Full EHR interoperability has recently received a lot of attention as a core attribute of a smoothly functioning national health care system. This is appropriate. However, interconnection may be of secondary importance compared to a related topic, namely the extent of a patient’s complete medical history that exists within the electronic record. Much less attention has been paid to this `extent` — I think that implicitly, we naturally tend to presume that an extensive prior history is part and parcel of our existing EHR file. So imagine my utter surprise when two months ago I had occasion to check my Epic record on the Yale system. Basically, there is next to nothing in the file, with only my records corresponding to billed procedures within the system after Epic went live, that is, from 2012 forward. This content is comprised of data on a routine colonoscopy, and on a routine minor gynecological procedure. Otherwise, there is an `empty` report from a physician whose name I know, whom I have never seen, plus a (mistakenly placed) mammogram that is not mine, but rather, from a patient of mine. That is the entire record; there is no prior historical data about me prior to March of 2012, except for a few brief comments on my familial history related to colon cancer and/or polyps. From the system’s perspective, it was as if I were born at the start of Epic interactions, that is, at age 60. These records provide no mention of my childbearing history, or of an operation to remove my gall bladder many years ago. Most crucially, there was no whiff of history about my two primary medical concerns, namely (i) infrequent but severe episodes of vertigo; (ii) my unusual baseline EKG, with ST segment elevations and flipped T waves. In the absence of comment, this EKG pattern often suggests ischemia and a significantly elevated risk of a heart attack. Indeed, I carry a paper copy of my EKG in my wallet so that in an emergency situation, a treating physician would be apprised that it is my stable, baseline pattern. Again, there is no notation of this anywhere in my Epic electronic record. Now, a relatively complete medical history for me does exist — in my primary physician’s office. Alas, my physician uses a separate EHR, entirely outside of any hospital-based system. In fact, my primary physician, like many private practitioners, utilizes an ambulatory EHR that is local in the most restrictive way, exclusively used for interoffice documentation only, and adopted primarily to comply with Medicare requirements for reimbursement. So until the day comes when every EHR large and small is interconnected, localized ambulatory records included, I expect that my most critical medical history will be unavailable electronically anywhere outside of my primary physician’s office, under present day protocols. For example, if I were on vacation far from home and had a medical emergency, I would be out of luck, unless I carried my own personal records with me.

In retrospect, perhaps I should not have been so surprised to see a scant electronic history on file, given the length of time that I require to satisfactorily abstract my own patients’ charts into Epic’s system. Yale’s EHR, like most commercially available systems, was built to a large extent to avail itself of (cash in on) Meaningful Use (MU) dollars in the Affordable Care Act. It is a direct and inevitable consequence of the MU incentives that nearly all of the large-scale hospital-based EHRs are billing-centric, rather than patient-centric in any broad sense. Since physicians are not reimbursed to abstract and upload histories to non-user friendly computer forms, I expect that scant personal histories exist on electronic files throughout most hospital systems, except for records or data that are directly pertinent to billed procedures performed within the last three years. Unfortunately, I can confirm that minimal personal histories are standard issue in electronic records from outside my office that cross my laptop.

However, I cannot overstate the importance to a physician of a relatively complete patient history, especially if the doctor is unfamiliar with the patient, and particularly in emergency situations. So this issue of largely unpopulated histories for patients in most hospital EHR systems requires our utmost and timely attention and resolution, if we ever want to approach the full potential of interoperability. Finally and furthermore, one of the selling points of the integrated electronic record has been the ultimate potential for big-data analysis of huge caches of medical records of a scope that is presently beyond us, to help us to develop new cures for diseases or to discover new uses for drugs. If we want to do the data-mining effectively to meet the hype and the hope of this promise, we will require strong patient histories to maximize our chances to achieve biologically plausible significant results and greater understanding.

Given that it is unlikely that any EHR will contain even most of your complete health history, the question remains as to how best to ensure that any medical provider has all of your essential background. Last June, Melinda Beck wrote a very nice survey article on this topic in the Wall Street Journal entitled `How to Take Charge of Your Medical Records and Health Information`. A number of her points are quite useful, and bear repeating. You have the right, under federal law, to obtain copies of your medical information from virtually any place you receive health-care services, including doctors and hospitals. By obtaining your records, you can help circumvent “data lock,” serve as your own data hub and give out information when you are consulting specialists, seeking second opinions, or sharing data as you see fit, for example with family members or close friends, to be accessed in an emergency. If you spot errors, you can request changes or add information to make the files more accurate. Also, about half of all Americans, including all Medicare and Veteran’s Administration patients, can access at least some of their health records, free, through the government’s Blue Button program. Several recently developed products allow consumers to create personal health records that are organized in an easy to understand way, often for little or no cost. These encompass apps, software programs and web-based platforms that can help patients manage and integrate their medical and health records, and include iBlueButton (based on the VA’s Blue Button initiative), Microsoft’s HealthVault, and WebMD’s Personal Health Record. To date, relatively few patients have accessed their records; many more of us may want to take advantage of these strategies.

However, I still have some major concerns here. The patients who would be most likely to act on these suggestions generally tend to be those who are already basically healthy, and more computer-facile. My worries here are about the significant segment of patients who are not especially comfortable with technology, those with lower health literacy, and those who are either seriously obese or present with more complex medical regimens. From a public health viewpoint, this sector of our populace is likely the most in need of excellent medical histories, since proper treatment is often more multi-faceted. From a financial perspective, these patients will often accrue the greatest costs to the system. Yet it seems unlikely that many such individuals would routinely go through the effort to obtain accurate, portable computerized personal health records. Furthermore, I am also quite concerned about the possibility of selective editing of one’s electronic records, either to omit records that the patients deems sensitive or embarrassing, or for more `nefarious` reasons. Omitted or altered records could lead to misdiagnosis and improper treatment, with possibly serious health risk to the patient. Finally, more generally, both privacy and hack-proofing must be assured, as best as possible.