DrFirst

Bridging Healthcare Data Gaps With Clinical-Grade AI

Automation Paves the Way for Smooth EHR Workflows

Change in healthcare is notoriously slow. Consider systems interoperability and data sharing, two related problems that the health IT industry has been trying to solve for decades.

In a recent Empowered Patient Podcast, host Karen Jagoda and our Chief Medical Officer Colin Banas, M.D., M.H.A., explored the persistent challenges of systems that don’t speak the same language, incomplete healthcare records, and the crucial role clinical-grade AI plays in filling those gaps. Dr. Banas, a former hospitalist and Chief Medical Information Officer for VCU Health System, sheds light on the hurdles clinicians face when they try to access information as seemingly straightforward as medication history data.

Understanding the Root Causes of Incomplete Data

A significant hurdle in overcoming missing medication data rises from incomplete data sources. The standard industry data feed is based on financial transactions from payers or pharmacy benefit managers, which often provide only basic information such as the name of a drug. Robust clinical details are often omitted, such as prescription instructions, or “sigs,” that include directions like, “Take one tablet by mouth once daily.”

Dr. Banas explains, “When those details are missing, it causes problems for the receiving system. And it causes problems for the clinician, requiring them to either re-input that data and potentially make a mistake, or worse, make a prescribing decision without complete data. With something as important as medications, this is where errors pose risks to patient safety.”

Addressing Data Gaps With Clinical-Grade AI

When essential details like sigs are missing from a patient’s medication record, clinical-grade AI becomes a powerful ally in bridging the gap. AI assists the clinician by inferring the medication instructions and placing them in the proper fields in the electronic health record (EHR). When there’s any doubt about clinical meaning, the data is not imported, which appropriately flags it for a clinician to verify before adding to the medical record.

Automating this workflow reduces the manual entry of data and the likelihood of a clinician introducing an error, such as entering an 11 instead of a 1 by holding down the number key too long. It also goes a long way toward combating the burnout that results from repetitive, manual tasks that take time away from direct patient care.

Listen to the full podcast to understand the interoperability challenges that prevent medication data from being accessible to providers and how technology can fill gaps safely so healthcare providers can elevate patient care, minimize risks associated with incomplete medication history, and pave the way for a more connected and efficient healthcare system.