The Curse of the Medical Record
As always, the goal of this magazine is to increase efficiency of medical processes and to lower costs. One process in need of fundamental improvement is the timely recording into and retrieval of medical records. To make the point, I will synopsize a type of conversation I heard many hundreds of times during my career (thousands of physicians will be nodding their heads in agreement as they read this).
An established patient of mine has just been released from a hospital and is here to see me for “follow-up”. He was admitted because of breathing difficulty. He believes that he was diagnosed with “pneumonia” but a physicians at the hospital reportedly told him that he may have some “heart trouble” also. He was to see me as soon as possible after going home. I ask him if he was given any paperwork to summarize the details of his hospital stay and his new medication regimen. He says that he thinks that he got something like that but regrets that he probably lost it before today’s visit. I then ask him does he know, by any chance, the name of any new medications. He does not know any names but offers that one pill was green and possibly oval and the other was blue and circular. At this point, apparently noticing my deepening facial furrows, the patient reassures me that everything is on that hospital’s computer and “all you need to do” is to get an account on the hospital’s computer and I can retrieve any information which I deem necessary for continuity of care.
What to do, what to do.
A modern hospital record for a patient stay is often voluminous. It contains all the progress notes made by the various physicians/clinicians seeing the patient. It contains observations by the rest of the patient’s care team including nurses, dietitians, physical therapists, and so on. Every laboratory drawn, every image obtained, every procedure done, every EKG collected (this list is not exhaustive) are recorded. At the time of discharge, a summary of the patient’s stay and relevant findings are often made along with a list of his new medication regimen. However, the information is hidden on a sector of a disk drive in the hospital and not easily accessible (and this author has a degree in computer science).
The transition of care is 1 of the most critical processes in the medical world. However, in spite of its importance, my observation over the recent decades is that this process is actually faltering rather than improving. Depending on the rest the patient’s history and findings on examination at his follow-up appointment, I will have some decisions to make.
One key decision is how much time will I spend attempting to recover the patient’s information from the remote hospital. The 1st obstacle is a law called HIPAA (Health Insurance Portability and Accountability Act) which was created in 1996. Among other features, it “protects” the patient from unlawful disclosure of his medical information. Until the medical records department of the relevant hospital has received a written release from the patient which allows me to obtain records, it will not allow me to have any information about his stay even though I am his PCP (primary care physician) in the need these records for ongoing care. Assuming that the patient gets the written release to the medical record office, I may be allowed to get the written medical record or some subset of it in a week or so. This is not acceptable. Formally, I could get critical information from the treating physician at the hospital by attempting to discover who that physician is and then calling him/her. In general, this is less successful these days because the attendings of record tend to “cover” the patient for an ephemeral 1-3 days during the hospitalization and don’t necessarily recall details. As often as not, they’re not available for discussion anyway. If I’m lucky, that person can read me the discharge medications. If I am very lucky, maybe I will get a verbal report on chest x-rays and cardiac test such as an echocardiogram. However, with my waiting room full of patients scheduled to be seen, compromises are made and it may be necessary to repeat recently done tests. My desire to avoid costly waste is sacrificed at the altar of expediency.
My suggestions to remedy this situation are as follows.
In the United States, we should put our medical records on one central site accessible to each individual patient, all physicians, and care team members. That computer will be home to 300+ million orderly, contiguous medical files (one for each of us). The security needed for this is unprecedented and should be arranged by a competent entity (e.g. NSA or equivalent). No longer will the record be fragmented across dozens of sites e.g. PCP, specialists, hospitals, radiology facilities, blood collection centers, etc. If a patient sees me, I will generate my progress note for the visit and it will be downloaded in real time to the central record when I close my note. There, it is incorporated into the record. Later, it will be instantly retrievable by date and also by dictating physician. All the blood work will also be there and retrievable by date and time. If I order a blood test as an outpatient, it will be incorporated into the record and available to all credentialed users (more on that later). If the blood work is from an inpatient stay, it will also be seamlessly incorporated into the patient’s record in time sequence. Later, if I wish to see a trend over time (a typical example of which might be hematocrit to assess for anemia), I could easily pull out and display the last year’s (or any interval’s) hematocrit values. Echocardiograms (and indeed every procedure) would be recorded, with both text and the actual images depending on the needs of the reviewer. Current medication lists and allergies/intolerances would be instantly available. A properly credentialed physician could update both of these lists.
In the real world, the benefits would be unrivaled. I could see in real time all progress notes including the patient’s discharge summary from the physically remote hospital. This would allow me to hone my duties for the patient’s transition (does the patient need a follow-up chest x-ray, does the patient need to see a cardiologist as an outpatient, and so on). I wouldn’t have to guess about his discharge medications as I would have them instantly (indeed, if I saw the night before his visit he was on my schedule, I could look up his most recent information on the medical record in preparation). Each time his record adds new information, I would be notified as his PCP. Outpatient reports would be in patient’s record and I would be notified of the availability as soon as the record was incorporated into the medical record. Redundant testing done by 2 practitioners would be largely eliminated (I would estimate that this alone would save the United States tens of billions of dollars per year).
Additional benefits would be legion.
If a given physician was ordering twice as many MRI scans on average as other clinicians for a given population, he/she could undergo physician peer review and a cause discerned (poor training, conflict of interest [the physician owns an MRI center], etc.)
The PCP would no longer need to scan into his/her own local patient chart a copy of many labs, reports, consultations, summaries, etc. (I would also estimate that this alone would save tens of billions of dollars per year not having to pay employees for an unnecessary task).
If a given physician does 3 times more hysterectomies than his/her peers do for a given population, he/she would also be eligible for review (i.e., utilization review).
If a cluster of fentanyl overdoses developed in area of town, without disclosing patient identity, law enforcement could be notified of the uptick to allocate resources.
If a cluster of leukemia cases triggered a threshold, without disclosing patient identity, CDC and EPA could be notified.
Once again, this list is by no means exhaustive.
As noted above, the information described above would have to be available only to “credentialed” individuals. To obtain credentialing for access, a clinician would need to sign on the system with the unique password or even biometric marker for identification. The patient would need to give that clinician permission for his records to be read. Finally, the entity under which the clinician is working would have to be uniquely identified also (a physician might be working at 2 practices but the patient sees her/him only at 1 and thus the other would not have permission to see the patient’s chart).
If we are to undertake universal healthcare, a robust, intuitive, and timely chart will be necessary. Having dozens of charts generated during one’s lifetime spread across time and space, i.e. a patchwork of information, no longer needs to be tolerated. With this article, I believe I have outlined a tractable, practical, and foundational improvement of the medical system.