective electronic communication strategies can ‘facilitate a transformation of mental health services in the years ahead’.9, 14 However, a number of concerns need to be addressed before this promise can be realized. Regarding psychiatry in particular, communication using social networking challenges established norms of ethical and professional practice. These include potential risks to confidentiality and boundary violations. Young psychiatrists, residents and medical students are particularly at risk for such adverse outcomes, being more likely to have an established online presence, and having greater comfort with using social networks in general.15 Awareness of these risks has prompted recommendations for the cautious use of social media by health professionals.

What do patients think? Although a number of sources cited above concur that the majority of patients are in favour of interacting with health‐care providers electronically, the extent and manner in which patients may wish to engage is less understood. Furthermore, interaction through social media has been comparatively less studied and is made more complex owing to the variety of communicative tools these platforms afford. In contrast to email or texting that typically take the form of one‐on‐one messaging, social media allows for a diverse set of interactions where patients may broadcast messages to a large network of friends, address individuals publically or privately and monitor messages made by other members of their online network.16 As social media has grown in its use, these types of interactions have come to characterize a predominant means by which adolescents communicate with peers and family. This raises the question of whether adolescents may hope to be able to approach communication with their health‐care providers, and specifically mental health providers, in a similar way.

In this study, we aimed to build on the existing understanding that patients value the potential for electronic communication with mental health providers and describe in detail the ways and extent to which adolescents may consider social media to be of value for this purpose. In addition, we sought to explore the particular concerns patients might have about this style of communication. To this end, we employed qualitative methodology – a decision informed by our sense that this was a complex and as of yet poorly understood issue, with participants likely holding nuanced views about their communicative preferences in different clinical situations. The goal of our study was to present a preliminary description of the preferences of adolescents with psychiatric illness regarding communication with their mental health providers and contextualize these preferences within existing thought on how to approach this new and significant challenge.

Methods

We obtained human investigations committee (HIC) approval to conduct 20 semi‐structured interviews with adolescent patients attending an intensive outpatient programme (IOP) at Yale New Haven Psychiatric Hospital. Informed consent was obtained from participants and their parents. Participants were recruited using a convenience sample over a 1‐month period, where all IOP participants were approached and offered the chance to participate. Twenty‐one individuals were approached, and only one declined participation. The process of recruitment was that participants were informed about the study and given a one page, HIC approved information sheet to take home to read and share with their parents. Participants were requested to follow‐up with the study psychiatrist once they had decided whether or not they were able to participate. Participants’ willingness to participate is likely attributable to them having periods of open time during the course of their IOP attendance, the relatively non‐invasive nature of the study, and the fact that participants were given $25 gift cards to thank them for their participation.

Interviews were open ended, but centred around questions regarding the participants’ experience of social media and how they felt about communicating with health‐care providers using social media (see Table 1). Participants ranged in age from 14 to 18 (see Table 2 for further demographic information). Interviews were conducted by the authors PM, KK and GVS and took place in a private office with only the interviewer and participant present. Interviews were transcribed and analysed using inductive thematic analysis, allowing themes to emerge organically from the data.17 This allowed us to give voice to the participants’ perspectives without imposing existing theoretical constructs. The process of preliminary data analysis began concurrently with the interview process. To this end, the authors met at several stages during the interview process to discuss field notes kept during and after interviews and impressions of emerging themes. Furthermore, transcripts were reviewed to gain an early sense of prominent themes. After 16 interviews had been conducted, it appeared that no new themes of prominence were emerging – this was determined collaboratively by the co‐authors, who reviewed the transcripts and results of preliminary data analysis. A further four interviews were conducted to maximize the probability that no themes of significance were being missed. Subsequently, the accumulated transcripts were analysed inductively by the co‐authors GVS and SVR. GVS conducted the primary analysis, and SVR reviewed the codes, comparing them against the raw transcripts to ensure inter‐rater reliability. The codes were subsequently again reviewed by all the co‐authors, and decisions were made on the establishment of hierarchies and identification of prominent relationships between themes. In this manuscript, we report on themes relevant to the issue of patients’ preferences regarding communicating with providers over social media. Subsequently, we contextualize these findings within existing descriptions in the literature and highlight ways in which are study confirms and extends extant theory.

Table 1.

Sample questions

Sample questions In what ways have you engaged with a therapist or counselor using social networks? Or your family? What was this like? If not at all, or is this something you would value? ○ Potential prompts: Quality of the interaction? In what ways was this/could this be helpful? ○ Potential prompts: To chat? Schedule things? Read your profile? In what ways could this be an issue for you? ○ Potential prompts: Privacy?

Table 2.

