Hackathon Learnings — #Hack4ChildMentalHealth

EXECUTIVE SUMMARY

Blockchain Learning Group and TMX Group proudly presented the “Hack 4 Child Mental Health” Hackathon, a not-for-profit initiative on May 11–13, 2018 in Toronto, Canada.

For this inaugural Hackathon, the goal was to have participating teams develop innovative solutions that address awareness, early detection, prevention, intervention and treatment opportunities for children and youth mental health issues. Using data and technology, 11 teams representing over 73 participants dedicated three (3) days to develop solutions within an agile sprint-like environment. The participants ranged from 15 year-old high-school students to seasoned software developers and data scientists, and everything in between.

The solutions developed addressed a wide cross-section of challenges and opportunities around child and youth mental health issues, including: improving access to and quality of knowledge and support tools; increasing collaboration within and between the practitioners and their patients; developing platforms to collect, analyze, and disseminate data; encouraging earlier detection, dialogue, positive behaviour and support among affected parties through incentive-based Apps; and reducing the stigma associated with Mental Illness.

The teams pitched their applications and solutions with 5 minute presentations that highlighted the problem they were attempting to solve, and the solution they were putting forward. Prizes were awarded to the top three (3) teams, by a panel of judges representing clinical, academic, entrepreneurial and technology industry sectors.

The event was very well received and yielded tangible and relevant solutions, some of which have the ability and promise to be utilized within the mental health arena today. The solutions created addressed real issues and opportunities facing child and youth mental health realities.

This first successful child and youth mental health hackathon served to establish a viable platform for leveraging technology and data solutions to drive innovation around mental health realities. Plans are already underway for next years event, tentatively slated for April 26–28, 2019, building upon the momentum from this year’s hackathon.

BACKGROUND

OBJECTIVES:

The goal of the hackathon was to enable a platform to develop innovative solutions that would address child and youth mental health issues. One of the ways we felt that we could achieve this was to leverage the insights and talents of the broader technology community in Toronto to collaborate within a hackathon format and build functional solutions for this cause.

LEAD UP:

The idea for the hackathon was borne from a desire to achieve greater inroads on mental health innovations for children and youth in Canada. In late 2017, a group of Toronto-based executives, mental health organizations and practitioners bound together and committed their time to addressing this stated objective.

A steering committee was created to help guide the strategic roadmap for hackathon. This group worked tirelessly in the months leading up to the hackathon to outline the framework and organizational requirements for the event. The committee included:

Chami Akmeemana | CEO | Blockchain Learning Group Inc. & Blockscale Solutions Inc.

Maureen Jensen | Chair and Chief Executive Officer | Ontario Securities Commission

Kathy Soden CPA, CA | Manager, Permanency & Adoption Competency Training (PACT) Program | Adoption Council of Ontario

Marc Folch | Senior Manager, Digital Payments Insights | Interac Corp

Vipan Nikore MD, MBA | Medical Director of Dept of Medicine at Credit Valley Hospital, Physician at Cleveland Clinic | Trillium Health Partners and Cleveland Clinic (Toronto and Cleveland sites)

Inga Dóra Sigfúsdóttir | Founder and CEO | The Icelandic Centre for Social Research and Analysis

Barbara A Fallon, PhD | Associate Professor & Associate Dean of Research
Canada Research Chair in Child Welfare | Factor Inwentash Chair in Child Welfare|Factor Inwentash Faculty of Social Work | University of Toronto

Jean Desgagne | Former President and CEO, Global Solutions, Insights & Analytics Strategies
TMX Group

PROBLEM STATEMENT:

After setting the date and location (generously provided by the TMX Group), the first order of business was to create a Problem Statement for the hackathon. For the inaugural event, we determined that a broad-scope problem statement would ensure we did not limit or constrain the solutions to any one particular track or focus: .

Develop innovative solutions that address awareness, early detection, prevention, intervention and treatment opportunities for child and youth mental health issues.

COMMUNICATION:

A website was created (https://hack4childmentalhealth.org/) to be the primary vehicle for driving awareness and interest for the event. The website was launched in February of 2018.

We then embarked on a digital campaign to drive awareness, interest and participation for the hackathon. Our committee members utilized targeted email campaigns and reached out to their respective networks to promote the event and solicit presenters, judges, volunteers and off course, the participants for the hackathon themselves.

DATA-THON:

The goal initially was to also include a datathon component for this event, leveraging available data sets to develop either actionable insights and/or predictive models and solutions. Ultimately, we did secure a data set from Kids Help Phone that was utilized by some of the teams at the hackathon. Unfortunately, the availability of data within the mental health arena proved difficult to identify and acquire leading up to the hackathon, despite considerable efforts. Healthcare data, particularly mental health and youth data, is confidential given the sensitivity of the data. We will address the go forward strategies for data collection in the learning section of the report.

