The demise of the hospital discharge summary

There was a time within my professional lifetime when old records would arrive on the medical ward in a wheeled wire basket, multiple volumes with shabby pages. And in the VA of the 1980s, which still had veterans of World War I who never lived anywhere else the previous half century, the forklift could deliver not only the charts but the x-ray films that had not yet been sent to the processor to recapture the silver they contained.

Residents on call in that era had no place to go until the next night with an obsessional morning report internist who should have been put on something ready to pounce at the next day’s intake conference, so selective and sometimes random perusal of past admissions made the efforts of the record room staff appreciated. Attending physicians and specialty fellows did have a time when they needed to depart the parking lot with a patient note placed into the current chart before that, so many of us became rather adept at wading through a series of records, most of which had the interest of the phone books that they resembled.

But amid the clutter, we learned to seek out each discharge summary, that succinct distillation of history, physical findings of uncertain accuracy, the salient diagnostic results, a paragraph noting uneventful days on the wards or misadventures bringing the patient to the ICU, and that all important medication list. Sometimes they were dictated while fresh in mind as the discharge prescriptions were being written.

Perhaps more frequently they were dictated a month later under duress when the records room staff in cahoots with the credentialing committee threatened to suspend the responsible physicians’ clinical privileges. Whether done in a timely way or the attending physician just set a two-hour appointment with himself to engage in a dictation orgy, physicians understood that these had to be done in a meaningful way, as the people doing the dictations were the same people doing the ongoing chronic office care. Whatever information the transcriptionist provides, even with typos that escape detection before the document becomes ossified by signature, formed the reference needed for future care to proceed in an orderly way, so they were taken seriously as a task needing the same attempt at excellence as the rest of the hospitalization.

That certainly does not appear to be the contemporary reality of more high-tech medical record keeping. Discharge summaries that I read seem more vacuous. The reason for hospitalization copied and pasted from the admission resident’s history and physical, populated boxes of normal funduscopic exams that were never really done, a blurb about the admission lab but no mention of how medical care altered those findings as the hospitalization proceeded, and too many hospital course summaries that go this consultant advised this and that consultant advised that but the word diabetes not appearing anywhere in the hospital course even though there were no glucoses under 250 mg/dL while under hospital observation. I used to receive courtesy copies of summaries sent passively to the office on my established patients even when I was not needed as a hospital participant. The last of these arrived about five years ago. I cannot honestly say I feel deprived in any way for not having them in view of the paltry value of most of the ones I read when I do hospital consultations.

One problem might be that the person doing the summary has no serious stake in the patient’s future. Getting a note on the chart to be included in the packet that the patient receives on discharge becomes its purpose. Sadly, the responsibility to professional colleagues that we once had to enable consultants or primary physicians or the next group of physicians who well be called upon to take charge at the next admission seems to have fallen victim to the administrative pressures that set promptness as a higher value than either thoroughness or thoughtfulness.

Still, it mystifies me that a patient can have wretched glucose control and multiple end-organ diabetic sequelae. Yet this devastating condition does not merit a word of recognition, even though obviously relevant to any post-hospital care to say nothing of the next hospitalization. Maybe the often deficient or even misleading hospital course paragraphs trace their roots to the author not being personally involved in the hospital course. In our era of handoffs and real-time writing displacing a more recognizable physician in charge and delay of document preparation that allows reflection in what actually happened to the patient, the author of the discharge summary might have no other resource than sound bite capsules of what each consultant put in their recommendation section. By doing that, though, he or she makes a transition from doctor to company scribe, and not a terribly accurate scribe at that.

So if we no longer can depend on the discharge summary to guide us, might we have a decent surrogate for what a pithy but appropriate summary once did? We return to that ornery attending internist belittling house staff at morning report. Invariably that physician was expected to know the lab with its trends and the x-ray results by the next day. While the discharge summary had little detail, when done well it made a superb index for finding just the right pieces of paper in a massive record cluttered with irrelevancy.

If the summary referenced x-rays, the reports or even the films can be retrieved. If the hospital course noted anemia, the studies on anemia can be selected out. We no longer depend on the summary as our index of where to find detail. Computerized retrieval will put all the CBCs in sequence, all the x-rays in another place sorted by type of study, all the consultant reports come up with a few clicks of the mouse. If the summary neglects to mention diabetes, no matter. A fully sequential accounting of venous and capillary glucoses lets me make my own judgment on diabetic control instead of depending on the impression of the previous attending physician.

So while the ability to categorize and retrieve prior information had been more than adequately replaced those, 5 to 10 minutes of forceful contemplation of what actually happened to a patient while under hospital observation seems to have been sacrificed in the name of expediency. Recovery of this once cherished element of medical care does not seem to be returning any time soon.

Richard M. Plotzker is an endocrinologist.