Demographics

	Age	Gender	Race	DX
1	16	M	Caucasian	ADHD, mood NOS
2	19	F	Caucasian	MDD
3	16	M	Caucasian	MDD, Anxiety disorder NOS
4	15	F	Caucasian	MDD, PTSD
5	15	F	Caucasian	Psychosis NOS, PTSD
6	18	F	Caucasian	MDD
7	15	F	Hispanic	PTSD
8	14	M	Caucasian	MDD, Anxiety disorder NOS
9	14	F	Caucasian	Mood NOS
10	17	F	Mixed ethnicity	Mood NOS, Learning disability
11	14	F	Caucasian	Mood NOS, ODD
12	16	F	Caucasian	MDD
13	17	M	Caucasian	MDD
14	17	F	Caucasian	PTSD, Mood NOS
15	18	F	Caucasian	MDD, Anxiety
16	16	F	Caucasian	MDD
17	17	M	Hispanic	MDD
18	16	F	Hispanic	MDD
19	16	F	Caucasian	MDD, Anxiety disorder NOS
20	16	F	Caucasian	MDD

Results

As described above, participants were asked to reflect on their experiences with social media, the ways in which they thought it could be of value in communicating with mental health professionals, as well as potential disadvantages. In the course of analysis, it became clear that participants identified three distinct ways in which they thought communicating with their mental health provider using social media could be of value. In addition, participants identified clear areas of risk associated with each theme. These are summarized in Table 3.

Table 3.

Summary of themes that emerged from this study

Summary of themes Constant access: ○ Can get quick feedback from mental health provider ○ Access support outside of scheduled appointment times ○ Value of positive presence on social media ○ However, ○ Potential for anxiety if response time too slow Easier to open up: ○ Less embarrassing than face‐to‐face interactions ○ Resembles style of sensitive communication with family ○ However, ○ Non‐verbal communication lost ○ Easier to hide feelings Monitoring: ○ Provider may be aware if safety concerns emerge ○ Provider may gain longitudinal sense of how things are going ○ However, ○ Content may be misinterpreted

Constant access

A significant advantage cited was the potential for having open‐ended access to providers, with questions being responded to quickly:

I actually think it would be good [communicating with provider on Facebook]… Because not only would I be able to talk to them, um, when I, I came to see them, or like when I was scheduled to, but then when I needed them outside, they were also there. [Participant 10 – 17 year old female, Mood Disorder not otherwise specified (NOS)]

The idea of not having to stick to scheduled appointments was emphasized by a number of participants and was linked to the idea that providers were most helpful during periods of acute distress:

Well the advantage of doing it I think would be like, the support's there when I need it, not only when I'm scheduled for it. [Participant 10 – 17 year old female, Mood Disorder NOS]

Another participant who also described having many stressful interactions with peers over Facebook felt that there would be considerable value in having a supportive person they could communicate with:

I think so, yeah, because like Facebook is so common, with teenagers and, so I feel like, if there was a positive like influence, and positive person, that they can go to, because they're always on Facebook anyway, they can always just, you know, talk to that one person. They can like check up on them, and stuff like that. [Participant 12 – 16 year old female, Major Depressive Disorder (MDD)]

At the same time, participants described how it would make them feel anxious if they initiated communication with a provider and failed to receive a response within a short period of time:

Participant 14 [17 year old female, Post‐traumatic Stress Disorder (PTSD)_ and Mood Disorder NOS]: I mean, what if they're not there to answer you, and you have to wait a day or so, to see if they reply. I mean, if you need an answer now, you can call them up, and talk to them, but if you Facebook message them, they won't, they probably don't even have a Facebook, you don't know.

Interviewer: If you send someone a Facebook message, do you expect, like how long do you think is reasonable to get a reply?

Participant 14: Five, six minutes.

Interviewer: Right, so if you're messaging a therapist, and their sort of pattern is, on Facebook, and they get back to you in a day, you think that would be too long a delay?

Subject: Um hmm, because what if I need help right then?

Easier to open up

Participants described a sense in which it could be easier to open up to providers about their lives over social media:

Um, well, some of the advantages I think are like me being able to um, tell them without like being, necessarily like, embarrassed to tell them in person what's actually going on in my life. [Participant 10 – 17 year old female, Mood Disorder not otherwise specified (NOS)]

This idea was part of a larger sense that it was easier to discuss sensitive issues over social media, even when speaking to friends or family:

Its actually helped me to talk to people and open up to people, because I've been able to like tell my sister what's been going on with me, through like Facebook, opposed to having to tell her face to face where I'd be really overwhelmed and anxious and I wouldn't be able to. [Participant 20 – 16 year old female, MDD]

The above was one of a number of examples where participants normalized the idea of communicating with therapists using social media by pointing out that this was the way they chose to communicate with many of their family members. Concurrently, a number of participants highlighted how in the case of therapy, communication over social media could be a less rich experience and potentially lead to misunderstandings:

