PROVIDENCE – It was a glorious, late summer morning when ConvergenceRI sat down to talk with Joel Stewart, MPH, the newly hired health data analyst with The Providence Plan.

As one of his first tasks, Stewart is working on developing a potential new approach to data around lead poisoning in Rhode Island, using a predictive risk analytics model developed at the University of Chicago.

The model uses existing data to predict the risk of a child being poisoned by lead so that an intervention can take place before that happens. It mines some two decades’ worth of existing data from Chicago, including blood lead level tests, home lead inspections, property value assessments and census data. The predictive risk model then allows inspectors from the Chicago Department of Health to prioritize houses and identify children who are at the highest risk.

“If you go in and fix lead hazards, you are going to protect kids in the future,” explained Rayid Ghani, director of Data Science for Social Good, a fellowship at the University of Chicago funded by Google Chairman Eric Schmidt and his wife, in a recent story in The Chicago Tribune. “It would be far better if we could protect the kids living in those homes now.”

The Providence Plan, in consultation with the folks in Chicago, is now in dialogue to see how that open source predictive database might be adapted as a data tool to combat lead poisoning in Rhode Island. And, if it proves to be successful in its prescriptive capability to address the issue of lead poisoning, could it also be used to address issues such as asthma or, more broadly, toxic stress.

The conversation with Stewart occurred in the garden oasis that is Olga’s Cup + Saucer on Point Street, one of the nodal points in the innovation ecosystem network in Rhode Island.

[On this particular morning, the conversation was interrupted not by fellow conversationalists, but by a steady rumble of trucks, because Point Street now serves as a primary connector between the city and the highway, a very real data point that probably has not yet been analyzed in discussions of traffic patterns for the proposed stadium. The noise pollution often forced the conversation to pause, creating gaps, as trucks rumbled by, interrupting the flow.]

Stewart called the work at The Providence Plan to be about “democratizing” data, configuring data to tell stories around health – putting together the visual mapping that shares data across any number of disciplines. “Data is one of the ways in which all of us can speak the same language,” he said.

The work around prescriptive risk analytics, Steward continued, connects everyone involved “in the pathways between information and enforcement.”

It changes the game plan in preventing childhood lead poisoning, Stewart added, “from reacting to anticipating the risk.”

Here then, is the interview by ConvergenceRI with Joel Stewart of The Providence Plan.

ConvergenceRI: What attracted you to working at The Providence Plan?
STEWART: I [earned] a Masters in Public Health at Brown University. I had often used [in my studies] a lot of The Providence Plan’s available data. I had been familiar with their work for a long time. A few months ago, they had an opening for a health data analyst there.

Their philosophy is about democratizing data, about making information accessible to individuals, community stakeholders, politicians, whomever. I think that it’s highly important.

ConvergenceRI: What in particular are you working on?
STEWART: My portfolio is mostly comprised of any health related topics. For instance, I’m working on the continuation of the lead technical study; you’re probably familiar with the work we’ve done previously.

[In that work], we’re looking at newer data, more recent data, [focusing] in on smaller geographies and different outcomes.

ConvergenceRI: I’ve heard that The Providence Plan is exploring the potential of adapting a new predictive risk analytical data model that’s being used in Chicago. Are you involved with that?
STEWART: Yes. The researchers in Chicago have developed a model that uses a lot of the same data that we have, and they are able to generate a risk score, looking at the risk of an individual being poisoned with lead.

And so, we’re kind of shopping around the idea, to see who’s interested, and evaluating how valid it would be if it were used here in Rhode Island.

ConvergenceRI: My understanding is that it is a predictive model, through which you would be able to identify which housing in specific localities was most at risk? Is that accurate?
STEWART: Yes, it’s changing the game plan from reacting to anticipating the risk of lead poisoning.

ConvergenceRI: Is the model open source?
STEWART: Yes, the data model is open source. We have access to the model and how it works. We’ve been in contact with [the researchers] in Chicago; we plan to talk to them again.

ConvergenceRI: Beyond lead poisoning, are there opportunities to use this as a risk predictor model for things such as asthma?
STEWART: That’s the hope. We are hoping that, in the future, we can use this to look at other health outcomes, like asthma, or for any sort of environmental problem.

ConvergenceRI: Who are the buyers for this?
STEWART: What do you mean, buyers?

ConvergenceRI: You had talked earlier about shopping the model. Who are the people or agencies that are going to benefit from this? The audiences? If you can predict those housing units in Rhode Island that are going to produce childhood lead poisoning, it would seem to me that those priorities would become the first priority for renovation.
STEWART: We just started talking about this with others. We met recently with some people at the R.I. Department of Health. Their resources are very limited, so they are very interested.

Having a better picture of where lead is going to be problematic allows them to redirect their limited resources. Right now, people at the health department are very excited about it.

ConvergenceRI: One of the fascinating outcomes about the way in which The Providence Plan is telling data stories is how you can engage with communities, the way that the data becomes visual, making the data accessible. Can you talk about why that’s important in today’s world?
STEWART: I feel as if we have a disparity of resources in a lot of different ways. One of these resources is information. Because information allows us to make the right decisions and it fuels action. As technology becomes more improved, as we are able to use many different data sources to tell stories, allowing people to know what’s good for them, and what’s dangerous. It’s allowing them to make decisions about their own health and habits. I find it very exciting.

