New ways forward in youth mental health

Children, youth and families

How is technology being used to do more than just keep young people safe in the “right now”, but keep them safe into the future?

Mark Davis, Technology Director at Today talked to Dr Steve Leicester, Head of Direct Clinical Services at headspace about the healthy risk-taking happening in the ever-evolving world of treatment and challenging common assumptions on who accesses the services headspace provide.


1 Sep 2019

In conversation

Dr Steve Leicester, Head of Direct Clinical Services at headspace

Mark Davis, Technology Director at Today

Tomorrow 7 Headspace1

Can we start with a little bit about your background and role at headspace?

Well, the only thing I was formally trained in was to become a psychologist. I was trained as a teacher, but I was really bad at it. Along the way, I also did a couple of months of a mechanic's apprenticeship - wasn’t so good at that one either.

So, I'm a psychologist and have been involved in youth mental health specifically, for over 20 years.

I work at headspace which is a large national commonwealth funded programme for young people from age 12 to 25. When considering mental health people primarily recognise headspace for its 100 plus headspace centres dotted around the country, but there is more to it than that.

The area I'm involved in is what's referred to as eheadspace, which is fundamentally teleweb. So, telephone and digital support for young people across the country. That comes in the form of web chat, so synchronous one-on-one sessions with a clinician, but can also be done over the phone or via email. There are a lot of people still really like email!

So my role is to oversee eheadspace, which is a real privilege.

And what is the scale and scope of headspace?

We're all centrally located and we're all clinicians - credentialed clinicians. Whilst I think volunteering is fantastic, we're not volunteers. We're a group of practitioners.

We've got social workers, occupational therapists, psychologists and mental health nurses and our service operates daily from 9am to 1am, all year round, with about 80 clinicians. Everyone's part-time. With shift work of this nature, which is high volume and involves very in-depth interactions, I wouldn't recommend someone working at this full time. I think a blend with face-to-face work is quite important, healthy and hopefully fun too.

In terms of the number of young people that would come through, across the past year, around 36,000 new people have created accounts seeking our services. Like any health service, we find that hard going, keeping up and making sure people aren't waiting too long.

eheadspace has been running for about six years now, so, I’d say, much of our service offering is precisely designed and tailored to meet client needs. In reality, however, there's a lot of shit we also make up as we go. There's not quite a roadmap on how to get things done while finding the balance between high quality and high volume. So, we’re always looking at how this tension, which will never go away, can interplay. Currently, we still have more questions than answers.

Can you elaborate on the role that technology plays?

Well, the program that we have today could not have existed 20-25 years ago. It's a really interesting one in that technology is the medium, but the relationship is still the guts of what we do.

Even in areas which we’d term ‘group chat’, which resemble a national Q&A where anyone can jump into the discourse, we'll be providing online resources as well.

These Q&A style forums go for an hour, hour and a half on set topics. I think the one we ran last night was for family and friends under the broad title of ‘blended families’.

Tomorrow 7 Headspace2
Large image for null

You mentioned before the idea of email still being popular. What do you think is behind this? Accessibility?

When I came into this role I was surprised myself. I asked, "Really? What? Why are we still using email?" Then I started getting some insight into the nature of these emails. Very often email is used as the first point of contact, and rather than expressing yourself in a web chat or forum environment, people are banging out their thoughts in an email.

A lot of what I read in emails indicates that a person has taken time to really gather their thoughts and paint a phenomenal picture of their current experience. It's interesting that in the fast-paced environment of webchats, queues and more people wanting service, you might think, "Oh, let's get to the emails." You're not going to fire back a rapid response back to someone after they’ve taken you through a two and a half thousand word–really detailed explanation of their background. What they're looking for, the hope that they’ve conveyed, you want to give them the consideration and respect they deserve. In an email format, a person can come across as more reflective.

That said, I don't think that we should assume that people are not reflective, even if some of their chosen outlets fail to demonstrate this. Mind you, the last thing the world needs is another freaking mindfulness app!

Many young people still do much of this reflection, which we’re referring to, in the form of diaries. A sort of reflective practice which allows them to really consider things on a deeper level.

