CIOs push for patient ID progress

"The real opponents are the privacy advocates."
By Neil Versel
10:36 AM
Bill Spooner CIO Sharp CHIME

The patient identification issue refuses to go away, mainly because nobody has quite figured out how to assure proper patient identity when engaging in health information exchange. At the College of Healthcare Information Management Executives Fall CIO Forum in Phoenix last week, some leading hospital CIOs emphasized the operational, clinical and financial importance of accurate patient matching.

Joey Sudomir, senior vice president and CIO of Texas Health Partners – the healthcare management affiliate of Arlington, Texas-based Texas Health Resources – said it costs about $600 to $800 to remediate duplicate patient identities following hospital discharge.

[See also: Rekindling the patient ID debate.]

The IT department at Sharp HealthCare in San Diego has 10 full-time-equivalent employees just to investigate and clean up duplicate records, at a cost of about $1 million a year, according to Senior VP and CIO Bill Spooner. Matching also goes on in other departments of the health system, so the overall price tag is probably significantly higher, Spooner said.

That’s only part of the problem, however. "I think there is a patient safety issue," Spooner added, as well as a privacy one. Nearly 20 years ago, Sharp mailed 15,000 mismatched bills sent to the wrong people, he said. Imagine if those had been patient records.

Panelists, also including CHIME CEO Russell Branzell and Stacie Durkin, RN, a member an American Health Information Management Association work group that produces a practice brief on patient identity integrity, expressed varying levels of support for a national patient identifier. The 1996 Health Insurance Portability and Accountability Act called for the creation of a national patient ID, but following privacy-related protests, Congress voted in 1998 to block the use of public funding to meet that requirement.

"I think it is an option," said Durkin, a health IT consultant in Kansas City, Mo.

A national ID would help foster interoperability and create a "universal view of each one of us" in a care setting, Sudomir added. He said there’s a high risk of errors from mismatched records as clinicians start to make decisions based on longitudinal patient records.

Spooner said there needs to be a national patient ID because trying to match records based on multiple criteria is difficult. "The real opponents are the privacy advocates," he opined. But he said the concerns might disappear in a generation or so, based on a reported statement from TEDMED owner Jay Walker, who, according to Spooner, said that nobody under 30 cares about privacy.

Spooner said he would like to see a study quantifying the savings from having an effective patient-matching algorithm. Sudomir also would like to see someone quantify the waste due to misidentification. However, Sudomir does not believe any of this research will take place before the Office of the National Coordinator for Health Information Technology – currently crippled by the government shutdown – completes its work on patient matching.

ONC kicked off the Patient Matching Initiative in September, with the goals of identifying the data points that lead to the most accurate record-matching across health systems and defining best practices for patient matching.

"Obviously, we've got to have standards," Spooner said. There also has to be cooperation between providers, vendors and government, though the Sharp CIO wishes ONC had included record matching in its early work on meaningful use. Since the federal agency did not, he worries about errors if ONC and CMS include standards that haven’t been fully tested in the forthcoming meaningful use Stage 3 regulations.

Sudomir said it was "realistic" to get to 75 percent to 80 percent accuracy in matching patients across organizations within five years, but it will take some quick work from the government. "I think we should set ambitious goals because it's a worthy cause," he said.

Durkin said the goal should be "one record for one person," something that could be achieved in three years with the right effort among many different stakeholders. "Give us a national patient ID and we'll knock it out in 12 months," Durkin said.