Um, I might not get the same level of attention and you know, kind of therapeutic qualities that I would if I was in a room with a therapist, and it's not like personal, you know, you know what I mean, because you're not right there with them, talking about it, you're on a keyboard talking about it, so. And a therapist can't really read you, or your texts or how you're feeling so, the way, like I could respond, and, like another hard thing is like I'm a sarcastic person, so that's, that's a hard thing to figure out when you're typing, because people don't know what you mean sometimes, it's kinda like, do you mean that literally or sarcastically? [Participant 11–14 year old female, MDD and oppositional defiant disorder (ODD)]

In a similar vein, another participant described how they are not able to hide their feelings in face‐to‐face encounters:

Um, I don't know, maybe, because like, being face to face with them, I feel like I'm more inclined to tell them the truth, whereas, I could hide my um, emotions and feelings more over social media, because like I know when I talk to my psychiatrist, she's always like, what was that face? Like what did you make that face for? And then I have to tell her, so, I think it could be problematic [Participant 18–16 year old female, MDD]

Monitoring

Beyond using social media for communication, some participants also highlighted how it could potentially be helpful for providers to monitor their ‘Facebook status’ and message them if necessary:

If there was like a questionable status, like, it made them feel like I was upset, they would like message me, I guess, and say you know, ‘what's going on’ like ‘the status worried me, is everything okay? [Participant 12–16 year old female, MDD]

And,

I think that would be smart. But I think you would have to have like the patient's like approval to do that, obviously, but I wouldn't mind it, if I was truly trying to get better, because sometimes when you're like in the midst of like, I don't know, like a crisis or just like not feeling your best, like you do stuff on social media that you know, can come across as like a red flag, and maybe a therapist would like see that and use it to like, for the better of the patient [Participant 18–16 year old female, MDD]

Another advantage of being monitored would be that this would give providers a sense of how things had been going that could reduce the need for the patient to explain things during therapy:

I guess it'd be positive because instead of having to go to therapy and explain what was going on, I mean they would have read it and have some sort of understanding. [Participant 1 –16 year old male, ADHD and mood disorder NOS]

Other participants felt that their providers would be at risk of misinterpreting their posts and understand jokes between friends as actual threats for self‐harm:

Yeah, I mean, there's inside jokes between me and my friends, and if he or she didn't know about it, she [provider] might take that the wrong way… I don't know how they [providers] would put it – as unsafe, or between me and my friends as a joke. And I wouldn't know how they would take it. [Participant 14–17 year old female, PTSD and mood disorder NOS]

Similarly,

Like sometimes, I post, like on Instagram, like a picture, like of like a poem, and it's like numb, and it talks about like self‐harm and stuff like that. Sometimes I post stuff like that, and if my mom or my therapist saw that, like they would freak out. [Participant 5 – 15 year old female, psychosis NOS and PTSD]

Some participants did state that it would be ‘weird’ or ‘strange’ for their provider to monitor their social media. In addition, some participants noted that their parents would monitor their social media and that this could lead to potential misunderstandings if their parents were not aware of the context of their posts. Despite extensive probing on this issue, participants did not describe privacy as being a significant concern.

Discussion

Summary of findings

Overall, despite highlighting some potential downsides, participants in this study were in favour of communicating with providers over social media, highlighting a number of potential advantages. Specifically, participants emphasized the properties of continuous access and prompt response, less potential for embarrassment than in face‐to‐face communication and on‐going monitoring. With regards to continuous access, participants were drawn to the idea of being able to reach out to providers during times of particular stress, rather than simply at scheduled appointment times. Participants also highlighted how such an arrangement would likely be contingent on the clinician being able to respond quickly as they might need help ‘right then’. When discussing how it was easier to open up when communicating online, participants pointed out how it was less anxiety provoking than sharing sensitive information face to face, but that this also meant that there was the potential for hiding important emotional content from the therapist. Finally, participants described the potential value of their posts being monitored such that a provider could reach out in a time of need or crisis, but that there was a risk that providers would misinterpret the information being posted.

Findings in context

Potential value of new information sources

In many instances, participants pointed out how this style of communication already characterized their interactions with family members. This points towards the hypothesis that adolescents wish to communicate with their health‐care providers (and in particular psychiatrists) in a similar way to how they communicate with friends and family – an idea consistent with some existing empirical data regarding the preferences of patients with asthma.18 What might be the consequences of trying to meet this expectation? A benefit is the potential for the mental health provider to access additional information that could facilitate treatment. For many chronic illnesses, providers are able to rely on certain laboratory values to assess the overall progress and management of a disease – such as measuring glycosylated haemoglobin in someone with diabetes. In mental health treatment, providers are generally restricted to cross‐sectional assessments for collecting objective data about a patient's mental state. Having access to multiple, small samples of the patient's thoughts and feelings on a more continuous basis could provide a more accurate account of how the patient is doing overall. However, there is also the possibility that overzealous patients could overwhelm the provider with details not pertinent to the treatment, especially given the lack of time constraints imposed by such asynchronous communication.

