Carolinas HealthCare’s Investment in Analytics for Community Health

Sept. 10, 2014
Clinician and other leaders at the 42-hospital Carolinas HealthCare System are moving ahead to leverage analytics capabilities to support community health and population health efforts system-wide

Exciting things are happening at Carolinas HealthCare System. The Charlotte-based health system, which encompasses 42 hospitals and 900 care locations, has been making significant investments in data and analytics, to support rapidly expanding efforts in population health and community health across North and South Carolina. Among the senior executives leading important efforts are Michael Dulin, M.D., chief clinical officer for analytics and outcomes research, and Allen Naidoo, Ph.D., vice president for advanced analytics, for the system. Dr. Dulin, who has been with Carolinas for 16 years, has held the chief clinical officer for analytics and outcomes research position since late 2013, and continues to practice family medicine part-time. Dr. Naidoo, a biostatistician, has been with the system for two years. Among other activities, they’ve helped to launch the Dickson Advanced Analytics Department (DA2), which is housing some of the key activity that is facilitating analytics and community health progress within the system.

Drs. Naidoo and Dulin spoke recently with HCI Editor-in-Chief Mark Hagland regarding their current initiatives in analytics and community health. Below are excerpts from that interview.

You’ve been doing some exciting things with analytics and population health recently. Can you share some of the recent developments in your organization?

Michael Dulin, M.D.: We’ve been making significant investments in terms of our data and analytics; we’ve built a single center for data and analytics, which Allen and I run. And we’re involved in a number of initiatives. One of the  big ones is around community health—how do we bring the data in, and then use the results to improve community health? And we have funding from the National Institute for Minority Health and Disparities, or NIMHD; NIMHD is part of the NIH [National Institutes for Health]. It’s about $250,000 a year for five years. We’re using the data capabilities we’ve built out to hotspot in the community—to engage with members of the community to co-develop interventions to improve community services and primary care. And asthma—we’ve risk-stratified our population for asthma. We identified areas in the community at high risk, and patients at high risk, for problems with their asthma. Then we’ve built interventions, to improve outcomes for asthma.

Michael Dulin, M.D.

What kinds of interventions?

The biggest one is for shared decision-making for care for asthma.

Do those efforts involve patient engagement in the care process?

Yes, patient engagement in building a care plan, and then, building an electronic asthma action plan, a tool to provide decision support for the providers that helps us use the NIH guidelines for asthma care.

That’s built into your EHR [electronic health record], then, as clinical decision support?

Yes. We launched it in August and just finished the evaluation now, which documented that that tool by itself reduced emergency department visits and hospitalizations for asthma, by nearly 50 percent, in eight months.

What was the key element on that program that has created that result?

Part of it was engaging the clinical team and even patients, in the development of the tool. We brought in family doctors, pediatricians, ED doctors, nurses, and respiratory specialists, and patients. We built an Internet tool, a web tool, that allowed us to iterate it and get feedback and improve it over a short period of time. As you know, changing things within the EHR is very hard; so we built this as a pilot, and once we got that feedback, we put it into the EMR. We had it on the web for about a year before we launched it in the EMR.

You can see the shared decision-making tool and the electronic asthma action plan; but you can see the different tools and see the asthma action plan, and it will show you how it works. It takes a 140-page NIH guideline and turns it into something you can use in a few minutes. We had found that providers couldn’t really use the guideline because it was so complex. So this has helped us improve care delivery.

What have been the biggest challenges and opportunities in population health?

Allen Naidoo, Ph.D.:  A couple of the big challenges from a data perspective have really been around the integration of data. The data is in many locations. So how do we amass all of the data into a centralized repository, so we can do the analytics we need to do for population health? So we continue to integrate data from multiple different sources. One of our biggest challenges, though, is around patient identifiers from our different source systems, and from other data sources that come from outside the organization.

Allen Naidoo, Ph.D.

From a people perspective, one of our other big challenges is around hiring the data scientist talent and the analytical tealent to really be able to analyze the data successfully. There's a tremendous shortage of those skill sets nationally.

What have been the biggest lessons learned so far?

Naidoo: I think the biggest lesson learned is that prior to creating this new infrastructure, work was being done in a very siloed fashion across the organization, but now that we have committed to actually consolidating analytics in one area, enterprise-wide, things have worked much more efficiently. Focusing on an effort as big as this from a centralized perspective is really key. The other lesson is that in general, people produce analytics, and they say, wow, that’s nice, or interesting. But we’ve found that when we engage with the people to whom we’re delivering, the physicians and care managers, and we show them how they can use that information to better care for their patients and demonstrate value, that’s far more valuable to the system as a whole. So it’s not just about producing the analytics, but consuming the analytics as well, for the betterment of the patients.

A good example of that is the readmissions work we’ve been doing. We just implemented it last year, and in fact, this quarter [the second quarter of 2014] will be our first measurement period, so we don’t have the results yet to publish, but we will have results by the third quarter of this year. This is around all-cause readmissions. We predict the risk when a patient is admitted, of readmission, while they’re’ still in the hospital, and we do that in real time, and we can what the factors are that are driving risk for the patient, and we offer customized support for that patient while they’re hospitalized. Some of that is education for the patient or the caregiver, before they leave our facilities. And the interventions might be guiding them to follow up with a primary care physician, or making sure they’re compliant with medications.

Lastly, what would your advice be for other healthcare leaders, especially CIOs and CMIOs, around all this?

Naidoo: First, if you want to do something to scale, you need a high level of organizational alignment around this, and you need senior-level commitment, and you need to be able to invest for success.

Dulin: I think Allen covered it very well. The investments are significant; not just the technology, but a lot of process-oriented change that is required. So you need high-level support and willingness to make investments, because those investments take time before they pay off. We’ve done this for quite a while—the group we’re in actually launched in 2010, when we launched the Dickson Advanced Analytics Department (DA2), but we had planned starting in 2008. Named after a former board member of the hospital; and we call it DA-Squared.

Sponsored Recommendations

Healthcare Rankings Report

Adapting in Healthcare: Key Insights and Strategies from Leading Systems As healthcare marketers navigate changes in a volatile industry, they know one thing is certain: we've...

Healthcare Reputation Industry Trends

Navigating the Tipping Point: Strategies for Reputation Management in a Volatile Healthcare Environment As healthcare marketers navigate changes in a volatile industry, they can...

Clinical Evaluation: An AI Assistant for Primary Care

The AAFP's clinical evaluation offers a detailed analysis of how an innovative AI solution can help relieve physicians' administrative burden and aid them in improving health ...

From Chaos to Clarity: How AI Is Making Sense of Clinical Documentation

From Chaos to Clarity dives deep into how AI Is making sense of disorganized patient data and turning it into evidence-based diagnosis suggestions that physicians can trust, leading...