Healthcare Analytics: Understanding Risk and Value at the Patient-Level

By Christian Nimsch and Graham Hughes, MD

The change is underway. The healthcare ecosystem is officially shifting from volume- to outcome-based reimbursement. With so much at stake, risk-bearing provider organizations are well aware of the importance of “getting it right” and “doing it well.” But to hit the mark, analytics are critical to success.

To fully balance patient outcomes with financial incentives derived from value-based payment models, providers need to re-focus their workforce and re-engineer their processes. The goal is to improve coordination of care to yield improved patient outcomes while at the same time, controlling costs. One of the inputs to reaching this goal is to understand risk at the patient-level.

By gaining a deeper, data-driven understanding of individual patient risk, care teams can understand, manage, and possibly prevent the causal drivers of that risk, tailor treatment options to an individual’s specific needs, and target interventions across the care continuum. The combination of the availability of ever-increasing patient-level data and the technology to gather, cleanse, analyze, and act upon this data is a necessity in understanding and predicting risk. When you think about it, from electronic health records, claims data, health information exchanges, digital imaging, streaming device data, and prescription data to physician notes, test results, and vital signs, the contributive potential of patient-level data is enormous and expanding at a daunting rate. This escalation in data volume, coupled with technological advances (think storage and speed), provides opportunities and challenges for enterprise analytics to act as catalyst for transformation into patient-centered care that balances both clinical and operational considerations.

Reconstructing both clinical and business operating models requires trustworthy, useful information to formulate strategy and make decisions about resource allocation, interventional investments, and risk-sharing negotiations. In this context it’s easy to understand why there is so much excitement about the potential for big data analytics in healthcare. The days of limited, retrospective reporting of events are over. True enterprise clinical and operational analytics deliver the scalable, predictive, and prescriptive means to mitigate risk and enable clinicians to optimize interventions and improve expected outcomes.

Today, most industries have successfully gained quantifiable value from big data analytics. In order to achieve comparable value in healthcare, a full spectrum of analytic capabilities—from descriptive to predictive to prescriptive—can be invoked to find new opportunities.

Capitalizing on these opportunities requires a clear multiyear roadmap that integrates some fundamental initiatives for making progress: 


Secure access to comprehensive patient data across the care continuum

Data builds the foundation for taking risk and measuring outcomes. Patient-level data may always be stored in disparate systems and formats across multiple settings of care and stakeholders, but this doesn’t have to be a hindrance. Creating a comprehensive view of patient-level data is likely the most difficult step in achieving enterprise analytics because it requires a willingness and ability of healthcare stakeholders to share data, i.e., a culture change. Add to this that data sharing across stakeholders can be encumbered by legacy reimbursement structures that compete with the need for data transparency under pay-for-performance. The good news is that the trend is moving toward transparency, helping to solve these data challenges.

Analytical tools and talent to transform data into knowledge

Big data in healthcare is often called the new gold rush. Data without access to scalable IT and analytic capabilities is like having a gold mine without access to the mining tools to get the gold. Analytics (at scale) requires access to scalable IT and analytics as a core competency. However, even when organizations succeed in integrating and preparing data for analytics, many organizations struggle to achieve scale in IT and analytic capabilities to derive key insights. It’s an area where the importance of a keen focus should not be underestimated.

Agility to disseminate and take action on analytical insights

Taking action on data-driven insights requires stakeholder alignment. Historically such alignment has been rare with healthcare providers, but is necessary in the new data and analytics- enabled arena. To succeed, this effort requires organizational alignment among clinical, IT, business management, caregivers, and patients—again, a cultural shift to secure the potential that insights can deliver.

The bottom line is that each of these initiatives requires an enterprise intelligence strategy using traditional and advanced analytic tools based on the right infrastructure to ingest, store, secure, integrate, transform, visualize, explore, and analyze patient data. Moreover, advanced health analytics requires a rare combination of clinical, business, and statistical expertise. As such, an extensible platform that leverages the skills of medical informaticists and data scientists to generate and disseminate analytical insights to care teams and business leadership—across any channel or device—is a necessary investment.

These efforts also require a powerful array of data integration, data quality, predictive modeling, optimization, forecasting, and simulation techniques to hone in on the interventions that will lead to the best results, while operating within resource limitations and other relevant constraints. For risk-bearing organizations to overcome their present challenges, scarce resources must be prioritized for the greatest impact, and a variety of accountable care risks managed, to deliver on the promise of value-based care. This all requires an unprecedented culture change across groups for stakeholders to align in their shared, overarching purpose.