They may decide not to post their reflections online, they may decide not to post them anywhere. It might just be something they sit with, that they scrunch up and throw in the bin. But it emphasises this concept of reflection. Of trying to understand what is going for an individual and the world around them. I hope that we never dismiss this style of reflection with the assumption that it bares no potential. I think if you scratch the surface, you'll be surprised how prevalent it is.

The idea of reflection in this fast-paced, social media age is probably runs against the current discourse a little. What are your thoughts on this?

Well, I think a really good example to counter that sentiment can be seen in alcohol and tobacco consumption. The rate of tobacco consumption in young people is decreasing. The age young people start drinking, and drinking regularly, is rising. People are thinking. Don't think that they're not.

I think people are just grabbing at catchphrases. Let's not confuse the mediums that are put in front of us with people’s level of consideration and thought. Say you're asked, "Can you please use Slack instead of email for the next week?" or an alternative new trend. Because this new medium is put in front of you, you end up feeling compelled to use it, whether you like it or not. My point is–don't judge the person by the medium they choose or feel compelled to use.

Having so many modern tools can make for a very noisy life. How do you use one-on-one tools at eheadspace to cut through the background noise? Earlier you mentioned ‘teleweb’?

Yes. Teleweb is a term used for a whole selection of tools. Fundamentally, it's telephone and other online interactions which constitute what we might think of as online forums. Webchat is a one-to-one, real-time interface with another person. It's a live chat through a web browser.

I type and someone sees it instantly and responds. I can also attach appropriate links and so forth. Often, certainly in our domain, I think around 86% of young people are coming to eheadspace webchat as their primary medium to communicate with us.

We also respond to the family and friends (FaFS) of young people using our service. Here, by comparison, we find over 90% opt to communicate via phone. It’s an interesting split.

I think there are a couple of elements at play here. For many young people, a text-based interface provides a medium they’re very attuned to so they’re more comfortable with it. Another significant factor, which I think is really important, is the potential for anonymity.

Nowadays we ask for details but, like with any teleweb service, you can go there anonymously. This means you, the user, can select the tone of how you present yourself. Being anonymous means I can come and go as it suits me. This is a little different from the traditional face-to-face services where you’re required to give your name, Medicare card and so forth.

I suppose the way I think about it is that it's much harder to walk out of a session with a face-to-face practitioner if you don't like them than it is to exit your browser.

So, I think it's more than just convenience, it's anonymity. Anonymity creates a different dynamic, a more equal dynamic between the practitioner and the help seeker. Obviously, it also removes a significant barrier to entry.

Absolutely. And would you say you promote this anonymity? Anonymity in a teleweb service?

No, we don't. That is, it’s not something that we push, but we do highlight the accessibility it facilitates for users. Anonymity is part of having as few barriers-to-entry as possible. We're certainly not hiding it, but it's not the characterising feature of eheadspace.

It's also worth considering that many people who contact our service will engage through multiple mediums. They might make their first contact via email, where we'll respond and suggest that they make an account so that we can interface via webchat. Then, if they don't turn up for a scheduled appointment - scheduling appointments is another facility we provide for many young people - then we’ll have other records of them online.

SMS is another interface and, if we're really worried about someone, we'll always try to reach them on the phone.

So, within an episode of care, the user may engage with multiple different communication mediums. Ultimately the intent is always the same: to look after the wellbeing of the individual.

"Ultimately the intent is always the same: to look after the wellbeing of the individual."

That makes a lot of sense. We’ve chatted in the past about the “stepped care” model of health, is that still applied? And if so, does that apply online and digitally?

I absolutely think it still applies and I think people are engaging with this model, possibly not even realising it. Keep in mind that the more traditional definition of stepped care sees step one as primary health care, like your GP.

The next step might be a hospital or a specialist clinic setting, right through to acute service providers. This model points to very distinct, traditional, face-to-face steps of care. Most people are probably familiar with primary care, so their GP.

Definitions of stepped care can be a bit clunky because it places you in one discrete category or the other. I try not to get too caught up in these delineations because in the online space there is a tendency for everything to be lumped into one basket, or perceived as low intensity. In actual fact, in the online space, the potential to have multiple steps is huge, it’s yet to be fully recognised.

Earlier you mentioned the idea of communicating, contacting or helping people across different mediums, be it text, phone call, “reflective” email, or a mixture of each. Is this an example of a stepped care model?

