HIStalk Interviews Robert Kahn, MD, Faculty Lead for Population Health, Cincinnati Children’s Hospital

Robert Kahn, MD, MPH is professor of pediatrics, associate director of the Division of General and Community Pediatrics, and faculty lead for population health at Cincinnati Children’s Hospital of Cincinnati, OH.

Tell me about yourself and the program.

I’m a professor of pediatrics at University of Cincinnati and the Cincinnati Children’s Hospital. I’m a general pediatrician, but also with a degree in public health. 

My interests have always been with the broader circles of influence on kids. Not just are they getting the right shots and the right medicine, but obviously kids live in families, families live in communities, and there are a whole lot of other influences that determine how well a child is doing and how they are in their development. 

To that end, in 2010 the hospital developed four county-wide health goals around asthma, injury, infant mortality, and obesity. Because of my interests, they asked me to help co-lead that effort, thinking how does a quaternary care hospital begin to engage more deeply in achieving population health goals? That’s the background to some of these projects that then involve through electronic health record and helping bridge between what a physician would normally do in a day-to-day clinic filled with patients to begin thinking about community and population health.

Can you give a brief background on population health management, particularly that involving the public health issues in children that you mentioned?

Population health management refers typically in two different ways. One is, how is my total panel of patients doing? How can I get a high-level overview of everyone I’ve seen? How are all my patients with obesity or with asthma doing? How should I shift my overall care and allocation of resources?

What we’re doing here in addition to that is thinking, what about all the children we’re not seeing who have asthma? How should we think differently about improving their outcomes, even if they aren’t going to walk in our door? For us in Hamilton County in southwest Ohio, with 180,000 kids zero to 17, we wanted to begin a journey to say, what would it take to improve the health of all kids?

I would say we’re very early in that effort. We started in just a couple of neighborhoods to think about population health outcomes for that neighborhood. For example, in the city of Norwood, which is nested within Cincinnati, we know there are about 800 households with children under four. We wanted to think about what would it take to reduce injuries in homes with those kids. Our head of trauma surgery, who typically spends all his time in the operating room or in the emergency room helping these kids, has helped lead a team to think in a population way about injuries in the city of Norwood.

Do you feel vindicated in a way that you were early on in something that now everyone wants to figure out?

[Laughs] I’m not sure I feel vindicated. I’m excited that more and more people are interested. I’m really excited to think that maybe payment mechanisms and healthcare reform will start bringing financial incentives to do the prevention-oriented work that could help out in the community.

You mentioned that some of your work involves targeting neighborhoods and subsections of neighborhoods. Describe how you use geocoding.

There’s two ways we’ve used geocoding. We use electronic health records and part of that is geocoding. One is around clinical care and one is around population health management. 

In clinical care, we’ve used the electronic health record to help drive key questions about these other influences. We have one of the largest training programs for pediatricians in the country. If we set in front of them a series of questions in electronic health record about the quality of the housing, what school does the child go to and how are they doing there, are they able to make ends meet, what we can do is drive the discussion to these determinants that are outside of the typical exam room or outside of the typical physiology of an individual child. That then leads everyone to say, hey, where does this child live? What is his address? What other resources in the neighborhood we can get to the child? That’s at the clinical level.

At the population health management level, what we can do is take every single asthma admission in the past year. We know the minute they register. We have their address. We can then link that address to a latitude and a longitude, or what people typically call geocoding. We can say hey, that means they live in this census tract or this neighborhood. Then you can begin to look for patterns of where the asthma is particularly high, or patterns of where the injury or prematurity is high. 

The minute you put a dot on a map, it shifts the center of gravity away from just in the exam room in that moment to, my goodness, I didn’t realize I had 15 kids admitted from a 10-block radius, or a 20-block radius. What’s going on there that might lead to such high admissions rates for asthma or a high emergency room visits for injury? Now we’ve gotten to the point where literally on a monthly basis we can chart injury rates, prematurity rates, and asthma admission rates from each of the 70 to 80 neighborhoods in Hamilton County.

How do you draw a box around how far you can go being a hospital-based project? Do you put people on the street or link up with social agencies?

That’s a fantastic, very insightful question. People are really excited about it, but the right question is, where does our mission end and another person’s mission or another organization’s mission start?

This is a frontier time. On the journey, we’re out there trying to figure out what is it we can do, and then how do we catalyze new relationships, new missions, shared missions. 

As an example, I do not see my job as improving housing for children, even if they have asthma. I see my job is to know that mold, cockroaches, water damage, or a negligent landlord are important in exacerbating this child’s asthma. But then I really need to find the agency in the community that has a mission to improve that housing. So to me, it’s about building new partnerships. Staying true to my mission about improving health and delivering healthcare, but doing it in a way that engages other people with complementary missions. 

We work very closely with the Legal Aid Society of Greater Cincinnati. One of the great cases we had is a child with asthma, middle of the summer. The mother came to the doctor with the child. The doctor said, tell me about the child’s housing. The mother said, well, I’ve wanted to put an air conditioner in, it’s 100 degrees outside in Cincinnati in mid-summer, but the landlord told me I’d be evicted if I put an air conditioner in my apartment. It turned out we had had three other cases with the exact same story in the past week, all with different doctors. Because of our relationship with Legal Aid, they asked the really simple question I don’t need to ask, which is, who’s this landlord?

It turned out this landlord owned 19 buildings and was in foreclosure doing no upkeep on any of these buildings. Almost 700 units were going into disrepair. Legal Aid took it on, developed tenant associations, started to work with Fannie Mae and the property management, and ended up with hundreds of thousands of dollars in repairs and new roofs on these buildings. 