This is new territory for many organizations and they should both look to the pioneers who are paving the way, as well as learn lessons from best practices adopted in other industries. It would be wise to recruit talent from areas outside of healthcare to seed leading edge thinking around how to develop and deliver an enterprise-grade analytics strategy. Connecting these outside resources of knowledge capital to key business and IT leaders in your organization will help to drive transformation.

Christian Nimsch is chief strategist for the SAS Center for Health Analytics and Insights where he focuses on innovation, collaboration, and thought leadership aimed at developing advanced analytics solutions for healthcare and life science. Prior to joining SAS, Nimsch worked at Quintiles and supported patient and provider engagement initiatives that combined market insights, education, and multichannel interventions. He has also managed teams of economists, programmers, and consulting actuaries to develop and support data-driven decision support tools. Nimsch has been published in peer-reviewed journals in the fields of public health, behavioral research, health economics, and multichannel marketing. He may be contacted at

Graham Hughes, MD serves as chief medical officer at the SAS Center for Health Analytics and Insights. He joined SAS in 2011, bringing more than 20 years of experience in developing and delivering innovative healthcare information technology products and services. Prior to joining SAS, Hughes spent six years as vice president of product strategy and chief medical informatics officer at GE Healthcare IT, leading a customer-facing advanced technologies innovation team, as well as spearheading the annual strategic planning process. He was the primary physician leader driving GE’s knowledge platform strategy and associated products in collaboration with Intermountain Healthcare and Mayo Clinic. He may be contacted at

Holding the Line on Health Costs

Before Ohio Medicaid officials required HMOs to focus on improving care for the most expensive patients, CareSource was doing just that.

Bob Gladden’s CareSource team was already saving the state money by identifying Medicaid recipients who visited emergency rooms for routine problems and filled multiple, duplicative prescriptions. “Their use was not only expensive, it was potentially ineffective and even dangerous,” said Gladden, vice president of decision support and informatics. The HMO works to redirect the patients to primary physician offices, where they can get better and more cost-effective care.

It’s not just an issue with Medicaid patients. The Agency for Healthcare Research and Quality reports that 5% of the population accounts for almost half the total of healthcare expenses, with 1% accounting for a fifth of expenditures.

So when Ohio mandated that Medicaid HMOs identify the top 1% of most expensive patients and find a better way to manage their care, Gladden’s team not only met the challenge, it took it a step further. Rather than just pulling a list of recipients with the highest bills in the past 12 months, Gladden identifies individuals with long-term chronic health problems and uses data to predict future high-cost patients. Then CareSource assigns nurses to help them manage their health better, including finding a medical home with a nearby physician. The goal is to save money and keep people out of the hospital for deadly complications related to treatable illnesses like diabetes and high blood pressure. CareSource has nearly one million Medicaid recipients enrolled in its program, 40% of the state’s total. It processes 2.5 million claims a month.

The efforts are working: High-risk patients’ hospital bills dropped by an average of $1,600 per patient. With close to one million members, that adds up quickly. In addition, the emergency department utilization rate dropped from 1.5 visits a year to 1.1 visits. Inpatient hospitalization utilization dropped from 0.5 visits per year to 0.4 visits per year for this high-risk population.

The HMO is finding that wellness and case management programs can’t be separated. It is rare for high-cost patients to suffer from only one chronic illness; they often have multiple health problems that need to be addressed together. “There is this point where it becomes very difficult to turn the patient’s health around. Using analytics, predictive modeling, and forecasting, we’re trying to find that patient before they reach that threshold,” Gladden said.

Taking a Closer Look at the Top One Percent

When the state decided it wanted to pay more attention to patients with the greatest expenses, CareSource made a convincing case to use the analytical approach it had already pioneered. “A patient could be in the top 1% because of a one-time incident, such as an otherwise healthy baby in the neonatal unit because of premature birth,” Gladden explained. In addition, the HMO has figured out that some patients are simply more receptive to efforts to manage their health—especially now that CareSource can predict the best time of day for a nurse to call them.

“We want to save money, but as a mission-driven, 

nonprofit company that isn’t the primary reason we do this,” Gladden said. “We’re really concerned about helping the members stay healthy.”