EVENT SUMMARY

HACKATHON OVERVIEW:

The hackathon took place over a three (3) day period from Friday May 11 through until Sunday May 13, 2018.

Presentations at Opening

For the kick-off on Friday, presentations were made by leaders in their respective mental health fields to provide greater context on the Problem Statement for the participants.

Representatives from a cross-section of mental health agencies and organizations were curated to paint this landscape for the attendees, including the Centre for Addiction and Mental Health (CAMH), The Hospital for Sick Children, Kids Help Phone, Adoption Council of Ontario, University of Toronto, The Icelandic Centre for Social Research and Analysis, and Covenant House. A full line-up of all the speakers and presentation topics can be found below:

Inga Dóra Sigfúsdóttir | The Icelandic Centre for Social Research and Analysis

“Improving Lives based on Evidence”

Adeena Persaud | The Hospital for Sick Children [Infant Mental Health Promotion (IMHP)]
“Infant Mental Health Considerations”

Kathy Soden | Adoption Council of Ontario
“Improving Systems = Improving Outcomes”

Nicole Thomson | CAMH
“Child and Youth Mental Health Services: Bridging the Disconnect”

Barbara Fallon | University of Toronto
“ How can Data help to drive practice and policy in child welfare”

Robert Fraser | Covenant House
“Realities of at Risk Youth who utilize Covenant House Resources”

Kathy Hay | Kids Help Phone
“Child and Youth Mental Health”

Dr. Mandeep Singh (Psychiatrist) and Dr. Vipan Nikore (Internist and Digital Health Expert) | Trillium Health Partners
“Observations and considerations from the front lines”

Following the individual presentations, the industry thought-leaders participated in a panel where they fielded questions from the audience. This provided the participants with additional insight and guidance for the coming weekend of ‘hacking.’

Teams:

After the presentations, the team formation was addressed. Many had already registered as a group leading up to the event, and for those that hadn’t, the mentors from Blockchain Learning Group helped to place everyone on a team based on skill set, needs and fit.

At the end of this process, there were 11 teams and 73 ‘hackers’ in total who participated in the event. The backgrounds of each team member ranged from experienced developers and data scientists to business analysts, product, marketing and high school students.

Use Case Ideation:

The balance of the Friday night was spent formulating the specific use cases that each team would solution against. This was the most nuanced part of the hackathon, where numerous ideas were discussed and assessed on a variety of merits, ranging from applicability of the contemplated solution to market readiness and originality.

Once again, the Blockchain Learning Group circulated from team to team throughout the night to help guide the use case ideation. By Saturday morning, each team had formulated their respective use cases and began the process of developing out their solutions.

The ‘Hacking’:

On Saturday through Sunday morning, the teams broke out into various meeting rooms and office nooks to develop their solutions. Throughout this development stage, mentors from the Blockchain Learning Group and Blockscale Solutions provided technical and business solutioning support to the teams, acting as a sounding board for ideas and helping the teams navigate the decisioning and trade-off criteria for their deliverables.

The teams had until 10am on Sunday morning to finalize their solution and pitches. Presentations commenced at 10:30am and continued until 12:30pm.

Judging:

The judges that were curated for the hackathon represented a diverse cross section of the mental health arena, from technology, practitioners, entrepreneurs and academia:

Barbara Fallon, Associate Professor & Associate Dean of Research, University of Toronto
Huda Idrees, Founder & CEO, Dot Health

Vipan Nikore, MD, MBA, Credit Valley Hospital & Physician at Cleveland Clinic
Nicole Verkindt, Founder & CEO of OMX

The criteria that the judges utilized to assess each ‘pitch’ was based on a 10 point scale (equal weight for each area, for a total of 100 points):

Criteria

Explanation

Design

Is the UI/UX attractive or pleasing enough for users to start using the project?

Simplicity

Is the tech used in the project overly complicated?

Originality

How original / unique is the project?

Market Readiness

How close to market is the state of the project?

Innovation

How innovative is the project, versus a marginal improvement?

Usefulness

Does the product or service actually have a use case?

Features

Does the project include multiple features?

Market Potential

How big is the market the team is going after?

Market Analysis

Does the team accurately represent the market and the opportunity?

