Every single time you borrow money—from student loans to credit cards or home mortgages and car loans—it all goes into several databases, all run by for-profit businesses who happily compile this data about you at no charge. The next time you want a new credit card, you apply online and in a matter of seconds, the application for credit is approved based on the lenders’ query of their preferred database(s). That’s database interoperability. The database owner makes money by aggregating huge amounts of data and charging lenders access fees and offering you a paid, premium level of access.
True interoperability for multiple hardware vendors should seem as simple as Wi-Fi connectivity. When you come to my house or that of any other friend or business associate, we simply tell you which of the many networks your device already sees and suggests is ours, and provide you with the password. If I don’t change my network password, the next time you come to my house (after the pandemic, of course) your phone or laptop will immediately, automatically connect to the network again. Most people can’t tell you anything about the steps involved in Wi-Fi connectivity except the password entry part. That’s seamless, password-guarded hardware interoperability.
Patient medical records, of course, must be far more flawless. While you can likely be sure you never had a Honda Civic loan from Wells Fargo if one appears on your credit record, most of us may not be able to read and understand every part of our health record, so errors would be hard to spot. And there is no guarantee that there will be time to appeal an error in your record even if you find it before a medical error occurs as a result of bad data. When it comes to personal health information and the mission-critical nature of reliance on that data, it’s clearly been wise to be very cautious about letting bad data into the system. It’s a fundamental place to stop medical errors—if my records reflect I am a diabetic when I am not (or the reverse), the treatment I receive would vary significantly, even dangerously, during a hospital stay.
Every time you go to a health provider they enter data in your Electronic Health Record (EHR). This should mean that somewhere up in the cloud (on a server) all your health data is organized and ready for the next visit, no matter who you see, why you see them, or where in the world you are located at the time of the visit. That’s how other data systems work. So why are EHRs failing to work as intended?
Each entity, from the software platform to the health insurer to the healthcare provider, has their own protocols, so the promise of the EHR has become the story of the Tower of Babel, with everyone speaking their own language and yet failing to communicate. While there’s a lot of talk about interoperability from both EHR makers and medical technology companies, it’s certainly not as timely and seamless as is the case with credit agencies and Wi-Fi. Added to that, health systems are hard enough to manage in and of themselves, with increasing pressure on their profit margins, so they can hardly be expected to align their data collection with that of every other healthcare provider in the community.
In an effort to manually bridge these territorial data gaps, we’ve got a lot of time-wasting data collection and re-input for every patient who sees a new doctor, whether due to geography or health issues. And a medical professional relying partly on a patient’s EHR data to provide the best possible care is building a treatment plan on a potentially faulty foundation, with information missing, mis-reported, or in other ways rendered unreliable. Knowing a database isn’t reliable, a healthcare provider must ask a patient or their family to waste time trying to get records sent from a former provider and while the system wastes money retesting to confirm data stored elsewhere.
Interoperability of healthcare data doesn’t need to be cobbled together this way. The fact is that non-patient players are guarding the data in the interest of some market advantage they perceive, but at the expense of the larger, overriding social goal of letting all clinicians work together in providing the best possible healthcare to patients over their lifetimes. And imagine what anonymized, aggregated views of this data might do to revolutionize public health alerts, surely a need we are all more aware exists post-2020.
Neither the patient nor the medical professional should be pushing or pulling data in and out of the system. Perhaps the patient should have to provide the permission to link these systems, but if so, only once. As with credit and Wi-Fi, I can set my security levels, and I am the one who grants permission for the linking of a new device with my network or a new potential lender with my credit record. If I end up in an out-of-town ER, too ill to reliably give my entire health history, I will need my health information made instantly available to the health professionals trying to provide the best possible care. And the pandemic has also made it clear many of us may need to communicate with medical providers long distance as they discuss incapacitated family members’ care, so perhaps we should have a way to permit long-distance caretakers to provide and acquire information within the EHR.
Since I live near several large, national health research institutions and have an interest in progress, I volunteer as a healthy control subject for a variety of research studies. While thinking about writing this post, I learned that my research data is all anonymized under a single federal Global Unique Identifier (GUID). If at some time that fMRI I did for the VA study in 2019 can be usefully reunited with blood I donated for a Parkinson’s Disease study at UCSD last month or DNA I provided to The WISDOM study some years ago, a researcher can cross reference that information. Each time I volunteer, the researcher checks to see if I’m in the database and unites my new, anonymized research records with my GUID. In each study I have always given permission for my data to be reused for any other study without my further permission. Perhaps some day research might choose to combine existing data like that and find a new, much-needed answer quite quickly. (Imagine the data possibilities if we could all also choose to link research to our complete EHR as well.)
We can easily achieve the same accessibility with our patient data in our EHRs—there is no technological reason why it doesn’t work as easily as a credit check. In all cases, I grant permission to my aggregated data. I choose the level of control I want over my data, weighing privacy with other considerations. Patient- and caregiver-provided information goes into a temporary holding place so healthcare providers remain able to effectively quarantine outside patient data until they verify it. There are many business reasons why this level of data sharing hasn’t been done in healthcare, but no patient care reasons exist. Patients and healthcare industry players haven’t demanded the free flow of patient data, but the patient, who has every reason to share their data as needed, is the logical gatekeeper. Data flow should be controlled by the patient, not the technology providers providing awkward, multi-step, labor-intensive ways of linking records. The cost of admission to the healthcare data world for all entities should be true, seamless interoperability. Only then can patients focus on their health, clinicians on providing great care of other people, while engineers and computer scientists work on new breakthroughs and innovations in medical hardware and software under unified interoperability standards.
To learn more about problems with the electronic health record from Susan, click here.