The alliance’s Technology Leadership and Policy committees have concluded that the current statistical process for matching patients to their records based on such attributes as name, address and birth date is too unreliable.

Alliance representatives say the organization has been focusing on the issue of patient identification for three years, and has held forums, reviewed research and gathered input from a range of experts.

“Outside of carefully controlled pilots, accuracy for the current process is roughly 90 percent, based on our collective experience and industry estimates,” said Tom Doyle, vice president and chief architect for HCA and a member of the Alliance’s Technology Leadership Committee.

 “That margin of error will only widen as it is applied to ever-larger populations,” Doyle added.