Pitch

Overall quality and professionalism in demo presentation

TEAM SOLUTIONS:

Each team was given five (5) minutes to pitch to the judges and audience. The following summarizes the solutions that were presented by each team, in order of presentation:

Team Name: Impetus

Members: Shams Mehdi, Talha Atta, Abhinav Boyed, Samarth Athreya and Shehryar Assad

Problem: In the world of academia and research papers, the existing ecosystem of funding may create an environment where researchers could be tempted to embellish or sensationalize their studies in order to secure or maintain funding. Additionally, the system also allows for large corporations and organizations to promote their own narratives and objectives through funding supposedly ‘unbiased’ research studies in a manner that masks their involvement.

Solution: The team developed “Impetus,” a platform that rethinks the economic model behind research and academia. It rewards the generation of good data as opposed to news making research papers. For researchers, this means coming to the platform to get funding, execute research and publish data — think Github + Kickstarter for the academia world. This system is also fully transparent, so that the likelihood of undue influence being exerted from self-serving interests of large corporations is severely limited.

This platform also recognized the value of good data points within a study that might never end up being published, yet there existed these specific data points that had considerable value and could be shared as part of the platform ecosystem. Their project utilized Blockchain and Smart Contracts for its properties of immutability and tokenization to support this application build.

Team Name: ReachOut TO

Members: Jenny Feng, Marian Hernandez, Francis Ho, Michelle Liu, and Trent Chappus

Problem: When it comes to mental health resources, a problem that communities are facing is access and awareness. The people in need of resources are often unsure where to look within their community, and thus are unsure of where to turn to in times of need or crisis. Communities are struggling to effectively communicate the availability and access to the resources they provide.

Solution:This team created a single-stop platform to connect youth and parents of children to mental health resources in the community. This model could be used by governments of municipalities, provinces, and college/university campuses as a way to communicate the resources they provide to the people in need of them.

The team developed a website in which the user is presented with a simple questionnaire to assess the situation they are facing (a few examples include depression, domestic violence, addiction, grief, or homelessness). The platform then uses the user-provided answers to best determine the situation and provide them with a list of nearby and applicable resources. If they wish, the user may provide some demographic information to help the platform in recommending more specific resources. This data could also be used to follow up with the individual and further drive research and assessment of needs.

Team Name: Autism Heros

Members: Jean-Anne McQuarrie, Denella D’Souza, Swapna Sangayya and Durga Lakshmi

Problem:To cut down wait times for a child to be diagnosed within the Autism Spectrum Disorder. Current wait times to see a developmental pediatrician are approximately 32 months and greater. This is costing both the province and society valuable resources and funds. Early diagnosis is key to ensuring positive and timely outcomes for an individual child. Ontario Health has also cited that getting both an earlier diagnosis and treatment is key to the success of the child.

Solution: This team created a simple but powerful common calendar system to connect General Practitioners and Pediatricians to the Developmental Pediatricians in order to get access to all possible appointments from a wider pool of practitioners. A gui front end was created that would interface with various APIs that works with all email platforms (ie. Outlook, Lotus Notes, gmail etc).

Team Name: Attentia
Members: Morgan Moskalyk, Shiv Gupta, David Nagel, Eric Bye

Problem: Screen time, and social media use in particular, can negatively impact children, youth, and young adults. Children aged 8–12 spend nearly six hours a day consuming media and teens are spending nearly nine hours a day consuming media. Too much screen time damages the brain and risks children’s mental health.

Solution:Their solution aimed to create social accountability to influence behaviors and create positive habits around technology usage. Their strategy was based on a gamified preventative intervention for communities. Their App worked to have:

• Sponsors provide leadership (volunteering time) and/or monetary support that feeds into the rewards system.
• Friend Group co-create the goals with the sponsors
• The goals and the corresponding rewards were coded into the Smart Contracts using Blockchain Technology, where the actions of the friend group could positively or negatively affect their progress towards their goals.
• For example, should a friend group play a bit too much PS4 and not quite reach their goal, the rewards might roll over to the next challenge, or be donated to the local community center. But when a friend group successfully reaches their goals, they’ve earned their collective rewards.

By understanding the tradeoffs between Netflix binging for hours versus exercising with your friends, this system is designed to encourage the friend group to support each other in developing mindful habits for intentional uses of their technology.

Team Name: PsyDex
Members: Sacha Sukhdeo and Hammad Ahmed

Problem: The time period between scheduled clinical visits between patients and Psychotherapists is under-utilized and can be used to used more effectively.

Solution: PsyDex created a mobile app (Android & iOS) that connects Psychotherapists with their patients between therapy sessions. For patients, PsyDex lets them report their levels of stress, anxiety, and anger on a daily basis. A therapist can view these daily vitals for each patient on their dashboard and submit personalized questions to better gage a patient’s mental state. A patient can view these questions and submit an audio response for their therapist. This allows for monitoring of a patient’s mental health on a daily basis and can help power more effective therapy sessions at the right times.

