Improving Health through technology

Editor's Notes

• #3: The agenda for this morning.
• #4: First Age Relevant Search Engine Backup and Recovery Data Storage Systems Even an Obituary Website
• #5: I wanted to try my hand at Healthcare Joined in Fall of 2009 Employee #1
• #6: Disease management company founded in 1981 Focuses on analyzing claims data to determine who to target Driven by cost savings Hundreds of certified nurses who reach out to patients over the phone
• #7: We built a company in the South End. Proud to have received Boston Globe’s Top Place to work in 2014. [CLICK] We even allow babies!
• #8: 2009 A simple idea, not so simple to build.. I’m a technologist, a few data bases, some UIs and we’re done right? Wrong.
• #9: Well, I need to get smart about this.
• #10: Dots = Do something once Spans = Do something for a period of time Path = Do something for ever [CLICK] Green = New (try Quinoa) Blue = Do a familiar (flossing) Purple = Increase behavior (eat better) Grey = Decrease behavior intensity (drink less coffee) Black = Stop a behavior (stop smoking)
• #11: Design systems that allow people to place bicycle at the top of the hill when they are at times of high motivation [CLICK] Think leaving the movie theater after a great sports movie Let them ride the bike down the hill when they are in periods of low motivation
• #12: Changing behavior is only one part of the equation. If you can’t engage people you can’t have impact.
• #13: We need to design product that people will actually use. This is a typical Health Portal. Notice the massive drop off. We need to do better
• #14: We need to design product that people will actually use.
• #15: The Health and Fitness engagement products traditionally have very low engagement.
• #16: By appealing to different types of users we maximize our ability to engage them Some people prefer prescribed interaction. Step by step guide on how to get fit Exploration: Give me a goal and let me explore my way there.
• #17: We know we’ll need engagement and we know we need to change behavior. Now which ones?
• #18: Target the ones that are behavioral choices that cost the American healthcare billions of dollars. Improve Well-Being Eat Better Stop Smoking Walk More
• #19: Great, we know what behaviors, know we’ll need high engagement. Now how are we going to get started?
• #20: Assemble a team of: Designers, Engineers, Product Managers, Data Architects, and Scientists.
• #21: We’ve got the team, we, know the behaviors, we know we’ll need engagement know how do you get started?
• #22: Talk to people and ask them lots of questions.
• #23: 34 participants Ask them about their well-being: Emotional, Financial, Physical, Community and Purpose
• #24: Lots of time in dark rooms seeing how real people describe their well-being Looks a lot like this [Click] And like this [Click] Ironically listening to people talk about their well-being isn’t great for your own. [Click] Notice the bag of Cheetos
• #25: What you’re left with is lots of qualitative data. Mountains of it. Papers, video and notes, lots of notes.
• #26: Take all that qualitative data and identify patters to describe your users.
• #27: Me Time – Mostly moms juggling work and family. Looking for something guilt free [CLICK] Validation seeker – driven by numbers wants to compare themselves against others [CLICK] Aware and Achieving – Already doing the things we’re targeting. Like being coaches to others [CLICK] Enlightened and Discovering – Are interested in learning about themselves [CLICK] Idle – Looking for social interactions in a group setting [CLICK] Excuse maker – Values privacy, doesn’t like to share needs a very low bar [CLICK] Enabled – Looking for fast results, likely to want a personal coach or sponsor to help keep them motivated
• #28: Realistic: This isn’t marathon training Convenient: It needs to fit into their day. It can’t be a time suck Immediate: Give them something they can do right now. Genuine: Can’t be quirky or feel outside their comfort zone
• #29: Got all the behaviors, know we need engagement, know our personas now we need to get to work.
• #30: MeYou Health’s process for designing software. I’ll review the process at the end if there’s time and interest.
