EHRs and the Paper Monster

The healthcare industry has yet to transition to a paperless model, but leading organizations have been achieving some success.

This article first appeared in the January/February 2015 issue of HealthLeaders magazine.

While electronic health records are helping to move the industry toward being paperless, the goal remains elusive, if not unlikely. Complicating the effort is that certain documents, including those created and forwarded by payers, researchers, and administrators, live outside the electronic health record.

The EHRs themselves have limited capabilities to search and organize scanned paper records, many of which are still stored as images, while others go through optical character recognition (OCR) or intelligent character recognition (ICR) in order to be indexed and searched.

Two efforts that promise to unlock and organize unstructured text are natural language processing (NLP) and IBM's Watson, a set of parallel-processing, cloud-based machine learning technologies. But while these futuristic technologies evolve toward maturity, providers are finding simpler tools to index, organize, search, and present information culled from unstructured, machine-readable documents as a way to coordinate care, speed compliance, and accelerate research such as clinical trials.