By Meridith Peratikos, Director of Enterprise Research
In January, researchers and technologists across a wide range of disciplines and job sectors converged at the RStudio::conf 2019 in Austin, Texas to celebrate and learn about R, a programming language for statistical computing and graphics, as well as discuss RStudio tools that make using R for business effortless. I was lucky enough to attend the conference with two colleagues- Amy Graves from our Medical Economics team and Chad You from our Data Science team. Although we all work on different teams, each of us was impressed by the depth of content discussed in the workshops, inspiring messages offered by keynotes speakers, and relevancy of content across the presentations.
I worked with R well before RStudio released their beta IDE (Integrated Development Environment), so I guess that makes me an ‘R veteran.’ In my role as Director of Enterprise Research at axialHealthcare, I’m responsible for both strategic planning and execution of high-impact research projects. With efficiency improvements crucial to my daily work, I relished in the opportunity to learn from practitioners across various disciplines at the conference. While there, I challenged myself to consider how I can apply these learnings to my daily work, which is focused on improving the lives of patients in pain and those suffering from opioid use disorder.
Demonstrating Value
Discussing the use of R to reduce employee attrition, Matt Dancho of Business Science discussed the usefulness of R for analyzing data, building models, translating findings, and having a notable impact on employee attrition. He demonstrated a model that included predictors of employee attrition (e.g. continuing education, years of service, business sector) and translated the impact of the model into dollars saved. This is an important takeaway for data scientists working on business-facing solutions: don’t get get hyper-focused on the accuracy of the prediction and ignore the demonstrated value.
Primary Data Collection
Hilary Parker, Data Scientist on the styling recommendations team at Stitch Fix, discussed how design thinking has impacted the ability of their data scientists to better inform individualized product recommendations. They were able to design data collection tools to map individual members, enabling her team to style profiles more quickly than if they were to wait for an individual to make a purchase.
This year, axialHealthcare will launch initiatives to deepen patient-level intelligence with the aim of improving access to quality pain care and addiction treatment. This recognized parallel between these two initiatives prompted me to consider whether health plan members would be willing to give direct feedback if they knew that the data would be used to inform individualized treatment plans. Thanks to this session, I’m adding a few books to my reading list, including “Design Thinking: Understanding How Designers Think and Work” by Nigel Cross.
Automating Work
Throughout the conference, speakers highlighted new and existing features in RStudio’s professional product line. Jeff Allen, Director of Engineering at RStudio, spoke about RStudio Connect, which is a platform we use at axialHealthcare to publish interactive applications (called Shiny apps) and research reports (produced using Rmarkdown). Some of the newer tools discussed will allow for better automation and improve visibility into customer usage. For example, I hope to use new features in RStudio Connect 1.7 to 1) automate emails of routine reports, and 2) demonstrate the value of our published Shiny apps and Rmarkdown reports by analyzing historical event data.