Data analytics, AI, and machine learning are improving clinical decision-making, personalizing treatment, and accelerating drug discovery.
But what do the fundamentals of data-driven transformation look like in healthcare and life sciences? And what pitfalls does the sector need to avoid as it continues its journey into the data-driven future?
Investment needs to enable interoperability
Investment in analytics for healthcare and life sciences is surging. A recent survey of senior technology decision-makers working in the sector in North America, Europe and Asia suggests that six out of 10 expected spending on data and AI to rise by up to a quarter during 2023-24. A further 38% expected even larger increases.1
Some of this investment, particularly in life sciences, is focused on leveraging AI and machine learning.
One example among many: the need for more collaboration as the next generation of personalized medicine comes to market.
“Personalized healthcare means that life sciences companies need to get involved in healthcare delivery, sharing data and becoming part of the workflow,” says Andrea Sorkin, Director of Healthcare Data and Analytics Consulting at EPAM Systems.
So far investments have been focused on single initiatives in one domain, but enterprises need to invest in models that enable interoperability across the entire ecosystem of healthcare and life sciences to expedite drug discovery, clinical research and to enable patients to be managed holistically.
Future of healthcare
Another key focus in healthcare is around improving systems management with data-driven approaches.
As Ajay Patel, Managing Principal for Healthcare Strategy at EPAM Systems, puts it: “No matter which national market you talk about, the ability to gather healthcare information, convert it into high quality data, aggregate it, and then develop longitudinal patient histories: all of this is vital in the future of healthcare.”
“Everyone recognizes that better health data analytics needs to happen,” says Sorkin. “They want to manage utilization and keep costs down.”
Other associated factors are driving the data revolution. Organizations must address risks created by uncontrolled data duplication and data security. They must also modernize to deliver wider deployments of AI and machine learning.
The key to data success at scale
Building effective data strategies at scale isn’t easy. In the UK, the National Health Service has started work on a Federated Data Platform, its latest effort to unify a national treasure trove of data, much of which is locked away in legacy silos.2
This will be a challenging undertaking: the key to its success will be interoperability and high-quality data.
In the US, many of the challenges are similar, although the existence of many small healthcare organizations complicates the picture, says Patel.
Today, efforts to build effective data strategies to drive successful use of AI are often underpinned by lakehouse technologies from specialist vendors like Databricks.
“All of these files in different formats, different volumes and different semantics create a very substantial challenge,” says Patel. “Platforms like Databricks have emerged as a one-stop shop for collecting, ingesting and transforming that data. On top of that, Databricks allows you to layer data governance policies and store those disparate datasets within a very uniform domain-based data model.”
As usual, the future has arrived, but it remains unevenly distributed. “Today, the big players in the US understand that interoperability and the ability to aggregate patient data is a game changer,” says Patel. “Have they all built data strategy solutions that open the way to expanded use of AI and machine learning? Not all of them. But they increasingly understand the need.”
Learn more about the EPAM and Databricks partnership.
1 MIT Tech Review Insights, Bringing Breakthrough Data Intelligence to Industries (January 2024) https://www.databricks.com/resources/ebook/bringing-breakthrough-data-intelligence-industries
2NHS Federated Data Platform (FDP): Digitising and Connecting Data to Transform Health and Care https://www.england.nhs.uk/digitaltechnology/nhs-federated-data-platform/