PsyDex is not a remote therapy service nor does it employ or pay Psychotherapists. It only seeks to enhance the relationship between a Psychotherapist and their patient.

Team Name: Thought Bubbles
Members: Jonathan Perez, Jin Xiag, Jordan Walwyn, Jason Domagala-Tang, Jimmy Domagala-Tang

Problem:50% of children have an adverse childhood experience (ACE). Researchers estimate that up to 2 of 3 drug addictions are caused by ACEs

Solution: This team developed an incentive-based App for parents to monitor and track early childhood milestones that may have an impact on later mental health issues. The app asks the parents for their general thoughts and then pushes out messaging to influence behaviour and general awareness of the environment their child is in (i.e. Did you know that spending 1-hour a day with your child reduces the likelihood of mental health by 85%)? The more interaction the parent has with the App, the more “Bubble Rewards” they collect, which could be redeemed at participating sponsors such as Starbucks or Freshii.

Team Name: Thera Notes
Members: Stephen Gibbs, Charlie Frantowski, Andre Di Genova, Frank Warnock, Ryan McCaul

Problem: One concern with introducing any enabling technology with health care professionals is that they already have so much to do, and any additional task added to their workflow — no matter how beneficial — would be inhibiting.

Solution: To address this constraint, team Thera Notes developed a Google home app capable of recording therapy sessions, executing automated sentiment analysis and producing reports. This allows data collection to occur with minimal impact to the interaction between the clinician and the patient.

This technology promises to take some of the workload off of healthcare professionals while providing a platform for data collection and analysis. The team leveraged cloud technologies and employed voice recognition, natural language processing and machine learning to build out the platform. The platform included a web application that aggregates the resulting reports which can then be searched and viewed by healthcare professionals.

They also analyzed a dataset of approximately 300,000 chat logs provided by Kids Help Phone. Using Latent Dirichlet allocation, they were able to find topics and map them to individual sessions. They demonstrated one possible application of this analysis by displaying topics with each of their respective word weightings as word clouds in our web application.

Team Name: Insominacs
Members:Tony Xu, Brendan Wu, Soroush Khoubyarian, Bamdad Sahraei, Amirreza Moghaddam

Problem:Through data-driven insights, this team found a direct correlation between a lack of physical activity/family bonding with youth suicide rates. They determined that exercise, sleep, and parental love were imperative to a child’s mental well-being and that technology could play a complementary role in enabling more family and leisure time.

Solution: They developed an incentive-based app that makes suggestions on family activities and other extracurricular activities. Examples of the recommendations they put forward included visiting the art gallery, taking a bike ride, and attending sport camps. Parents are rewarded with journey points for the completion of these activities with their children which can be used as discounts for futures events. They also made a website to serve as an admin panel for organizations to add/edit events and verify attendance and reward points earned.

The goal was to remove the hassle in the planning process, and promote face-to-face interactions to bridge the disconnect caused by technology.

Team Name: Ansar and Friends
Members: Ansar Butt, Victoria Chernoukhova

Problem: This team looked at ways to support Kids Help Phone enhance operational efficiencies. While Net Promoter Score was recently rolled out as of March 2018, not a significant amount of data had been collected on it to-date, and survey response rate was random. The current counsellor scorecard focuses on adherence, volume and overall average handle time (AHT). There are currently 228 call issues types listed and assessed on an overall or LOB level.

Solution: The team set out to create a Proficiency Rating.

• Firstly, they narrowed and grouped the 228 call issues into a more manageable set of issues (in this case 15, but this customizable)
• Next, for each call issue, they standardized a counsellor’s AHT by the mean AHT of the population. For each counsellor, this creates a standardized AHT metric for each call issue as well as an average standardized score overall.
• The score allows a counsellor or supervisor to easily see how their talk times perform with respect to the overall population.
• As counsellors approach the mean talk times in each category we may be able to capture several underlying benefits. For instance, when this system was implemented in a US health care call centre the following benefits were realized: agent occupancy down, morale up, absences down and quality scores up (source: http://www.swpp.org/member/newsletter-archive/spring10/proficiency.html)

Team Name: mSights
Members:Andrea Wong, Yuki Wu, Will Zhang, Terry Wang, Hang Wu

Problem: Mental health providers and resources are stretched thin, particularly for children and youth under the age of 18. While the demand for services is climbing, and there are currently over 1 million Canadian children afflicted with mental health conditions, a great number of children are still not obtaining access to the specialized services they need.

