What Do National Patient Identifiers and Donald Trump Have in Common?
By Catherine Schulten
Over the past several years (decades?), the call for a national patient ID has moved beyond discontented grumblings by hospital CIOs to a hot button topic that has garnered national attention from the likes of CHIME, HIMSS, the US Congress, and practically everyone with an opinion who is involved in healthcare data exchange.
A HIStalk poll conducted 2/8/15 asked, “Should the federal government issue a national patient identifier?” The overwhelming response was yes, as 79 percent said yes while 21 percent said no.
Interestingly, a poll done by the Wall Street Journal asking, “Should patients have unique electronic identification numbers for their medical records?” revealed that 44 percent said yes while 56 percent said no.
Industry leaders who support the use of a national patient identifier point to the use of universal patient identifiers (UPIs) in the UK, Ireland, Canada, and elsewhere. They tout efficiencies gained, increased patient safety, the ability to easily pull together a longitudinal record across disparate systems, lower administrative costs, accelerated medical discovery, and the ability to preserve patient privacy. They also cite patient privacy advocates and the existing ban on any federal funding to study or promulgate a national patient identifier as the reason why no forward momentum on this issue has occurred.
Those opposed to the national patient identifier typically cite two primary deterrents: patient privacy and the role of the federal government in establishing an agency that has the ultimate authority to create, distribute, and manage these identifiers.
But before we get into the pros and cons of each side in this debate, let’s first agree on a few items that seem to be overlooked when we talk about a national patient ID.
First of all, let’s quit calling it a national patient identifier. In practicality, it is actually a national ID. From the moment we are born until the day we die, we all have the potential to be a patient. In all countries that have adopted this type of system, the ID is assigned to the patient by the government at birth. In some cases, not only is this ID used to identify an individual for healthcare purposes, but it is also used when securing other government benefits.
Secondly, healthcare is a service that applies not only to US citizens born in this country, but others who may be here legally or not. Nationalized citizens, foreign visitors, individuals with work or student visas, and even illegal immigrants would need to use the ID. Otherwise, how does one know for sure whether Jean-Luc Picard with an ID and the one without an ID are the same or different individuals? For this design to work, an ID process must be supported for non-US citizens as well.
Back to the question at hand: what do national patient identifiers and Donald Trump have in common?
Both are light on details and heavy on promises. We hear what we want to hear when told that a national patient identifier is the only option that solves for true data interoperability, that privacy advocates and their concerns stand in the way of this enlightened future, and that an ID, once introduced, will be used consistently and accurately.
We seem to forget that HIT systems, no matter how well they claim to be protected, are vulnerable to sophisticated security hacks and low-tech identity theft schemes. We forget that healthcare is a service that anyone can secure even if you purposefully choose to anonymize yourself or — in the case of an emergent care situation — are simply unable to provide identity credentials.
But here’s another way that a national patient ID is like Donald Trump. We are fed up with the status quo. We struggle for a way to achieve the promise of unencumbered health information exchange. We’ve invested millions, more likely billions of dollars into the systems and exchanges that are supposed to support data liquidity and yet we still stumble over the seemingly simple matter of accurate patient identification and record matching. We are fed up and we aren’t going to take it any more! We demand action!
As a result, the promise of a national patient ID takes the spotlight and many cycles are spent touting this concept as the deliverance we need. If only the federal government would get its act together and those pesky privacy advocates would quit proclaiming doom and gloom.
However, the truth – as is typically the case – lies somewhere in between.
A national strategy and design for health information exchange that considers the unique challenges of patient identity and record matching is required. The ability for a patient to manage his or her own credentials if they wish to promote or even prevent exchange is necessary. Ultimately, we need a design that doesn’t rely solely on a set of individual attributes to properly identify or match the patient (I refer to the oft-cited “Maria Garcia in Harris County, TX” study.)
We need visionaries at the table who understand the nuances and challenges and can chart a new path forward. We need to be looking at the role of existing forms of patient identification such as insurance cards, driver licenses, passports, smart cards, and biometrics to assist in the process. National identity standards and concepts such as OpenID and NIST’s Levels of Assurance are paramount to the design. Finally, peer-reviewed pilot studies that reveal the strengths and weaknesses of different approaches will help ensure the best ideas rise to the top.
Catherine Schulten is director of product management with LifeMedID of Citrus Heights, CA.