• #31: Here is what we’ve built. DC: Engage with people’s well being everyday Hello 200: Reduce a person’s diet by 200 calories each day QuitNet: Get people to stop smoking Walkadoo: Get people to walk more
• #32: I’ll introduce each product and do a deep dive on Walkadoo Daily Challenge - Video
• #33: Before we move further, I want to give you a sense of what a challenge looks like and introduce a few challenges throughout this talk. Here’s the first one. Take a few minutes to complete. Why it matters Balance is critical for a healthy, happy life, but it can be hard to achieve. Seeing your life as pieces of a pie can help provide a reality check, and may motivate you to look more carefully at your priorities.Are there some pieces that are too small? Are there ways to increase the size of those pieces, and cut down on others?
• #34: Here’s my chart. I think I need to increase my family time. Maybe I’ll steal some time from sleep
• #35: Hello- Video
• #36: QuitNet - Video
• #37: Deep dive into Walkadoo
• #38: Walkadoo - Video
• #39: Why focus on walking? It’s where the people are. The vast majority of Americans are either Sedentary [CLICK] [CLICK] or Low active
• #40: So how do you go about designing a program that gets people to walk more? [CLICK] You listen some more
• #41: Interviewed 27 participants Motivations and Barriers Where they get their Physical Activity How do they think about it?
• #42: Create an application that helps people move through the behavior change model Appeals to a large audience. Keeps people engaged. We cam to the conclusion that we’re going to need a pedometer.
• #43: We learned it’s all about the devices [CLICK] And we’re going to need one. STOP FOR A CHALLENGE!
• #44: Why it matters Stretching your upper back and shoulders builds better posture and relieves stress. If you've been doing housework or gardening, hugging yourself will give your body some relief. The same thing goes for people whose work is more physical, or those sitting at a computer for hours.
• #45: Ecosystem of wireless devices Devices collect data on steps. Number of steps and intensity of activity (walking vs. running) Transmit data to a gateway, usually over Bluetooth Data is stored in the device maker’s database We access the data through public APIs We use the step data to power Walkadoo
• #46: Which device? We play test a lot of devices. This is a small sample of the devices in the MeYou Health device graveyard. They Vary by: Weight Size What they collect User Experience
• #47: We wanted to achieve a few things To get smart about how they worked and what makes them different To understand the economics of the building a device To prove we could build one for under $5 It has all the basic components of today’s wireless pedometers
• #48: Early ideas for how we might design a different device [CLICK] Wrist band [CLICK] KEY FOB [CLICK] CREDIT CARD [CLICK] But should we? We were built as a software company. Focusing on hardware would distract us from our core mission In the end we decided to take a BYOD (Bring your own device) or device agnostic approach to our fitness product.
• #49: Early wireframes of the first user experience. Code named Hot sauce. We designed a fitness game where you are dealt a card each day with your step goal. Complete the goal. Earn points Interact with other members through leaderboards
• #50: After play testing our our MVP we learned a lot about how people actually walk Even in an office with healthy young adults we found: Lightly active – long commutes lots of sitting Moderately active – mass transit, dog walking etc. Highly active – runners/athletes We used this data to understand how different types of users would interact in one system
• #51: Cleaned it up to prove the proof of concept. Even with an MVP you need to get people excited with some eye candy Put this website out on a URL to gauge interest
• #52: After months iterating on a new version of the product ready for consumers and partners Walkadoo – Let’s walk more together
• #53: Here is what Walkadoo looks like today. Using similar card metaphors with some modifications to the UI and the user experience We added points to help normalize the experience for all types of users
• #54: Personalized goals for each member. We learn your walking patterns send you goals based on what we think you will walk and add in some game mechanics.
• #55: Limited: Earn points up to your goal All or Nothing: Only earn points if you make your goal Unlimited: Earn points all day
• #56: Points to teach user preferred interactions Streaks to encourage adherence Stamps to create a sense of mystery and intrigue Levels to show progression within the system
• #57: Social drives engagement. Think about how you’re prompted to return to Facebook and Twitter. Strong engagement from being prompted by a social interaction. Walkie Talkies are a Daily prompt or conversation starters to encourage members to share Posting milestones or achievements to other members steams to create interaction moments
• #58: Derbies provide members the ability to invite 5 other members and engage in a competitive week long walkathon. You can compete at a step derby or a points derby. Points derbies allow all levels to interact. It’s like handicapping in golf. It allows everyone to participate on a level playing field.
