A new Capgemini white paper addresses how the healthcare industry can overcome barriers to adoption of electronic health records.

The paper names several key challenges and obstacles to widespread implementation of electronic records.

"The challenge of identifying the patient is perhaps the most significant one that a national EHR initiative faces," Capgemini researchers write.

But two physicians, both industry leaders say other obstacles remain even more difficult to overcome.

"Lack of a unique identifier is a major problem if one is striving for a guarantee of 100 percent accuracy," said Peter Basch, MD, medical director of the e-Health Initiative. "However, I believe there are existing identity algorithms that are very good, and would adequately serve the purpose of helping the U.S. to move away from siloed healthcare to one that is more patient-centric."

"From my point of view," Basch said, " the biggest challenge is not a unique national patient identifier, but rather that in a fee-for-service system that rewards only episodic face-to-face care, there is no business case for information management and quality.

"Health IT can dramatically improve quality, safety and efficiency, but unless this misalignment of costs and benefits is addressed, IT will not be widely adopted, or more importantly, appropriately used."

Mark Leavitt, MD, director of ambulatory care for the Healthcare Information and Management System Society, echoes some of Basch's remarks.

"I don't think identifying the patient is the biggest challenge to a national EHR initiative, " he said. "I think that honor still goes to the financial barriers to adoption."

Leavitt and Basch and Capgemini noted the firestorm provoked by the proposal for a unique patient identifier, triggered by concerns over threats to privacy. HIMSS has advocated a voluntary patient identifier in the past, but that position is under review, he said.

"A number of technical experts continue to insist that we need a unique identifier, while others say current MPI matching algorithms are adequate (using name, birth date, and other demographic information)," Leavitt said. "Unique identifiers would reduce the possibilities of misidentification which is estimated at perhaps 1 percent with current MPI matching technology.

"As a clinician and pragmatist," said Basch, "I would urge us not to let the perfect be the enemy of the good, but rather to compromise - use the existing matching algorithms we have now, and move forward."