In America we don’t have a unique number assigned to the health of each person in the country. We also don’t have a unique identifier for each person’s finances. We don’t have unique identifiers for education, retail, transportation, hospitality or any other industry. Commercial service providers usually create internal identifiers for their own customers in the course of normal business to facilitate accounting and other business processes. These identifiers are usually called “account numbers” or IDs. You have bank account numbers, utilities account numbers, student IDs, member IDs, and sometimes you have multiple types of identifiers issued by one entity for different purposes.
Federal, State and local governments issue their own identifiers, usually for taxation purposes, licensing operations, administration of entitlements, and on rare occasions for the purpose of drafting citizens into military service. Commercial entities are currently piggy backing on several types of government identifiers and government issued ID cards to validate identities of their customers. For example, practically all financial institutions, insurance companies and utilities require the use of social security numbers (SSN). Commercial airlines rely on government issued driver licenses and passports for security purposes. Business purchasers require tax identification numbers in order to transact with other businesses. And the list goes on.
As far as I can tell, there is no crisis in mis-identification of bank account holders, where fortunes flow erroneously from one account to the next on a daily basis. Frequent-flyer miles do not accrue to random people. Credit card charges don’t randomly appear on people’s cards. Even hotels and grocery stores seem to do a pretty good job at accounting for reward points. Yes, there is a certain level of fraudulent activity in every market and, yes, it is possible that a national identifier for shoppers could curb some of the fraud, but I am not aware of any Chamber of Commerce lobbying for such identifier to be created.
Healthcare is different. The healthcare industry is clamoring for government intervention to create a national health identifier that is attached to every resident of this country from birth to death and beyond. We need to emphasize here that it is the industry that is pushing for nationwide machine tabulation of people’s health and not individual patients, doctors, nurses, or other caregivers. It seems pretty clear that national health identifiers are needed to create a seamless national interoperability infrastructure for surveilling, locating, collecting, querying, and exchanging health information. Perhaps less clear are the reasons why we would need such voluminous infrastructure.
The canned messaging from industry is mostly about patient safety. As restless Americans “move across the country” we must be able to retrieve their medical records from across the continent, or risk harming them for lack of data, particularly in emergency scenarios where patients are unable to relay any information to clinicians attempting to provide life-saving services. Thought leaders who are captains of industry, or high ranking government officials, can easily imagine themselves in such situations, seeing how they travel frequently from coast to coast for important business meetings and an interminable stream of healthcare conferences. It is also not unusual for the revolving door set to move from the suburbs of Washington DC to wherever the next big gig materializes (and vice versa). The rest of us are leading slightly different lives.
According to the U.S. Census Bureau, 36 million Americans over the age of 1 changed residence last year. Of those, 63% stayed in the same county, 81% stayed in the same state, and only 6.7% moved to a different census region. Younger, and therefore healthier, people make up the bulk of movers.
How about travel? According to the Bureau of Transportation Statistics almost half of all trips are less than 100 miles roundtrip, with 80% of all travel, business and pleasure, occurring within a radius of 250 miles. And finally, for almost 60% of us, the state we live in today is the state we were born in yesterday. So why, oh why, is it of such great importance that our perfectly identified medical records crisscross the nation at the speed of light, mixing and matching every bit of information from every imaginable and unimaginable source, while we all stay put in our little habitats?
Just like your SSN is being required and used for myriad commercial applications which have nothing to do with the public pension fund, you can expect your national health ID to be required by every “wellness” app, every employer “wellness” program, every CVS rewards card, every school, every background check, and whatever else you can think of (for your own safety, of course). This is the missing piece of the national learning system puzzle, where health research (some of it medical perhaps) is conducted on hundreds of millions of oblivious people all day every day. You can’t have precision health if you can’t precisely identify your research subjects, and precisely connect all available data (genotypes and phenotypes in the broadest sense) to each one, no matter where he or she is located at any given moment.
