Making Every Data Point Count: A Journey to Health Equity
Based on a TED-style talk given at NACCHO 2021 by the author
A paradigm shift is upon us.
Better health, equity, and community readiness NEEDS better insights.
But while there is more data at our fingertips now than ever before, many decision-makers still feel utterly data starved!
So the paradigm shift is this:
We don’t need MORE data.
But we do need to get more FROM our data.
And I’m going to tell you how we’re going to do this.
I’m Stefany Goradia, independent health data SME, and I’m here today to talk about how we can take what we learned during the pandemic to up our data game to the next level and rise to the occasion.
Because there is no denying it. COVID put a spotlight on major gaps in our communities — exacerbating some that we already knew existed and revealing others we didn’t even know about yet.
There are gaps in our data, our policies, our emergency response, and within our most vulnerable communities — just to name a few.
I’ve been working in healthcare data analytics for over a decade. And I can tell you that the data, and data methods, that we have been using are just not robust enough to really understand and get at the root of these gaps, let alone inform how we can truly address them effectively.
And these are gaps that we NEED to understand to move forward.
My Life in Data & How My Story Could Have Ended Much Differently
Let me tell you why this is so important to me personally and then I’ll tell you how we’re going to get there…
I was an eager little learner. Top of my class. Straight A’s, always the first to volunteer or ask questions — in fact I still am that girl.
My family is decidedly middle-class. My parents weren’t exactly around when I got home from school; we didn’t have a lot of experiential vacations and active weekend activities. In fact I have always been quite literally last in gym class.
I grew withdrawn in my teens and after skipping out on most of my 8th and 9th grade years, I ended up dropping out at 16. I was working at Taco Bell then, with a pretty bleak future ahead.
I wish I could say something amazing happened to motivate me back to school, but it didn’t. Through happenstance, I did end up going back, after changing majors quite a few times, I graduated — with $90,000 in student loan debt — landed in healthcare, curiosity led to starting a small company, and that company was acquired by RS21.
But, my story could have easily ended much differently.
And I share this story now, as an example of how easily people like me can fall through the cracks, even though all the signs — and what I understand now to be DATA — were there as early as age 5.
My datapoints are a few of millions of datapoints that — if leveraged together — we can use to be more proactive, deploy targeted and meaningful programs with precision, and even predict in real-time the risks of an individual or community.
Asking Better Questions for Deeper Insights
So how do we do it?
It starts with asking questions.
I like to follow the WHO WHAT WHERE WHEN and WHY to inform the HOW. Let’s use the risk of poor COVID outcomes as an example.
Sure we might think we know anecdotally where the worst areas of town with the highest rates of crime or poorest health — but do we really know them?
- Within one Census block, do we really know the differences between the communities of houses that are right next to 3 big apartment complexes? Because that impacts their behaviors and outcomes.
- Do we know how often the people in these neighborhoods go to restaurants, grocery stores, or church, and which ones? Because that impacts their behaviors and outcomes.
Understanding and asking questions like these is how we’re going to get more from our data.
I’ll share how I’ve seen this concept put into action in the past.
A public health-focused medical group received a Centers for Disease Control (CDC) grant to improve COVID vaccination rates in Laredo, TX, specifically among migrant, undocumented, underserved communities, and the vaccine-hesitant — a very specific population.
They wanted to provide “pop-up” mobile vaccination clinics to reach these community members, and they needed to know where to go. Now, if you look at Laredo on a map of adverse social determinants the entire region along the border lights up.
But resources were finite, and you can’t really go door-to-door across all of Laredo’s vulnerable communities.
They used data to help them identify WHO — were the most vulnerable community members based on WHAT — unique set of characteristics the CDC were interested in.
In doing this, two specific neighborhoods emerged — one in the North, and one South.
At surface level, they appeared to be equally “high risk” neighborhoods. But further analysis revealed:
- The North neighborhood was closer to the border crossing, had poorer health overall, and tended to be an older age demographic.
- The South neighborhood had slightly better health on average, but greater social vulnerabilities (e.g., crime, poverty, access to care), closer to the Colonias, which is basically one step up from a homeless camp, rented by slumlords primarily to the undocumented.
Those were the people we wanted to reach the most.
But they didn’t stop there. They then used cellphone movement data and GIS analysis to understand exactly WHERE the residents of that South neighborhood tended to travel within the city, WHEN and WHY they visited different areas, by day and time, so clinicians knew how much traffic to expect and if they were in an accessible location.
From this, they identified 3 specific locations to deploy the pop-up clinics, down to the street corner and time of day.
Getting More From Your Data
I’m lucky that I get to use data and solve these kinds of problems every day through my healthcare data consulting with very passionate clients.
I’ve always been fascinated at how much insight is waiting when you bring together very complex and disparate data sources, merge it, and then what we can DO with that data using data science to find new insights.
There are 3 steps to getting more from your data.
1. It starts with curated data that is easy to exchange, ingest, and use.
Having curated data sets — like community risk factors, social determinants of health, and your own internal data — takes out 80% of the challenge of data analysis.
Having the data allows you to move so much faster that you can begin to get insights in days or weeks rather than months or years.
But it’s not enough to just have the curated data.
2. The data must be unique, usable, and novel.
Using novel data shows us so much more, revealing new insights because the data never existed together before.
There has been a lot of attention lately on geospatial and location intelligence data to understand how people move through and utilize their community resources.
This kind of anonymized data can be paired with other data sets, as described in the “pop-up” clinic example, to support efforts such as:
- Discovering when and where to reach the most people for improved public health outreach.
- Determining how far people have to travel for important amenities, like healthcare and food.
- Identifying travel patterns that indicate where congestion could bottleneck or impede emergency evacuations or responses.
3. Use data science to extract more insight.
This step is a bigger reach because data scientists are in high demand, but they are in high demand for a reason: these techniques provide extremely valuable insights that a typical analysis can’t deliver.
There is a myriad of new and exciting methods and techniques being pioneered by data science companies and practitioners today.
Even without fancy data science, there are many tried-and-true methods that are underutilized in healthcare analytics, in my experience. Things like Process Behavior Charts/SPC charts, Monte Carlo simulations, discrete/stochastic optimization, and more.
These new models and methods help us predict, visualize, and tap into new potential from the data we have at our fingertips. Even if you don’t have a data science team in-house, there are organizations that can provide this level of expertise to support your team.
We don’t need MORE data!
But we do need to get more FROM our data.
And we need to start doing it RIGHT NOW.
We’re on the right path, but we’ll need to work together to take it to the next level.
So… how will you get more from your data?
Like my content?
Follow me on Medium, LinkedIn, and Twitter.
Want to master healthcare, data, and healthtech?
Subscribe to my weekly newsletter, Health Data Guru. Every Sunday evening, you’ll get 1 actionable idea about using health data for change: https://weekly.healthdata.guru
Want to work together?
I work with healthtech startups, investors, and other health organizations who want to transform healthcare and achieve more tangible, equitable outcomes by using data in new ways. Learn more at: stefanygoradia.com/healtheq
Stefany Goradia is a health data technical expert, independent consultant, and facilitator.
She has spent her career “in the trenches” of healthcare analytics and delivering insights to internal and external customers across the entire healthcare value ecosystem. She now teaches others to use health data in new ways and develop equity initiatives that matter and speaks/writes about how to: interpret healthcare data, communicate it to stakeholders, and use it to support strategic decision-making and program execution. Learn more at stefanygoradia.com/about