Totally. There are a lot of scenarios where social media content goes out at the level of population health. One example of this which many people are probably familiar with is campaigns like “Feb Fast”, which is breaking from alcohol for February. There’s also Dry July. This messaging doesn't necessarily lead people to further steps such as consulting a GP, so I think of them as ‘population health’ messages.

Another example is the RUOK campaign. This messaging is designed around initiating, and being open to discussion. Hammering home that you’re not fully aware of what's going on for other people, but that we're not alone in our experiences. I think of this communication occurring at the population level, they target everyone rather than a discrete group within the population. These messages are important and I absolutely think they should be considered within our health framework.

So, you would place this level of messaging in the awareness category?

Yes, I think this level of communication plays an important role and it’s something we cannot afford to skip.

It's interesting because very quickly this messaging leads to the realm of combating stigma, which can be tricky to navigate. Is it fear about getting help? Is it ambivalence about getting help because people don’t want to be placed on a waitlist?

All of these elements are at play, so population-level messaging, I think works partly at addressing stigma, reducing fear and increasing understanding. It encourages people to seek help rather than trying to face their challenges alone.

So you see a big role here for education?

Absolutely. To be informed, first you need to understand what's going on. Then you need to know what will help you, even if it’s at the most basic level.

I think that population health absolutely needs to be assigned a level in this stepped care model, and to be understood for the significant role it plays. I think the next level looks to start tailoring an experience slightly more specifically. This may include self-guided assistance for more specific concerns like anxiety. If someone can identify that they’re struggling - that they’re not sleeping, that they’re mood's dropped - at this point, people might be seeking more specific, personally relevant resources.

And where do people find information that they can trust? ‘Dr Google’? :)

Yes, that’s a real point of concern. The commonwealth government has initiated a website, Head to Health, to help address some of these issues. It can be found at So again you see here that we're not necessarily talking about different stepped models, but we're starting to see a little more refinement in available resources.

Now, I'm beginning to consider what can an individual, who knows they’re not travelling that well, do to seek help? A possible answer: really raising the level of awareness and tailoring an experience so that someone can identify the steps they need to take. After that, you want to give people a way to connect with others, and social media plays a big role here.

Creating an interface starts to become quite a personal experience. It's not just automated features or static pages. We're talking about something that can connect communities.

Ultimately the intent is always the same: to look after the wellbeing of the individual.

Dr Steve Leicester

Based on our previous conversations, it seems like a lot of specialist communities benefit greatly from finding other like-minded folks. Autistic groups for example.

Absolutely. There is something profound about connecting to people of like-minds. I suppose prior to our major social media platforms, people used applications like MSN messenger and similar forums.

These types of platforms were coming in around the mid to late 90s. It was groundbreaking stuff!

People realised, "My God, there's a community somewhere out there that also collects green plastic toys on Saturday nights."

It's really interesting. I think there’s just so much that can be done in the space of online communities. It's more than just one or two people in there. In a group chat, we can talk about a specific topic for a designated period of time, moderated by an expert. It's the more organic forums where peer moderation is critical for it to work.

There are so many examples of online communities, and equally great examples of moderation. I see enormous potential for automation in moderation, but that it’s a little way off before we can apply it confidently. . For machine learning to gain an understanding of language patterns and to moderate beyond standard safety checks would be pretty incredible.

Have you got any examples of that type of thing?

Well, there's no shortage of data dictionaries which have been created such that if certain words come up, they're flagged. As a precautionary measure to potential risk, these posts may be held back, with potential for action to be taken.

Then, expanding on this you could create functionality with phrase recognition.

We’re not currently doing this, but there is a whole other level in which you create functionality which begins to recognise language patterns, tone, pacing, and context. It can indicate content which is inappropriate and block text, where appropriate.

It also has application beyond purely flagging risky language and behaviour. It could also be used to identify positive, affirmative contributions, which leads to some fantastic opportunities.

As with anything algorithm related, to create highly nuanced output you need mountains of data. But it is important to think about optimistic takeaways. How we can encourage valuable contributions rather than just stopping people from swearing.

