Data: The first step to meaningful change

People often assume that the biggest database is the best. Size is important, yes, but equally important are “data types,” “data velocity,” and “data integration and accessibility.” Although William Cowper certainly wasn’t thinking of databases in 1785 when he wrote “variety’s the very spice of life that gives it all its flavour [sic],” he was on the mark.

At Optum, we manage a de-identified health care database that covers 216 million individuals — in some cases, spanning more than 20 years. The fact that it’s de-identified means that while we have a wealth of information on millions of patients, we’re unable to trace personal information back to a specific patient. What makes the database powerful is that it includes information from a variety of internal and external sources, including electronic health records, claims and employer benefits data, along with consumer insights.

Of course, data does not mean much until you do something with it. That’s where expertise comes in. When professionals such as actuaries, health economists, biostatiscians, epidemiologists and data scientists use that robust reference data set to build models and analytics to measure, anticipate or act, real change happens. But where along the spectrum of analytic disciplines do you need to begin?Descriptive analytics

Are you looking for quantitative analyses of a population to understand total cost of care opportunities? Descriptive analytics provides these measures, like PMPM (per member per month) calculations or utilization per 1000 measures, which health plans and employers use to manage the cost of patient care.

Descriptive analytics are the most common type of analytics used in health care today, and it will certainly always have a place. However, those types of metrics are not sufficient to gain a competitive advantage. Solely relying on descriptive analytics is like driving your business forward while only looking in your rearview mirror.

Predictive analytics

Are you interested in probabilities? Predictive analytics will provide data on what is likely to occur in the future, such as identifying people who have low, medium or high propensity to develop a chronic condition or be readmitted to the hospital. Or perhaps you are interested in finding people who are “at risk” of having high cost care in the next 6–12 months. These types of measures are helpful ingredients in developing population health targeting activities or actuarial risk adjustment practices.

Prescriptive analytics

Are you hoping to act on those probabilities or risk thresholds? Prescriptive analytics provide information on potential next steps for the population being studied, like identifying the best rehab center for a patient being discharged by matching the patient’s specific conditions and comorbidities to the outcomes and measures of the rehab center. You could also identify certain types of “next best actions” for chronically ill members who are trying to follow a path toward wellness.

Data and analytics require a clear business strategy. In fact, if you don’t have a business strategy or know what you’re going to do with the insights gathered from an analytic process, don’t bother investing in data and analytics in the first place — it will be a waste of money. And most importantly, moving a business forward requires that your team readily apply insights from analytics to their day-to-day job responsibilities, influencing behavior and driving infrastructure decisions that support your organization’s strategy and associated success metrics.

The right mix of data, analytics, insights and actions can help you make changes that result in real value.


About the author:

Steve Griffiths Headshot

Steve Griffiths, PhD, MS
Senior Vice President, Chief Operating Officer
Optum Enterprise Analytics

Steve Griffiths has more than 20 years’ experience in health analytics management, and currently heads up the Optum Enterprise Analytics organization. His main focus is driving growth and innovation through Optum products and services. Steve has a master’s degree in biostatistics from the University of Washington and a PhD in health services research, policy and administration from the University of Minnesota.

One thought on “Data: The first step to meaningful change

Leave a Reply