The risk of rapid response expectations

In a similar vein, we found that patients had high expectations in terms of response time to messages and expressed concerns that they might not be able to get help ‘right then’. This is a concern of particular salience in psychiatric treatment, where patients may wish to indicate thoughts of imminent desire for self‐harm or even suicide. This discrepancy between patient expectations and what is feasible for the provider has been reported previously for other forms of electronic communication, and a number of guidelines have been proposed to protect both patient and provider in the event of an emergency.14 Nevertheless, this discrepancy should be considered alongside the opportunity of extending interactions beyond the limited temporal confines of a traditional consultation. The fact that such unscheduled, short interactions were viewed favourably suggests that they could be a valuable adjunct to longer, structured sessions. Additionally, it emerged that simply knowing someone supportive was ‘also there’ would reduce the stress associated with the use of social media.

Electronic communication facilitates disclosure

A further theme that emerged from the data was the greater perceived willingness of patients to ‘open up’ via social media and reveal aspects that would normally be regarded as too embarrassing in face‐to‐face encounters. This reflects the results of a large linguistic analysis of UK‐based adolescent email database: electronic communication allowed the patients to disclose information they would otherwise find embarrassing.19 This is also in line with qualitative data in a study by Andreassen and colleagues, in which patients had a lower threshold for contacting their doctor electronically.20 Importantly, the patients in that study emphasized the need to already have a trusting relationship in place before engaging in electronic communication.20 Our study suggests that this effect is not limited to emails but may also apply to social media posting.

The challenges of ‘Monitoring’

Participants cited the ability to post a ‘status’ online as advantageous. Rather than messaging a provider directly, this is a way of broadcasting a short, open‐ended description of how they are doing. These can be responded to by peers and used by the clinician to monitor the patient periodically and remotely. Of note, it was suggested that such monitoring could increase the efficiency of scheduled, face‐to‐face visits. The provider would already have a sense of how the patient was doing and could focus on specific issues that had been posted online. Two related, negative themes that emerged involved interpretation of posts. The first was that the nuances of body language, facial expression, tone and similar would be lost in translation, decreasing the richness of the clinical interaction. The second theme, which is especially applicable to social media, is that posts often have a specific intended audience or may be intentionally ironic or poetic. As such, situations could arise where the provider misinterprets, for example a sarcastic but innocuous ‘tweet’ as a cry for help. These findings are of particular interest given recent work drawing attention to the potential for monitoring of Facebook posts related to suicide. Ruder et al.21 highlight how suicide notes posted on Facebook have lead to timeous intervention, and therefore may have prevented suicides in the past. Another advantage was described by Ahuja et al.,22 who report on a case where a patient's Facebook post history was used to help construct a timeline after a failed, impulsive suicide attempt – a process ultimately considered to be relevant to his treatment. To our knowledge, very little has been written about the challenges of health‐care providers being able to accurately interpret Facebook posts, and this finding in our study warrants further investigation as it may highlight an important limitation to the use of monitoring by mental health professionals.

Strengths and limitations

Our work presents an important initial exploration of the ways in which adolescents with psychiatric illness may consider social media to be a helpful platform for communicating with their mental health providers. Social media is a complex and multifaceted communication platform, and a particular contribution of our study is that we report on some of the less explored aspects of online communication – such as public posts – and their potential role in mental health treatment. Nevertheless, our study possesses a number of limitations. Firstly, as a relatively small study, these results should be considered preliminary in nature, and further work is required to confirm our findings. Secondly, our analysis was complicated by the sheer number of social media platforms used by our patients, and further work could be helpful in delineating whether participants have different preferences for different services.

Conclusion

In our study, participants drew attention to the advantages of communicating with mental health providers using social media with regards to having constant access, being more comfortable opening up and the potential for on‐going monitoring. These findings correspond with existing ideas in the literature. Our study suggests that these ideas are considered to be salient by health‐care consumers. But significantly, we also report on a number of concerns patients had related to each of these ideas. Specifically, it is important to recognize that with the potential for constant access, patients may expect a ‘rapid response’ that may be beyond the capacity of clinicians to facilitate. When communicating online, providers need to be aware that although patients may appear to be opening up more easily, significant non‐verbal information may be lost. Finally, although monitoring appears to be a way to facilitate rapid intervention during times of crisis, the risk of misinterpretation may limit the ability of mental health providers to make accurate assessments of posted content.

Conflict of interest

The authors have no conflict to report.