This state of prevention is something that we need to be mindful of. This area is a pet frustration of mine and someone presented me with the following term to describe 'coronoia' which is paranoia of the coroner's court. I think this is something that creates substantial paralysis in our health systems. I think we absolutely need to assess, understand and respond to risk as best as we can, however we must find a balance to look beyond only containing suicide risk and doing nothing else. This is a tough challenge for our mental health system, but if we are genuine about addressing rising suicide, self harm and high risk behaviours we need to engage people in thinking beyond the immediate moment. It’s a fair concern but what if we challenge this thinking a little and extend beyond the binaries of life and death? If we’re only dealing in a risk-averse paradigm then all we’re doing is preventing an adverse event. We're not proactively doing anything to promote growth, recovery and greater health.

Imagine if Volvo was left to focus on creating a product which spoke to its primary marketing, that they are "the safest car in the world". You might end up with just a giant airbag, which doesn’t really address the goals of a function, usable vehicle. That's what it sometimes feels like. We're not actually engaging people in active treatment to move beyond where they are. I guess to do so would require some healthy risk-taking.

I think we need to have a clear purpose. Obviously, part of that is to keep people alive, but we also want people to thrive. In order to thrive you have to push yourself and, in a supportive mindset, you really need to shift from the immediate question of, "How do we keep you safe right now?" to "All right, you might not be safe right now but let's work towards understanding what's really going on."

I can call the police and get someone to a scene to deal with an immediate situation. The person will be a bit safer, I'll feel better and on my documentation will show that we did all the ‘right things’. But if that’s all we do, you're going to find the situation repeating.

I like to think about the concept of treatment as changing. Reacting can have a positive impact in the moment, but it doesn’t encourage people to change that much.

I think there are ways we can develop more forward momentum and the digital environment lends itself very well to this. We want to provide immediate resources for people in immediate need, but we also want to be talking about treatment.

That’s interesting. Can you talk a little bit more about the difference between one-on-one sessions and the experience of using a digital interface?

This is a fascinating point. It's something that we're working on at the moment. So with digital platforms, you’re looking at stepped care, population health, a large number of people, somewhat self-directed, tailored for the individual with the potential to connect with others in communities. One-to-one tends to be more specialist, with a mental health clinician or practitioner

The thing is, the one-on-one face to face counselling format is the only modality that our entire professional body of people, nationwide, are trained in. There is no tertiary training course which caters to online mental health interventions.

As a practitioner, we’re often trained in the concept of engagement in a face-to-face setting. Engagement is important when it comes to building rapport with a patient. It lets you get to know each other and create a sense of understanding. This goes a long way, and without rapport, people are unlikely to say much.

This level of engagement and sense of commitment might take a couple of sessions to foster in-person. In a webchat it can be quite different. "Bang," in the first sentence, "This is why I'm here, this is the shit. Let's get cracking."

So then, how would you describe the customer journey, so to speak, for a young person online to get to a one-on-one session?

Okay. In terms of steps, on the headspace website we used to have a section which asked, "Do you want to chat?" The user could click a button and ‘Bang’, they’re there. The trouble is people were bypassing other really important content. Really important self-guided material. I think it's important that people have the opportunity to see that there are a multitude of resources available in addition to one-to-one.

Then, we start getting into that idea of making informed decisions. It has the added benefit that, from a service demand and response perspective, it may lighten our load a little. But the main purpose of this set-up is people should be making informed decisions about what they access.

In the future, we’d like them to see information which gives people options. They’ll see self-guided content which they can work through. There'll be an increase in potential ways to connect with others - peers in moderated community group chats. But they'll also have the option of talking with a clinician and there will be a resource to facilitate that connection.

The way I’m describing this might sound a little slow, but with each of these options listed out, along with a little bit of personal data, hopefully, we’ll be able to collect more context and a greater understanding of the people we’re helping.

If they want to chat, we do ask for their name and email. The email address is then verified. This all takes less than a minute and a half. Then they're asked to complete what we refer to as a minimum data set, which is a series of questions: "Why am I here? What is the main issue?"

From clicking, "I want to talk" to completing the minimum data set would take around eight to nine minutes.

There is also a section for ‘free text’ where people can add additional comments. What’s particularly interesting is that most people will put something in this ‘free text’ section.

All up there are 86 questions. 94% of people complete these questions.

This sends a clear message that people who complete these questions are making a strong statement: “I damn well want a one-on-one”.

And the minimum data set provides a fair amount of information to help that?