Solution: The mSights tool uses chat, voice, EEG, and visual data inputs to interpret trends in the mental health status of patients, in both clinical and home settings. The utility of this technology is in helping providers monitor the in-session, between-session and ongoing improvement of a child’s mental health symptoms. Using individual trend and population aggregate analyses, the end product — a virtual assistant — will be able to interpret mental health service needs based on risk level, allocating a network of trained chat bots (low-risk), therapists, counselors, and doctors (medium-risk), as well as psychiatrists (high-risk) to patients so that each child can receive the right care, through the right channel, at the right time. The team anticipates that their technology has the potential to save the healthcare system up to $365,000 in lifetime health costs per individual supported.

Team Name: Fresh Data Mints
Members: Manik Chaudhery, Dona Maria Absi, Antonio Kim, and Lori Desjardins

Problem:Kids Help Phone provides a text and chat service as a method of communication to its users. One challenge with this communication vehicle is to be able to gage tone in order to truly assess the state of mind and what the child or youth is actually trying to convey.

Solution: Live Chat and Text Messaging App. The team analyzed the data files provided to them by Kids Help Phone. They targeted the problem of the lack of the ability to express emotions while chatting with counselors through text and built a solution around it. They mined the data to analyze the different words that convey a variety of emotions such as Loneliness, Depression etc. and then built a real-time chat analyzer that proposes to help train new and help existing volunteers and counselors to better gauge the main issues of the child or youth who has contacted them.

OUTCOMES

The following three teams took top honours at the hackathon:

1st Place: Team “PsyDex”

Team Members: Hammad Ahmed and Sacha Sukhdeo

Prize Money: $7,500

Prize Sponsor:

2nd Place: Team “Thera Notes”

Team Members: Stephen Gibbs, Charlie Frantowski, Andre Di Genova, Frank Warnock, Ryan McCaul
Prize Money: $5,000

Prize Sponsor:

3rd Place: Team “Fresh Data Mints”

Team Members: Manik Chaudhery, Dona Maria Absi, Antonio Kim, and Lori Desjardins

Prize Money: $3,000
Prize Sponsor:

LEARNINGS & NEXT STEPS

Learnings:

• Obtaining access to the data was challenging. Either organizations who possessed the data were unable to share due to confidentiality reasons, or in some cases, simply did not see the value in providing data.
• The data set that was utilized yielded valuable results. Multiple groups worked on this data set and the solutions developed were judged to be useful with potential to add immediate value to the organization that provided the data. Having the appropriate cautions around privacy and disclosure were important measures that allowed the data remain secure, and could be used as a model for other healthcare organizations and holders of data to use in the future.
• Engaging healthcare practitioners is challenging. There is only a small number of practitioners with experience and interest in data and technology. Their involvement and their insight is critical. It will be important to promote widely and early to practitioners, and also to emphasize that they do not need to be experts in technology and that instead their clinical insights add the most value.
• The education talks and panel from day 1 of the hackathon was critical with the hackathon participants given the lack of experience of most developers and data analysts in mental health and given the complexity of the field.

Next Steps:

• Based on the overwhelmingly positive feedback from this year’s event, we have already set a tentative date for next year’s hackathon: April 26th, 2019.
• Leading up to next year’s hackathon, we will engage the various Ontario Government agencies related to Child and Youth Mental Health as well as a broader set of healthcare organizations to encourage their participation in the early stages of planning.
• Based on the resounding success of this hackathon, we are confident that we will be in a much better position to demonstrate the value of opening up the data for an event such as this, and thus be able to bring a greater amount of meaningful data to the table for next year’s event — we are already having discussions with the appropriate parties to enable this.

We will highlight the success of Kids Help Phone data as an illustrative example of the value to organizations in opening up their data. As such, we will aim to have next year’s event be a true 50/50 Hackathon/Datathon split.

• As much as we can, we will continue to help assist and support the groups that are moving forward with their projects from this hackathon.

Authored By:

Chami Akmeemana

CEO Blockchain Learning Group Inc. / Blockscale Solutions Inc.

Barbara Fallon

Associate Professor & Associate Dean of Research, University of Toronto

Inga Dóra Sigfúsdóttir

Founder and CEO, The Icelandic Centre for Social Research and Analysis

Vipan Nikore

Site Medical Director at Trillium Health Partners Credit Valley Hospital,

Physician at Cleveland Clinic’s Toronto and Cleveland sites

Jean Desgagne

Former President and CEO, Global Solutions, Insights & Analytics Strategies
TMX Group