Enterprise data strategy in the quantum age

This is the 1st blog in our series “From data diffusion to insights with enterprise-wide data strategy”

sameer-siraj-1200x628An interview with Sameer Siraj, Vice President, Strategic Product Management, Health Care Technology as a Service, Optum

Across all industry sectors, U.S. health care payers are the least advanced in digital business strategies. Many health plans continue to rely on legacy technologies to power their businesses. Meanwhile, all around them, enterprises have re-engineered data storage, compute and analytic capabilities to make data work harder for their businesses. If you can capture and capitalize on it, big data in health care has the power to:

  • Improve and connect care
  • Reduce fraud and waste
  • Predict outcomes
  • Cut costs

If payers don’t capture and capitalize on big data, the threat of payer obsolescence is real and imminent.

In the first of a four-part series, Sameer Siraj defines the strategy and supporting technology of moving from data islands to big data. Siraj is vice president of Strategic Product Management for Optum Health Care Technology as a Service.

Q1: What is an enterprise data strategy?

An enterprise data strategy is a natural derivative of the Digital Revolution that began in the 1980s. It is the blueprint for migrating and assembling data from across the organization into a single source of truth. It allows payers to consume the data for broad analysis.

The role of the health plan must rapidly shift from merely managing health care costs. In order to significantly improve outcomes and member experience, payers must use untapped data. The first steps entail:

  • Understanding and separating all data sources from analytic tools
  • Defining how data is used to meet business objectives
  • Determining what technology will securely deliver the right data at the right time in the right place

Q2: Why now?

The cost of data storage and big-data computing has dropped dramatically in the last 20 years. We’re moving from the Digital Age into the Fourth Industrial Revolution — the age of optimization. Health care generates as much as 30% of stored data worldwide. Today, a single patient generates close to 80 megabytes each year in imaging and electronic medical records. This data has clinical, financial and operational value for the health care industry — if we can capture, contextualize and use it effectively in practice.

To remain competitive, payers need to free data from siloes and compile data in a system that makes all data types available to powerful business analytics. The analytics must continuously illuminate areas where we can reduce medical spend, enhance care quality and improve the customer experience. Without a holistic enterprise data strategy, health plans will lose their relevance in the health ecosystem.

Q3: How will a holistic approach to data management change the health plan paradigm?

Data in health care is very fragmented today. Health plans need upwards of 50 business applications to support health care operations and members — each with associated data — and new data sources are coming online rapidly. However, most plans use a fraction of structured data and next to none of their unstructured data to make decisions.

Aggregating and preparing data for consumption on a single platform enables relationships among data sets. It also supports reuse of data without additional expense. Not previously possible, infinitely scalable cloud competencies:

  • Facilitate a modular approach to data storage, computation and analytics
  • Shift the financial paradigm for plans from a capex to an opex service model

By harnessing the full range of its data, a payer sets the stage for a single, 360-degree point of view on the health of each member. A whole-person view of each member is a global shift from retrospective views of discrete data sets. Discrete data sets provide a partial and incomplete snapshot of an individual, but a cognitive or “learning” point of view allows health plans to confidently predict outcomes, improve care quality and lower costs.

With big-data analytics (such as machine learning) applied to aggregated and enriched data sets, payers can make accurate predictions. Enriched data may include longitudinal patient history, demographics, claims, facility and quality data. Enriched data analytics allows a payer to accurately predict, for example:

  • The right facility for an ER patient’s follow-up care
  • The types of specialized care and equipment that may be required
  • The likelihood of chronic disease

Payers could also set accurate payment expectations and predict whether a patient may need financial assistance. Enterprise data strategy goes to the root of personalized medicine or “precision” medicine and sets the stage for delivering a more consumer-focused experience.

Sameer Siraj describes how an EDP addresses the structured-unstructured data challenge. Listen now.

Q4: Are health plans ready to adopt an enterprise-wide model of data management?

I would say 40% to 50% of the health plans we work with understand what they’re facing today, but are using one-off applications that tie isolated data and analytics together to respond to specific business needs. The trouble with this approach is that when analytical tools are tied to data, it’s very expensive to integrate new tools, scale or deploy quickly. A better approach is to free the data from silos, prepare for new data sources and architect across the enterprise to exponentially increase the value of information assets.

Until health payers make this shift, they will have minimal ability to reduce medical spend. For example, let’s say an organization uses personal, population and predictive data to recommend the right treatment and length of stay at the highest-performing facility with the highest-performing provider for the particular patient demographic. In so doing, the payer reduces readmissions. The plan could save $35,000 per patient (the mean cost of readmission). Whereas, with siloed data and analytics, health plans will continue to chip away at administrative costs. That’s an important effort, but ten times less valuable than medical cost avoidance.

Siraj compares an EDP to earlier technologies. Which one is right for you?

Q5: What are the priorities in selecting the right cloud partner for your enterprise data platform?

In selecting a platform to fulfill the enterprise data strategy, choose a consumption-based cloud model. This is one in which storage and analytics can be utilized and paid for as needed. The model should support efficiency, scalability and data security, which are critical to reducing costs in both the short and the long term. EDP technology should support:

  • Machine learning and artificial intelligence (AI)
  • Agile response to new data sources
  • The ability to “sandbox” applications as a way to test analytical scenarios

Q6: What are the pillars of design to move from siloed to compiled data governance?

For a payer that is serious about using data to provide insights and business direction, a good enterprise data strategy proposes:

  • Separating data from operations and analytics
  • Moving data to a purpose-built, highly resilient cloud for cost-effective long-term management of large data sets
  • Evolutionary processes and practices for ongoing changes, data types and analytical scenarios
  • Integration and connection with everything that comes after it or from it

Formerly, the payer CIO was tasked with ensuring technology-enabled operations run smoothly. Today, he or she will work with internal stakeholders to release data and anticipate new data sources for cloud-enabled, enterprise-wide data analysis.

Why is it important that a CIO build an enterprise data strategy before embarking on an EDP deployment? Siraj explains.



Visit Optum.com/CIO to learn more about moving from data diffusion to insights with enterprise-wide data strategy.

Who has a stake in the payer journey from data islands to big data? Read the next blog in our series to find out.


About the author: 

Sameer Siraj, Vice President of Strategic Product Management, leads product strategy and definition for the Health Care Technology as a Service portfolio at Optum. His team is responsible for developing digital domain products and services spanning infrastructure, applications and data for health care payers and providers.

One thought on “Enterprise data strategy in the quantum age

Leave a Reply