Yes. The combination of the minimum data set with that free text is really important. The other thing is that for return users - most people will come back for more than one session - the senior clinician will be going back through previous case notes and care plans, so they'll be going through a lot of stuff.

They'll also be facilitating brief interactions while someone's in the queue. "We'll be getting to you soon, hang in there," - that kind of thing.

As queuing theory suggests, I think people are willing to wait a while provided there’s even a small amount of contact. Even if the contact doesn’t relate directly to their concern, it's still engagement, and people will wait a lot longer.

Even if the ‘no contact’ option means that you’d actually see a practitioner twice as fast, people prefer to be in communication.

When the clinician picks someone out of the queue, they'll screen everything, including the information the user has provided.

But they'll send a message through to the help seeker saying, "We've got you, we're just going through the information, we'll be with you in a matter of minutes"? Then they'll scan every bit of information they can.

And then go, "Hi, I'm Steve. Let's roll."

Great, okay. So there's a lot going on, just before that connection happens.

There's a heck of a lot and like with any health service, the front end is where the most expenses are incurred, the most adaptation and change occurs, the most potential confusion occurs, the area for highest potential risk.

When you say front end, you mean that first engagement?

When people walk in the door.

Unfortunately, when many of us have had to go to the hospital if you've been to an emergency department, the steps you go through, from filling in forms to the complexities of patient prioritisation, the time it takes to be seen, and if you're unlucky and need to be placed in a ward, the steps through that process are long and arduous.

Very complex. Very resource intense and if it's not done right, lots of bad things happen.

Yeah, absolutely. And that same resource cost applies in a different context here doesn't it? In a teleweb service I mean.

Yes, it's an interesting one. I think there are still some fundamentals which absolutely apply. So the therapeutic techniques or the core ingredients of good therapies, probably the most well known of which is cognitive behaviour therapy, still make up the essence of what we do.

The delivery of this, however, can be quite different. A lot of content is often provided. So we have a large library bank for our clinicians to use. We want to make sure that for every session someone engages with, they're exiting with resources and content which they can access. We want people exiting with a plan.

Even if that plan is to come back in two days, we want to make sure that this is a specified time. There'll also be an SMS sent through, to want to make sure the user knows what the next step is.

"We want to make sure that for every session someone engages with, they're exiting with resources and content which they can access. We want people exiting with a plan."

Gotcha. And the comms from that clinician, are they really different? Is the language adaptive?

Yes, we’re working with a very diverse group and our peak age range of engagement is probably 16 to 19-year-olds, tapering off to the early 20s range. We don't get a heck of a lot in the 12 to 15 bracket.

We really try to gauge where the young person's language is. So, for instance, in the past 5 years, we've really had to look closely at the investment and use of emoticons within our library.

We've got practise principles regarding how they're used.

We notice that people say, and I'm not being flippant when I say this, "Ah, I tried committing suicide." Well, don't send back a sad face.

We want to make sure our language use is appropriate. The use of emoticons can only be initiated once we've got evidence that the person's using them. The interface that the young person is using on our website will have the capacity for them to draw from an emoticon library. So this is actually a really major issue and we're learning as we go.

You spoke earlier about data collection from users and their ability to remain anonymous. How does that work for the clinicians?

The clinician will come across only with their first name only, so the anonymity of the practitioners is important for a number of reasons. This poses an interesting question because now we've got the functionality for video interface ready to go up on the website in the coming months.

We really need to think about “how and who” that would be most appropriate for. Most likely it'll be with those people that we're working with over a longer period of time. But there are also logistical considerations to take into account. Rooms need to be set up for these sessions. Additionally, does the practitioner want their identity known?

Reason being, many of their clinicians might be doing private practice and could be easily found. It’s an interesting quandary in this area, but that whole thing - one-to-one video interfacing - is another, more intense step.

headspace does provide telepsychiatry. So getting psychiatrists to do consultations with young people in clinics across the country. Especially, when we're getting into very specialist mental health treatment.

We're also seeing a lot of people who are specifying that this is the medium they’d like to use. The premise that webchat or phone sessions should only be available for mild to moderate, is a nice idea. The reality, however, is that we're seeing a lot more people with high complexity concerns, an extensive history including face-to-face and specialist mental health services, who are now choosing these options instead.

