Web App Masters: Adventures with Data Visualizations

by May 28, 2011

At the Web App Masters Tour in Seattle, Kate Brigham talked about the lessons PatientsLikeMe has learned from allowing people to share and discover medical information using data visualizations. Here are my notes from her PatientsLikeMe: Adventures with Data Visualizations presentation:

8 out of 10 people go online to look for health information. There’s a lot of good information out there but it is hard to put into context. How does it relate to me? If you have a chronic condition that you need to manage throughout your life, PatientsLikeMe (PLM) can help you tackle long-term issues. On PLM, people share information about their health and learn from people that have the same conditions and symptoms as them.

  • If you are asked how you feel, you usually respond with a story. But there are limitations to seeing stories in text. PLM creates real data-rich profiles to help tell stories with more than text. This creates a different kind of understanding and helps get people to answers.
  • Translate stories into data: this helps us process information, find patterns, and locate similarities, outliers, trends, etc.
  • PLM’s objective is that every patient on the site should be able to answer: “Given my status, what is the best outcome I can hope to achieve and how do I get there?” This question informs the company’s thinking and the solutions they evaluate.

Make it easy for people to create data

  • Most of us think in stories, not in Excel.
  • Give people questions that are easy to answer. Ask in clear simple language. Pair it down to basic questions.
  • For long forms, aim for simple points of interaction: big colorful buttons that have meaningful values. This can help visualize a pattern of answers and communicate better than words.
  • Ask as few questions as possible to get the information you need.
  • To ensure better accuracy in inputs, use inline validation (instant search results) to make sure appropriate selections/formats of data.
  • In addition to long-form surveys that detail conditions, PLM supports simple inputs about how people are feeling. This keeps re-engagement up and is a nice stepping-stone for more comprehensive data gathering.

Highlight benefits of sharing data

  • If you are asking people to put time into something, you have to tell them the benefits of taking the time to enter their data.
  • You can do searches on PLM without being a member. This allows you to find out what value exists on the site before you sign up.
  • Give something... get something. For every piece of data you collect, give something in return. On PLM, you can see where you fit in, how many other patients there are like you on the site. The site always reflects something back to you.
  • PLM has a simple quick start guide that focuses on reflecting answers back to people when they provide information about themselves.
  • Results are always shown in context for everyone with their specific condition and treatment highlighted within the broader set of results.
  • Sharing information in a community makes you part of the bigger story. People can learn from what you have experienced.

Help people use data to answer questions

  • Knowing your status is good, seeing yourself in context helps inform decisions, being part of something bigger is great... but ultimately you have questions to which you need answers.
  • Data visualizations can help paint the picture of the effect of new symptoms, and treatments on people. On PLM you can get answers to questions about recommend dosages, new treatments, and more.
  • The amount of data you share is your profile & credibility on the site. Your data is your reputation.
  • Charts are not warm and fuzzy but they can create connections between people. Seeing what’s happening with others can create bonds and opportunities for conversation.

Don’t be afraid to experiment

  • Wow week at PLM is once a month to focus on projects that people are interested in. This gives the team time to explore side projects and see what works.
  • You sometimes have to take data and play with it. Many times you just don’t know how things are related.
  • Experiments help you think about where to go next.