To me, the boxes fit together neatly. We did our job about saying this isn’t just about the kid’s lungs, it’s about where he’s living. They took it on to improve the conditions in where they’re living. But it was only because we had tracking systems through the electronic health record to know who these kids were and what their addresses were that then Legal Aid could go ahead and really understand what the pattern of the housing was and what the problem was.

What struck me as admirable in your model is that the hospital didn’t have any way to make money from this and hospitals a lot of times are guided by where the revenue comes from. How do you think hospitals can create a business case for these kinds of public health projects?

Luckily I’m in a place where very senior leadership at the very top has supported this notion and the board itself had endorsed these community-based goals. As our CEO says, our mission is to improve health, not to improve healthcare, or to simply deliver healthcare. It’s to improve health. If this is what it takes, this is what we need to do.

In an era of accountable care organizations in which there would be a global annual payment or a per-member, per-month payment to keep a child healthy, certainly then there’s a financial incentive to move out into the community and figure it out. Then every emergency room visit or an ICU admission for asthma becomes a loss. In that scenario, really beginning to go to the next step where you would say, what would it take? Would it take community health workers on the ground? Would it take hiring paralegals, or simply contracting with these other types of organization that could be effective in the community?

We also have a great collaboration with the Cincinnati public school nurses, who are really trying to think, how do we work hand in glove to help manage these kids? Again, to the extent there’s a huge financial incentive on a per-member, per-month basis to prevent illness, it becomes more and more feasible and desirable to build these relationships.

Where do you see information technology fitting in?

I’d say our approach has been relatively rudimentary. We work off the back end of our electronic health records system. There is a huge challenge because the school system or the pharmacies or the Legal Aid Society all have different technologies. It is not seamless right now and I’m sure it will take a while for it to be seamless, to figure out, how do we have shared responsibility for the patient? How do we share consents and get through some of the privacy issues? How can we track over time? 

My sense is, I haven’t seen that kind of technology developed, certainly between hospitals. There’s a lot more health information exchanges that work between hospitals. There’s a few folks, I think Nemours in Delaware, who have figured out how to get electronic health record look capability to the school nurses. But I think we’re a long way from true interoperability between everyone who might be touching a child or a family in terms of health.

I sometimes compare it to FedEx. If we were FedEx, I would know exactly when the patient showed up at the pharmacy, what time they checked in at school, how the lungs were doing there, and when they were going to come back to me. That level of tracking and monitoring to help the family with the family’s permission would be great to try to get to in the future.

Have you seen tools or thought about tools that would help what you’d like to do?

I’d like to say yes. [laughs] I’m intrigued by some of the new self-monitoring biosensors that are linked to, say, phones and then back to management software. Propeller Health is one example of a company that’s trying to think, how do you move the information from where the family is, where the child is, and bring it back to a central management point? That notion is a pretty huge advance. 

It’s still a long way off from saying, I’m co-managing these patients with the pharmacy, with the school nurse, with the community development corporation who’s thinking about green space in parks for the kids. We’re moving in the right direction, but there’s a lot of integration and a lot of issues to overcome. With the geocoding software, we’ve only scratched the surface, and even that’s not something hospitals typically use in their health analytics.

How would the average academic medical center or their physician practice organization create a model similar to yours?

I would think a health analytics group five years now, whether they’re working in a hospital or they’re working in an accountable care organization infrastructure, would have a geospatial group working with them. With that, they would be understanding where their patients live, what are the key local and regional determinants of health in that region, and then beginning to deploy healthcare resources differently. Being able to almost predict when there would be problems. Even knowing pollution and pollen patterns might be the kind of information that could be brought in, and then more anticipatorily, trying to get medicine out to the community if they know there’s going to be a surge in asthma morbidity.

Will be hard to get hospitals to do more public health outreach work instead of comfortably treating people who show up within their four walls with a complaint?

It’s going to take some time. It’s out of the comfort zone of where most hospitals are right now. Schools of public health and public health departments around the country could help healthcare a lot in trying to move the ball further faster. But I think until there’s a real financial incentive where there’s a big loss involved unless we’re preventing illness, it will be relatively slow-going.

The other caveat would be until we truly demonstrate a significant return on investment by thinking this way, it may also keep the work moving slowly. That’s our goal — to demonstrate we can actually reduce morbidity and cost by developing this kind of a platform.

Is there existing literature of where that’s been done, or are you finding that what you’re doing so far is promising?

We’re working really hard right now thinking about how to prevent prematurity with this kind of an approach. Every time a baby is born at 24 weeks gestation, it’s a $300,000 to $500,000  immediate cost and probably millions over their lifetime. If we can use a place-based strategy to prevent prematurity, we’ll have a much better argument for deploying the resources necessary, like community health workers, to get the job done.

There are various models of community health workers or home remediation, but I don’t think there’s been an integrated set of interventions put together that would really make the argument at the level of a hospital or an insurance company to push this strategy.

Do you have any final thoughts?

I’m excited to keep trying to push the boundaries. I see the electronic health record and geocoding is a way to break down the walls. 

I would just add, I have found tremendous, capable, and highly interested partners in the community who are really excited to have these kinds of partnerships, whether it’s the school nurses or the pharmacies or even Legal Aid. We’re now 10 percent of all Legal Aid’s cases in southwest Ohio because of this progress. It’s almost always a win- win-win — a win for the hospital, a win for the organization, and then a win for families that we can break down these barriers using electronic records and geocoding.