Healthcare data management challenges hold back adoption

Limited investment in data management techniques is a major challenge

The healthcare industry has not been as quick to adapt to the massive influx of data. Today, few healthcare organizations operate as modern, digital businesses.

"This is an area both payers and providers have underinvested in and have remained behind compared to other industries, while their consumers have grown accustomed to data-driven experiences from other industries," said Munzoor Shaikh, senior director of healthcare and life sciences at West Monroe Partners, a business and technology consulting firm.

This puts healthcare organizations in a difficult position. Most providers have traditionally focused on improving clinical operations. Now they're being asked to continue that push while also doing more with data.

"Healthcare organizations have the compelling need to leapfrog to where their customers are already at," Shaikh said. "A secondary, yet difficult, challenge is governance as it relates to treating data as an asset."

Traditionally, in healthcare, data has been seen as an IT problem. But this mindset cuts against the grain of modern approaches that treat data as a valuable business asset.

"In reality, data cuts across different departments -- IT, digital, marketing, clinical, operations and strategy," Shaikh said.

Shaikh believes most healthcare providers have yet to accomplish this unified view of their data. Implementing a holistic approach to data remains one of the biggest healthcare data management challenges.

"This [unification] is especially important in healthcare. As organizations take on [greater financial] risk or engage patients and consumers to match acuity to utilization of service types, data becomes absolutely necessary," Shaikh said.

Healthcare data management challenges

Data management has not become a standard in healthcare, but this is due to more than organizational inattentiveness. The unique obstacles that stand in the way of these organizations are daunting.

One problem is due to the nature of the data. Evolving clinical research data sets can be a useful source of information, but they present issues. An overreliance on these data sets can hurt health organizations when it comes to generalizability.

Oscar Marroquin, chief healthcare data and analytics officer at the University of Pittsburgh Medical Center (UPMC), said that one of the Achilles' heels of academic medicine is that insights gained from research endeavors can be inaccurate. The ways in which studies are conducted don't always match real-world conditions, which makes their findings sometimes difficult to implement in clinical settings.

This means the data cannot lead to knowledgeable inferences, which creates a major data quality issue. Without effective data governance, inconsistencies can crop up in different systems across the organization. This can lead to incorrect information and delay treatment.

Creating a hierarchy within the data team

"These are really the nuts and bolts of running an operation," Shaikh said. "The reason the need is heightened in healthcare is because data comes from many sources, and the discipline or capability of cleansing that data has never quite been built by healthcare organizations."

Marroquin said UPMC has tried to tackle this problem by centralizing responsibilities.

UPMC has created a leadership hierarchy built specifically to handle the organization's analytics and data management, which Marroquin leads. This gives the organization a more focused approach to the challenges of data privacy, management and governance.

"A good way of thinking about it is that this analytical stack needs to be linked together and has to have a singular vision and goals," Marroquin said.

Prioritizing data management through the creation of this hierarchy enables UPMC to tackle the unique challenges of healthcare data management.

According to Marroquin, if one doesn't establish these hierarchies in a holistic way, then it is very difficult to get the most out of all the data. Slow adopters of data management best practices need to first prevent infighting by creating clear lines of communication and responsibilities.

Otherwise, "it is more likely to lead to disjointed, piecemeal approaches that don't work as well," Marroquin said.