Getty Images

Predictive Analytics Helps Providers Address Total Cost of Care

The clinically integrated network implemented predictive data analytics to create a total cost of care plan, improving care quality.

Many healthcare systems have made creating high-quality and value-based care a top priority. Southwestern Health Resources (SWHR) used clinical data and predictive analytics to put a total cost of care plan in place to provide affordable and quality care to thousands of patients in North Texas and cost-effective care for their employers.

Oftentimes in the healthcare market, purchasers primarily focus on discounts. According to SWHR Senior Executive Officer Andrew Ziskind, discounts are a priority because purchasers search for predictable and short-term gains.

“We fundamentally believe that a focus on the total cost of care is the right place to be and the right thing to be focusing on. What that means is that there are times where you invest more services now to create long-term health and candidly better outcomes and lower total cost of care,” he told HealthITAnalytics.

The total cost of care plan was implemented because it was the right thing to do in terms of best serving their patients, Ziskind noted. Within the Dallas-Fort Worth Market, SWHR is characterized by high utilization.

“Texas is second only to Louisiana in Medicare spending. We have high utilization rates and so the question is do you approach cost of care by trying to lower the cost but still have too much happening, or do you step back and say, ‘We need to focus on appropriate utilization.’ And when you start to look at appropriate utilization, you start pointing much more towards total cost of care.”

Over the past three years, SWHR has limited the average annual increase in healthcare costs of their networks to one-quarter of the region’s average. With the use of data analytics, physicians could become proactive in early intervention efforts  and keep members healthy by preventing conditions that could create expensive healthcare costs.

“When you look about using data to identify which patients are going to need what, if you have access to the clinical information, you have a real advantage. Being able to pick up early indicators. On a population basis, we’re focused much more on addressing rising risk, rather than just wait for people to get sick.”

Using predictive analytic methods, SWHR identified rising and high-risk patients and worked to close gaps in chronic care management. However, different gaps in care management were identified pre-COVID-19 and currently.

“In the pre-COVID world, we know that healthcare needs to change. We have focused far too long on using data to report through the rear-view mirror,” Ziskind explained.

Data needs to start being used to predict which patients need which resources, a system which Ziskind identified as the Rs — getting the right care in the right place at the right time by the right providers for the right costs.

“As the science gets better, we ought to be using our data not just to report backwards, but to determine how best to create healthcare value. Now, one of the real advantages that a provider-sponsored network has is we have clinical data and the claims data,” Ziskind said.

Currently, in the time of COVID-19, many gaps in care are due to inaccessibility or fear of getting sick when going to an in-person doctor’s visit.

“Early on in the COVID experience, people were not seeing their physician, they were not leaving their house, they were not getting screening tests that are important for health and safety like diabetic eye exams and telehealth was growing,” Ziskind continued.

“But telehealth was limited because the sickest, most complex patients weren’t the ones who were comfortable using telehealth, the elderly, the ones with multiple comorbidities.”

To address care gaps, SWHR has created a three-part data and analytics approach. The first part improves upon the organizations’ traditional data work and ensures that the system is working as efficiently as possible. The second is having SWHR move towards data science and predictive analytics. The last part is incorporating artificial intelligence into the process.

Through working with ClosedLoop, an outside vendor specializing in AI, SWHR expanded its artificial intelligence capabilities to improve data analytics.

“It’s very hard for provider organizations to develop the expertise to fully embrace AI. Now, the new thing that we have come to appreciate is data and predictive analytics are only as good as your ability to act on the information. And so, it doesn’t help you if you’ve got all these great predictive models and you either don’t use them or you can’t use them,” Ziskind said.

By incorporating data analytics into care management, providers will see a significant improvement in care quality. Ziskind advocates for fundamentally redesign how care is managed. In previous care management models, providers would wait for people to become sick and then treat them. However, Ziskind explains that physicians should intervene before conditions become severe and harder to treat.

“The future of care management is more data-driven. It’s transforming the way you coordinate care. You can’t do that without good data analytics. I fundamentally believe our problem is not that we don’t have data — we have plenty of data. Our fundamental problem is we’re not acting on the information we have. So that’s the interface between data analytics and clinical operations.”

While COVID-19 imposed significant challenges to the healthcare system, Ziskind said it also enabled rapid transformation in how the system works and functions to deliver care standards.

“We should take this opportunity to learn quickly, use data to do that, and keep focusing on how do we deliver greater value, which I still characterize as value can be improved by improving outcomes or lowering cost or ideally you do both together,” Ziskind concluded.

Next Steps

Dig Deeper on Artificial intelligence in healthcare