• #59: Why is it important for everyone to play together? [CLICK] Creating systems that involve multiple users allows us to engage the network into the intervention. Norming users by introducing them to other people doing model behaviors changes the behavior of the user.
• #60: Dr. Nicholas Cristakis and Dr. James Fowler proved that behavior is contagious both good and bad behaviors. [CLICK] What if we infected the network with healthy behaviors?
• #61: Remember these people. Well, we asked them to draw diagrams of the people who effect their well-being. We found some interesting things.
• #62: The drawings identified themselves in the middle of the drawing The people in the outer circles are people who influence their well-being Directional arrows show whether they influence the person in a single direction like a child or a pet Bidirectional arrows show that they influence each other’s well-being like a spouse or friend. I documented all the drawings and found 43 participants identified 217 people who have some influence in their well-being
• #63: Here’s where it gets interesting Of the 217 people in the drawings only 21 were identified as coworkers
• #64: Why does that matter? Most workplace wellness programs only allow employees to participate in the program. They are missing 90% of the network who can have an influence. The blue dots represent a sample of users in a healthcare plan  The red dots represent everyone else in the system. The lines connecting them represent social interactions If the network is the intervention, you need to use the whole network
• #65: Wal-Mart understands how this works They have a clear understanding of how friends and family members influence the lives of their associates. This is a sample network graph of Wal-Mart's population interacting on Daily Challenge. To build behavior change through technology, you need to understand the DATA LET’S TAKE A LOOK AT SOME DATA BUT BEFORE WE DO IT’S TIME FOR A CHALLENGE!
• #66: Why it matters Social interactions play a big part in our emotional and physical health! Studies show that people with a wide circle of friends live longer, healthier lives than people without friends. And having a circle of friends who you interact with regularly helps stave off dementia. Taking a minute out of your day to connect with someone you like will also provide a boost for the recipient of your message - a real win-win!
• #67: Let’s talk Data. Here’s a snapshot of me as Walkadoo sees me. We have lots of data on Activity level Social Interactions Whether I’m achieving my goals But that’s just the surface
• #68: A product that collects information on social interactions AND on activity collects a lot of data Here’s just a few variable that are collected per user.
• #69: As of Thursday of last week there were over 720 million data points collected in Walkadoo And that was just me. How do you start to make sense of all that data?
• #70: You build tools. This is an internal tool used to understand a user’s activity level over time. This is my data for the last 30 days The redline represents my total steps per day The green line represents my goal for the day The blue line represents what Walkadoo thought I would walk that day On the bottom the red line represents walking steps The green line represents running steps. You can tell a week ago I started getting up in the morning and going for a short run
• #71: Ok so we’ve built a product that has Designed for behavior change Appeals to walkers Open to everyone to participate Uses game mechanics to keep people engaged But does it ACTUALLY WORK?
• #72: To do that we’ll need to run a Randomize Control Trial to measure the effect. Last fall we conducted a trial on a medium sized business involving 388 participants We created two groups Both groups were asked to wear their pedometer for a week for a baseline One group participated in Walkadoo, the other put their pedometer away. 6 Weeks later the control group was asked to wear their pedometer again for another week. We compared the two groups and found that using the pedometer + Walkadoo significantly improved the amount people walked While the study is under embargo while it awaits publication, Wakadoo has proven increase the step count of users enough to have a significant impact on their health.
• #73: What is in store for the future? Wearables and fitness devices are moving fast
• #74: iPhones 5S and higher have the M7 chip embedded in the hardware to collect activity Wearables will expand to be multipurpose Activity tracking will become passive
• #75: Wearable heart rate monitor by FitLinxx Temporary tattoo offers needle-free method to monitor glucose levels The bathroom that is the Dr.’s office
• #76: Questions?
• #77: Questions?