The effectiveness of predictive capabilities in improving health outcomes starts with understanding where we can effectively apply these new and powerful technologies.
Atlanta, GA (PRWEB) July 31, 2014
Atlanta-based Jvion, a leader in clinical predictive algorithms and machine learning, released their top three lessons in predictive population health analytics as part of an ongoing series dedicated to understanding and applying predictive analytics in healthcare. Ritesh Sharma, Jvion COO, commented, “it is critical that we work together as an industry to understand the impact of and potential within predictive analytics. There is a lot of information out there and a good portion of it is conflicting or confusing. The effectiveness of predictive capabilities in improving health outcomes starts with understanding where we can effectively apply these new and powerful technologies. When applied right, predictive analytics empower hospitals to proactively do a lot of things.”
Top Three Lessons - Summary Findings
Why focus on population health?
Population health initiatives have the overarching goal of targeting specific at risk populations to apply low cost interventions across the care continuum. New value-based payment models and systems like accountable care organizations are forcing many providers to rethink their approach to prevention and evaluate the effectiveness of predictive technologies in targeting specific, high and rising-risk segments of the community.
What role do predictive analytics play in population health?
There are three types of population health solutions in the market: first, traditional analytic solutions (retrospective data analysis); second, Enterprise Data Warehouse (EDW) based predictive analytics; and third, (new) Machine Learning algorithm powered solutions. Traditional analytic solutions involve retrospective data analysis, benchmarks, and trending and are primarily focused on looking back in time to understand how things worked in the past. The problem with this approach is that it only helps maintain a scorecard; it does not provide enough actionable information for organizations to proactively plan and change future results.
The other two forms of population health solutions are predictive: EDW and Machine Learning. EDW based solutions compile large amounts of data into a single data warehouse. This data is then consolidated and analyzed for patterns that lead to predictive insights. Machine Learning based solutions start by building models and clusters, and analyzing individual risk levels for all patients across a population. This individual-level risk is then aggregated and stratified into risk cohorts that can be targeted for specific interventions.
Feedback and findings suggest that Machine Learning solutions are better suited for the healthcare industry because they tend to deliver more detailed and more accurate results that don’t require a heavy investment in an EDW. The time it takes to deliver value is also significantly different. EDW solutions take a minimum of 18-24 months to stand up whereas Machine Learning solutions can start to deliver outputs in weeks. Additionally, the more advanced Machine Learning solutions currently available are able to use publically and readily available data elements to quickly stratify risk and define cohorts at accuracy levels that are much higher than earlier generation models.
How do you articulate the Return on Investment (ROI) for these solutions?
The ROI for a predictive analytic solution can seem a little “fuzzy” because it is about cost avoidance. The abstracted benefits of population health initiatives are well known. Employers benefit from lower absenteeism and injury rates, and the subsequent increase in productivity. Taxpayers benefit by lowering the number of chronic conditions treated through the Medicare and Medicaid systems. And communities benefit from the overall strengthening of the economy and associated influx of federal investment/benefits.
For individual hospitals, the numbers can get a little blurry and slippery if they are not already proactively managing population health. However, with risk-based and value-based contracting practices becoming more prevalent across payors, population health initiative ROI becomes much more straightforward and concrete for hospitals. In addition, findings suggest that predictive analytics actually enable better ROI measures. This is because they can quickly and comprehensively analyze historical data to determine the dollars saved if preventative measures were applied based on predicted risk insights. Using this approach, a hospital can assign a hard dollar ROI not only to population health overall, but down to a specific disease and subset of a risk stratified cohort.
For more information on Jvion and their population health solutions, please click here. And for information on the firm’s entire suite of predictive analytic solutions, please visit http://www.jvion.com.
Jvion is a healthcare technology company that develops software designed to predict and prevent patient-level disease and financial losses leading to increased waste. The company offers a suite of big-data enabled solutions that combine clinical intelligence with deep machine learning to help providers protect their revenues while improving patient health outcomes. The objective is simple—stop the waste of resources and lives by predicting and stopping losses before they ever happen.