Arizona to Study Unique Patient Identifiers | David Raths | Healthcare Blogs Skip to content Skip to navigation

Arizona to Study Unique Patient Identifiers

March 21, 2014
| Reprints
State’s HIT 2.0 Roadmap focuses on 19 initiatives

Like most states, Arizona is in a transition period in terms of health IT initiatives. It has made great progress in EHR adoption (80 percent) and has a statewide health information exchange with a sustainable business model that includes funding from both hospitals and payers. But as federal grant funding for HIE and regional extension center outreach work subsides, Arizona is shifting gears to focus on reaching more providers and expanding the types of services the HIE offers.

The Arizona Strategic Enterprise Technology Office funded the creation of the recently published Health IT Roadmap 2.0, which received input from more than 300 stakeholders and describes 19 key initiatives to advance HIT/HIE in Arizona. This morning I had the chance to speak with Arizona Health-e Connection CEO Melissa Kotrys about a few of those initiatives.

The first one that caught my eye was the study of a statewide unique patient identifier. Last fall I saw a debate at the AMIA conference in Washington, D.C., about whether AMIA should advocate for a unique health identifier for care and research purposes and if so, what kind of standard should it support under what conditions? There were strong feelings on both sides. An audience poll following the debate seemed to suggest that a majority wanted AMIA to advocate a unique health identifier, but there was widespread recognition that political headwinds continue to make consensus on the right direction unlikely.

I mentioned that meeting to Kotyrs, and she agreed with my description of it as a political hot potato. “It is a controversial issue,” she said, “but when we convened stakeholders to talk about key issues, this topic of the difficulty of patient identification came up, even with master patient index solutions. So there was enough interest to explore this topic further.”

According to the Arizona Health IT Roadmap 2.0 document, the lack of a unique identifier for every patient causes difficulty in ensuring that: 1) patient records are always matched to the correct patient and 2) all records belonging to a patient are identified and associated with the patient. As the number of patients and the number of organizations participating in HIE grows, the complexity of correct patient record matching is magnified.”

Arizona plans to explore best practices and identify alternative approaches it can use to address the increasing complexity of correctly matching patients to all of their available records across multiple organizations.

Another change Kotrys mentioned was that during the course of developing the roadmap, AzHeC tightened its affiliation with the statewide HIE organization, the Health Information Network of Arizona (HINAz). AzHeC will continue to serve as the multi-stakeholder organization that handles policy development, education, stakeholder engagement and various HIT/HIE programs such as the Arizona Regional Extension Center, while HINAz will serve as the “network” under the AzHeC umbrella that connects providers through the statewide HIE.

Here are a few other interesting initiatives on Arizona’s to-do list:

An HIE Consent Management Engine: This would develop a technical infrastructure to support the common statewide patient consent approach and processes. The consent management engine should be able to:

Manage consent from multiple sources (providers, HIEs, and patient-generated sources)

• Manage consent for various types of health information

• Support current levels of consent for secure sharing of patient information

• Support increasing levels of granularity of patient consent

• Harmonize approaches and consent engines from multiple organizations to ensure there is consistent compliance with a patient’s consent for the sharing of their information

• Integrate with other technologies to deliver notices to providers if there is additional information in the HIE that has not been consented to be shared – as appropriate within current laws and regulations.

Integration of Physical, Behavioral HIEs: Arizona has an existing network for sharing data between behavioral providers. It is called the Behavioral Health Information Network of Arizona. Kotrys said there are efforts underway first to offer providers on BHINAZ to access data from the statewide exchange and eventually to work through the more difficult regulatory and granular consent issues to allow physical providers to access the behavioral health information about their patients.

Arizona HIE Marketplace: The Arizona Health Information Exchange (HIE) Marketplace is a program established by AzHeC through ASET funding to get providers on the HIE on-ramp by using Direct secure messaging. Kotrys said that the first-phase uptake was low, but that meaningful use requirements should lead to greater usage.

I came away from our conversation impressed with both the scope and level of detail of the initiatives in Arizona’s roadmap. I’ll check back with them in six to nine months to get a progress report.






In response to David Raths’ March 21 blog post in Healthcare Informatics, here’s a short quiz on unique patient identifiers.
True or False: There is a government prohibition that prevents AMIA, the states, and others from advocating for and evaluating unique patient identifier solutions.
False: The federal government is indeed prohibited from spending federal funds on patient identifier research and implementation, a prohibition that was put in place in response to the 1996 HIPAA legislation which called for a national identifier card and caused loud public outcry about potential privacy risks and high costs. However, the Office of the National Coordinator for Health IT (ONC) managed to find a way around it when it funded a study on how to improve patient matching across disparate systems (including unique IDs) in 2013. While this study stops short of making specific recommendations, it reinforces what many of us believe: the lack of unique identifiers is perhaps the biggest challenge we face in achieving secure, safe health information exchange (HIE). Many states have enacted legislation calling for unique IDs (this 2012 study provides a good overview --, but little progress has been made as most states have been challenged to put basic HIE infrastructure in place, along with the necessary governance and sustainability plans – at the same time, meeting increasing government requirements in the healthcare sector such as the Medicaid portion of meaningful use, implementation of the Affordable Care Act, and ICD-10. But the list of organizations that have come forward in support of a national unique patient ID is long, and there is no shortage of research that justifies serious pursuit of a solution.
True or False: Current enterprise master person index (EMPI) technologies are adequate to ensure accurate patient matching and linkage to records across multiple provider sites.
False: There is a well-documented error rate with EMPIs of 8-10%. That is, 8-10% of the time patients are registered and a new medical record is created even though one already exists or, worse yet, they are linked to someone else’s medical information. The resulting waste of resources (e.g., duplicate or unnecessary testing, staff time in searching for and correcting registration errors and medical records) and medical errors (e.g., drug allergies, drug-drug interactions, unnecessary and risky procedures) costs us all and, sadly, often results in patient injury or death.
True or False: A unique national identifier will require a central database of private health information (PHI) which will be costly to design and implement, and adds risk of security and patient privacy breaches. The federal government is notoriously unsuccessful in managing such large projects, and private industry cannot be trusted with the kind of sensitive information that employers could use to make employment and insurance decisions.
False…and True: There is at least one technology that is available today that does not require a central database of PHI, nor government oversight of complex systems and security protocols. In fact, the solution from Global Patient Identifiers, Inc. ( is designed specifically to take the government out of the equation. The design employs distributed, redundant data centers to issue and manage identifiers via the Internet. If the system were to be “hacked”, the only information that the hackers would obtain are the identifiers themselves since GPII has no PHI, not even patient names. GPII never sees any sensitive healthcare information for any patient.
GPII is established as a not-for-profit private company with the sole objective of making unique identifiers available to help resolve the high costs of mismatched patient records, erroneous registrations, and medical errors.
GPII recognizes the potential problems with patient privacy and has very explicitly addressed this in its solution.
Coincidentally, GPII is located in Arizona. Our solution is operational today. We believe that a pilot project in Arizona or elsewhere would quickly demonstrate the efficacy, the acceptance, and the cost benefit of this approach to solving a very serious problem.
For more information on GPII visit