ConvergenceRI: In the digital age we live in, the way that the flow of information often works best is when it is shared across networks, creating engaged communities. Could you talk about the role that data and information plays in the potential of creating an engaged community? How are the traditional definitions of neighborhood and community changing?
STEWART: That’s a good question; actually, there are many questions there. A lot of what we do, with a lot of the technologies that we implement, I feel that they are able to engage communities at various levels. For instance, we do a lot of mapping using GIS technologies. Incorporating space into our analyses allows people to directly make this information or mapping more relevant to themselves [and their lives].

I see [sharing] our information, our data stories, as efforts to democratize data.

ConvergenceRI: You used that term earlier. What do you mean by democratizing data?
STEWART: Making it accessible at various levels, making it accessible to individuals, so that regular, every-day people that want to can gain access to the information.

It can also free up the information for someone who runs a hospital, for someone who changes laws, for someone who runs a school, whomever; there are numerous groups of stakeholders.

ConvergenceRI: Have you wrestled with how you define community beyond traditional boundaries?
STEWART: I don’t know if we’ve figured it out yet; I think it’s a very complex issue.

ConvergenceRI: What makes it such a complex issue?
STEWART: When I was in the MPH program at Brown, something I did when I was learning how to do qualitative data analysis, I interviewed people about how they perceived their neighborhood environment. A component of this was asking: what is a neighborhood to you? What is a community to you?

And, some individuals said: well, the city of Providence is my neighborhood. It starts here, and ends here, and that’s how they define it.

Other individuals would suggest that there was a different sort of, almost a psychological boundary, where they frequented, or where they felt safe. Things that had a positive affinity for them were part of their neighborhood, whereas things they didn’t like as much, they didn’t see those things as part of their neighborhood.

It’s an interesting idea, trying to figure out what a neighborhood is.

A neighborhood exists at a [geographic and political] level. It’s a place, for instance, where the Armory neighborhood begins and ends. But also, for an individual, it might be something that expands across the whole city, or it’s very limited to their block.

ConvergenceRI: Does it also have to do with how and where you come into contact with people, with people that you see as neighbors, and where those nodal points are?
STEWART: Yes.

ConvergenceRI: Is Olga’s a neighborhood?
STEWART: It could be. Olga’s in the scheme of things could be considered part of an individual’s neighborhood?

ConvergenceRI: In health IT, the dynamics of information flow, and the way in which information is being analyzed is changing. Hospitals and health systems are redefining themselves as accountable care entities, in order to be reimbursed across a continuum of care through population health analytics.
At the same time, there is a explosion in mobile health and wearable devices. People are self-reporting, in a way that often excludes the traditional health care delivery system. The market forces seem poised to collide.
At The Providence Center, have you thought about health data in the new age of mHealth?
STEWART: We have not completely thought about this. I think it’s something that needs to be digested a little more, as a society, before we start implementing sets of data. It will be interesting to see how it takes off; and, how we can use it.

ConvergenceRI: With The Providence Plan model of data stories, you are changing the way that data is being used and understood in health care. Will mHealth create new opportunities? What are the politics involved in sharing this kind of data?
STEWART: You are going to encounter a lot of obstacles in sharing this kind of information. It’s where I walk, how much I walk, how fast my heart rate is; these are data sets that many people regard as very private information.

It’s going to be interesting to see how this becomes integrated systematically.

ConvergenceRI: Getting back to lead, if you can identify the properties that create the most risk for children being poisoned by lead, who then needs to be at the table to change the equation?
STEWART: We need to get the various agencies that enforce compliance and healthy housing standards – everybody involved in the pathways between information and enforcement. I’m pretty new to this and these projects. I don’t know the whole picture or all the players involved.

But the goal is to get everybody involved.

ConvergenceRI: Beyond the interventions to prevent lead poisoning, are you working on other data sets to track what happens once children are poisoned by lead?
STEWART: Yes, that’s something that Anna Aizer is developing as part the second lead technical study, looking at what the health, behavioral, educational and economic outcomes.

ConvergenceRI: Does data sharing provide a way in which to help get people to talk across systems? In my work setting up the conversation/convergence on toxic stress at Rhode Island College, one of the reasons for creating the event is the difficulty in breaking down silos and the sharing of ideas.
STEWART: Yes, data is one way in which our efforts can be combined. Data is one of the ways in which all of us can speak the same language.

ConvergenceRI: The dream of a common language?
STEWART: Yes. We’re becoming a more data-driven society. Technology systems are making it easier and easier to find and share different data sets.

ConvergenceRI: Is there an overarching problem about who will control the data? Is transparency one of the fundamental principles that needs to be attached to data sharing?
STEWART: I think so. There are obstacles we face just trying to link the data between agencies. Trying to get different agencies to share their data, sometimes it takes a really long time.

We’re still early in the stages of this new, data-driven world. I think that there are a lot of people and a lot of groups that are still wary about sharing information.

ConvergenceRI: Why is that? What is behind the resistance to sharing data, and sharing information?
STEWART: Because, we don’t know what the outcomes are, necessarily, when we share data. What’s going to happen after I give up all this information?

I think that’s where we are as a society, and it’s going to take some experience with it, to make people more open to it, to become trustworthy.