And it is popular?

It is really popular.

And it's not just geographically driven, is it?


I think there was a conversation a while ago that these services were only of value to those in rural or remote areas and that's just not the case is it?

Steve: Look, it really isn't. I think these services carry immense value and I think one of our challenges is to really look at how we can properly target our engagement in regional Australia. I think that's a really important one. But I think one of the key things is that our population distribution kind of mirrors where the primary points of access are, so down the eastern seaboard in big capital cities.

But getting back to one of the comments I made before, senior clinicians are tasked with prioritising a queue of young people who want these services. If someone's acutely at risk of doing something to themselves or others, we'll get to that as quickly as possible. But this needs to be balanced because there are numerous at-risk groups but we still need to make a decision on who to treat. An example of this has come up. We had two people in there who were reporting being suicidal. One was a 14-year-old male. Now, we know he’s a young man from regional New South Wales and it's their first time ever contacting the service. We'll go for him because if we stuff it up, the likelihood of him returning is limited.

These clinicians are balancing this concept of risk and need. For this 14-year-old there’s a high risk of him saying, "Blow this, I'm never using this again".

Obviously, this is a conversation based around research pieces, evidence and amazing background learnings. Do you ever know you're getting it right? Or is it the best choice at that time?

Ultimately, it’s about making the best choice you can in a given moment. That can be really tricky. But it's also having those mechanisms in place to ensure that there are clear resources available, no matter what time you’re seeking assistance, it’s always there.

So again, those points of interaction are so key. Even if you're waiting, you know we will get to you really soon. This is critical.

When your resources are so short and the demand is so high. I suppose any save is a win, isn't it?

Yeah. So I think that one of the key purposes of your work is creating a good design, understanding the nuances and experience of service pathways. This is so critical.

It is, isn't it? I think that's a really good insight. Considering process pathways and users' potential behaviour, and then balancing that with resources and outcomes is important. Improving efficiency and saving time in this context can be redirected to contacting and helping a whole lot more people.

All of that going from the population health through to specialist mental health, through psychiatry and psychologists and OTs and social workers and nurses, is quite a spectrum.

Across what might just appear as a website, there is absolutely an opportunity to create these levels of care and support.

We want to make sure that for every session someone engages with, they're exiting with resources and content which they can access. We want people exiting with a plan.

Dr Steve Leicester

Steve, it's been really good. In wrapping up, is there something you want our readers to know about youth mental health, about the service you’re offering service? A final takeaway.

Give us more freaking money!

No, no. I think one of the main things is we need to stop making assumptions about who accesses materials online.

You can purchase any service, any commodity, really, anything online and healthcare is no different. Now, your capacity to do other things online such as purchasing a car, clothes, a cup, is not decided upon according to your level of incapacity. Everyone can do it.

We've got to stop making assumptions. Assumptions that this is only for specific groups. There's some fantastic evidence coming out from the likes of Nick Titov from Macquarie Uni, which suggests that when delivered well, online interactions with practitioner delivered psychological therapies are absolutely as good or even better than face-to-face service delivery with regard to client outcomes.

And if we're not looking at client outcomes, we've really got to ask, what are we doing?

And this is the other thing, with all of these steps I've referred to, even if they're in the most tricky situation and seeing a psychiatrist, population-level health material is still something you should be accessing.

I don't know what the solution is, but if we can construct services in such a way that people can make informed choices for themselves, I think it will absolutely lead to better outcomes.

Like many industries, in mental health and the health sector, we’ve got big, big dilemmas which are only going to grow with regard to workforce distribution.

We need to be smart about the distribution and what that looks like for practitioners. Smart, accessible, digital design is absolutely going to be part of that.

Thanks again for your time Steve. Wishing you and the team at eheadspace all the best for helping young people across the country!

We've got to stop making assumptions. Assumptions that this is only for specific groups. There's some fantastic evidence coming out from the likes of Nick Titov from Macquarie Uni, which suggests that when delivered well, online interactions with practitioner delivered psychological therapies are absolutely as good or even better than face-to-face service delivery with regard to client outcomes.

Dr Steve Leicester
Today Team Kate

Get in touch

Want to chat about news in the impact space? Get in touch with Kate.

Kate Bensen


Next Article

Invention and the health tech revolution

Site footer