Local medical systems do generate patient identifiers, and always had. Regional health information exchange organizations have pretty reliable algorithms to match patients, usually based on five pieces of demographic data, which can yield almost perfect accuracy levels. Resolving the outliers manually is rarely an insurmountable problem when you work with relatively small populations and limited data sets. However, when you do “research” with big data generated 24/7 by over 300 million people, using thousands of modalities and millions of data points, the story changes in a hurry, particularly if your research includes real time interventions to modify subjects’ behavior. A reliable unique identifier becomes imperative to ensure correct attribution to subjects, completeness of information for each one, and responsiveness of experiment tools.
If we restrict our discussion to medicine, instead of the full spectrum of industries dealing in health, there are emerging business models that may very well depend on the establishment of a national health identifier. Direct-to-consumer telemedicine is one of those Christensen disruptive innovations, entering the market with a product that is barely functional, but good enough and cheap enough for the increasing numbers of customers who can no longer afford a traditional doctor visit. Virtual doctor visits are dislodging medicine from its traditional local and personal roots. There is no material difference in quality between an online visit with a doctor across the street and a doctor across the country or across the ocean.
The telemedicine lobby is slowly but surely dismantling state licensure requirements for physicians, and any remaining professional objections to Internet medicine. Diagnostic tools are increasingly being ported to iPhone form and to a variety of cheap Internet connected gadgets targeting consumers directly. It stands to reason that eventually some type of physical examination will be available for online visits, and the range of medical services provided via telemedicine will move beyond scrapes and bruises or minor viral infections. At that point, teledoctors will need medical records to expand scope of services.
A national health ID should allow the telemedicine industry to do to traditional medicine what Southwest Airlines did to commercial aviation, i.e. make it cheap, crowded, cramped, exasperating and widely accessible, before prices start climbing again.
Telemedicine is a great solution for the unwashed masses, unless all evidence-based watchful waiting alternatives were exhausted and they end up needing some sort of procedure. Historically, patients would be directed to local facilities that contracted with the insurance company serving the patient, but a more cost effective alternative seems to be emerging for large national employers who need not be bound by geographical limitations. Instead of negotiating locally, a large employer could negotiate nationally with “centers of excellence” to ship its employees to wherever the procedure price is lower. This is the precursor to another business innovation – the focused factory, where medical facilities specialize in a narrowly defined service, such as hip replacements, and are able to utilize best manufacturing principles to provide standardized quality at the lowest price.
Obviously when you start shipping people for repairs across the nation and perhaps across the globe eventually, you must have a reliable national health identifier to attach medical records to the patient. Also, employers or health insurers will need good data to evaluate and select focused factories for their respective businesses. And, since patients will be coming from all over the country in a variety of health conditions, proper analysis of expenditures and savings will need to cut across employers, insurers, factories, and geography. It would be practically impossible to construct a good analytic model without detailed health data for each patient and without a unique identifier to facilitate inclusion of big data from all available sources.
What about Privacy?
In summary, once you shed the traditional lens through which we studied medicine for centuries, the need for a unique national health identifier for consumers becomes crystal clear. Modernizing healthcare – and bringing it into the 21st century sharing economy – cannot happen without a solid analytic infrastructure, and such infrastructure cannot stand without precise longitudinal identification of both providers and individual consumers.
Another traditional lens that we will need to redefine is the lens of privacy. As the health system integrates into virtually Uberized global markets, privacy becomes a localized phenomenon. A democratized health system will place control of health data in the hands of the uniquely identified individual to share with or to withhold from other uniquely identified individuals as he or she chooses, while allowing the learning system unfettered access to all information.
To illustrate this seemingly paradoxical state of affairs, and to end on a lighter note: imagine that you just caught the cooties following a brief encounter with a gorgeous brunette during your recent health care conference in Vegas. The system will need to know about your symptoms, treatment, and outcomes, so it can learn—but as an individual you are fully empowered to lock down this portion of your medical record so your wife never finds out about your extracurricular indiscretion. You can maintain full privacy protections from those equal to you, while allowing the system to observe everything from above. Sort of like living in a house with sturdy walls and doors protected by security systems, but without any roof. If you never look up, you won’t know the difference.
Latest posts by Margalit Gur-Arie (see all)
- Democratizing Medicine - January 7, 2016
- The Imperative of a National Health Identifier - December 8, 2015
- The Effects of Digital Health on the Moral Universe - November 3, 2015