Health Reform WK-EDGE Predictive analysis uses data to improve costs, quality
News
Tuesday, February 9, 2016

Predictive analysis uses data to improve costs, quality

By Patricia K. Ruiz, J.D.

Since the enactment of the Patient Protection and Affordable Care Act (ACA) (P.L. 111-148), provider payments have increasingly been tied to quality and efficiency. The increasing complexity of the industry, paired with the government’s focus on improving the quality and lowering the cost of care requires health care organizations to make smarter, more informed decisions, according to a report by IBM. The expectations of consumers are also shifting, as they expect more access to information and increased accountability from their doctors, nurses, and health plans. As a result, health care providers are turning to big data and predictive analysis to improve patient care and control costs. Data analytics helps to sort through this complexity and allows these organizations to deliver on the demands of consumers as well as the government.

What is data analytics? Analytics is “the systematic use of data and related business insights developed through applied analytical disciplines (e.g., statistical, contextual, quantitative, predictive, cognitive, and [other models]) to drive fact-based decision making for planning, management, measurement, and learning,” according to IBM. Analytics may be used to describe trends, predict outcomes, or prescribe actions.

This Strategic Perspective will discuss the benefits of using predictive analysis in health care and show how providers have adopted technology as a means of improving the quality of care. Finally, this Strategic Perspective will discuss the challenges providers face in instituting data analytics.

Advocate Health Care Collaboration with Cerner

Advocate Health Care and Advocate Physician Partners (Advocate, collectively), has been collaborating with Cerner, a global health information technology supplier and developer of electronic health records (EHR) system, to work toward advancing health care, lowering costs, and improving patient and clinician experience. The collaboration creates and refines advanced analytics and predictive models using Cerner’s HealtheIntent platform to collect, streamline, and structure clinical, financial, and operational data across the continuum of care. Advocate was able to integrate collection and analysis of data into its EHR workflow.

In 2012, Advocate joined with Blue Cross Blue Shield of Illinois to create its first value-based contract and, shortly after, it enrolled the resulting accountable care organization (ACO) in the Medicare Shared Savings Program (MSSP). Advocate realized that, in the transition to value-based patient care, it needed to provide physicians, nurses, and other caregivers with comprehensive information on which to base its patient care decisions. Unfortunately, the claims data it usually relied on tended to be stale. With the goal of aggregating real-time patient data that would allow it to predict medical events, Advocate, already a customer of Cerner’s EHR product, signed on to use Cerner’s HealtheIntent with the goal of automating much of the data collection and analysis. In some cases, results of the analysis are integrated directly into Advocate’s EHR system.

An algorithm to reduce readmissions. Together, Cerner and Advocate have developed algorithms to help predict an individual patient’s level of risk for a given medical outcome. One particular algorithm leverages data to help physicians choose the next level of care for a given patient (for example, skilled nursing facility, long-term acute-care facility, or in-home health services) for patients about to be discharged from the hospital. The algorithm helps support the goal of reducing the patient’s risk of returning to the hospital within 30 days of discharge.

“To some extent, any unplanned readmission is really a failure to ensure that we are providing the best transitions of care and the best patient management. So it became a top priority to ensure that we had the data analytics capabilities that we needed to tackle that problem,” said Tina Esposito, vice president of the Center for Health Information Services. “We really wanted to make sure that we could identify patients at highest risk for readmissions while they were still in the hospital, so that we could intervene appropriately.”

The algorithm developed by Cerner and Advocate allows the care manager or clinician to review a patient’s risk score as part of his or her EHR workflow, Esposito said. The risk score is updated every two hours during the patient’s stay. Rather than requiring the clinician to go into a separate application to input data and create a risk score, the algorithm is incorporated into the EHR system. It uses real-time clinical data to evaluate the patient’s readmission risk, allowing the clinical to act quickly and appropriately. A test of the algorithm performed in late 2014 through early 2015 showed a lower readmission rate.

Instituting Data Analytics

Analyzing data to improve the quality of care may not be an easy feat for all providers, said Patricia M. Wagner of Epstein Becker Green in an interview with Wolters Kluwer. “One of the challenges facing providers in instituting data analysis is having the capability to have a complete set of data points for each individual. Missing data elements (for example, data related to treatment at an unaffiliated provider) could be critical to any meaningful data analysis.” This issue can be particularly crippling for a smaller provider that is simply gathering less data than a larger provider. As with any data analysis, it is important to have enough data to make the findings statistically significant.

Smaller providers may work to form collaborative arrangements with other small providers to gather significant amounts of data. Providers who are accustomed to using EHR may have an easier time undertaking a predictive analysis project. Providers that are already active in Meaningful Use are ahead of the game because according to Wagner, “it would not be reasonably feasible to conduct any significant data analysis on paper records. . . . Providers who are active in meaningful use are capturing the data elements in an electronic format in a more consistent manner.”

Finding the best data and the right projects. The availability of data is not necessarily what is preventing providers from adopting a data analysis program. Rather, the sheer amount of data that must be analyzed can inhibit the development of meaningful insight. Nearly 40 percent of participants surveyed by IBM cited the ability to get relevant data as the biggest stumbling block for widespread analytics adoption.

Organizations eager to jump into big data often gather and use data that is the easiest to obtain without considering whether it is really the best data to address the specific problem they are investigating, said Steve Escaravage and Joachim Roski, PhD MPH, of Booz Allen. “While this can speed up a project, the analytic results are likely to have only limited value,” they said. Decision-makers and project staff can forget the importance of knowing the provenance and lineage of data—where it came from and what has been done to it. When data is not properly scrutinized, patterns and anomalies found later turn out to be either irrelevant or grossly misleading.

Providers undertaking a predictive analysis pilot also must ensure that the pilot has wide applicability. “Health organizations will get the most from big data when everyone sees the value and participates,” Escaravage and Roski said. “Too often, though, initial analytics projects may be so self-contained that it is hard to see how any of the results might apply elsewhere in the organization.” The IBM report suggests focusing on the biggest and highest value opportunities, starting with the organization’s biggest and highest priority challenge. Within each opportunity, the organization should start with the questions it seems to answer, rather than letting the data identify the issues. Providers may use the information gleaned from the data to plan for the future. Most importantly, IBM stressed, “[a]nalytics has no value unless it is acted upon.” Insights should be used to drive actions and deliver value. The most effective analytics programs embed small, action-oriented analytics into key decision points of specific business processes used widely in the organization.

Conclusion

The use of data analytics can help to ease the transition of health care from a volume-based to a value-based system. With EHR technology becoming more widespread, health care providers and insurers have greater opportunities to reduce unnecessary medical procedures and lower admission rates and, thus, may maximize the benefit of incentives flowing from initiatives such as the Medicare Shared Savings Program. While undertaking the task of collecting, streamlining, and analyzing data may be daunting, providers can collaborate to create robust data sets that can shed light on methods for delivering efficient and high-quality care.

Attorneys: Patricia M. Wagner (Epstein Becker Green)

Companies: Advocate Health Care; Advocate Physician Partners; Blue Cross Blue Shield Illinois; IBM; Booz Allen; Cerner

MainStory: StrategicPerspectives NewsFeed AccountableCareNews AgencyNews HealthInformationTechnologyNews HospitalReadmissionNews MedicarePartANews MedicarePartBNews MedicarePartCNews MedicarePartDNews PhysicianNews PreventiveCareNews ProviderPaymentNews VBPNews

Back to Top

Interested in submitting an article?

Submit your information to us today!

Learn More