The Drawbacks of Data-Driven Medicine

"I have often beheld two such sages almost sinking under the weight of their packs."


In Gulliver's Travels, Jonathan Swift tells of a community of scientists seeking to improve human life through technological innovation. One of their many schemes aims to replace words with implements. Instead of speaking, people will carry backpacks laden with the items about which they need to communicate. This new communication technology will eliminate misunderstandings and also serve as a "universal language, to be understood in all civilized nations."

There is only one drawback to the plan, particularly among the wise. Namely, if a person's business is great and of various kinds, he is obliged "to carry a great bundle of things upon his back, unless he can afford one or two strong servants to attend him." In short, the burden of communication quickly becomes almost too great to bear. As Swift puts it, "I have often beheld two such sages almost sinking under the weight of their packs."

Swift's story sounds absurd, except to many who labor in contemporary health care. I was speaking with a medical student recently who expressed astonishment about by the brevity of physicians' notes in patients' charts at a small, rural hospital. "In our academic medical center, the notes are usually a few pages," she said. "But in the community hospital, they were sometimes only a few sentences."

"In school, we are taught that notes need to be thorough, leaving nothing out. To avoid lawsuits and ensure that you get paid for everything you do, you document absolutely everything. In fact, physicians often copy and paste whole paragraphs of information from one day's note to the next. So I asked one of the physicians, 'How can you get away with such short notes?'"

"He told me their concern was not to avoid lawsuits or get paid, but to take good care of the patients. They try to avoid communicating with each other through the electronic medical record and take every opportunity to talk with each other about the patients they have in common. The charts really only reflect what's needed to summarize those conversations."

These small-town physicians are on to something, a truth about medicine in particular and human communication in general that we, in our rush to take full advantage of the latest information technologies, sometimes forget. Simply put, the amount of information in the patient's chart is not necessarily positively correlated with the quality of the patient's care. Often the presence of more information is symptomatic of decreased levels of communication and understanding.

There is a big difference between extracting, recording, and retrieving large quantities of information and actually understanding what is happening. When the quantity of information gets to be too great, it actually begins to clog up the channels of communication, interfering with the physician's ability to discern what is really going on. In my experience, the best doctors are not the most verbose. To the contrary, they generally formulate cogent impressions.

We must guard against the temptation to allow the medical record to supplant the patient. A recent study at Johns Hopkins University indicated that hospital interns -- physicians at perhaps their most formative stage of training -- spend only about 12 percent of their time interacting with patients. By contrast, they spend 40 percent of their time -- more than 3 times as much -- interacting with hospital information systems. The flesh-and-blood patient is getting buried under gigabytes of data.

This shift in attention can have truly Swiftian consequences. An intern recently presented a newly admitted patient on morning rounds, reporting that the patient was "status post BKA (below the knee amputation)." "How do you know?" the attending physician inquired. "It has been noted on each of the patient's prior three discharge notes," replied the intern, looking up from his computer screen. "Okay," responded the attending physician. "Let's go see the patient."

When the team arrived in the patient's room, they made a surprising discovery. The patient had two feet and ten toes. Where did the history of BKA come from? It turned out that four hospitalizations ago, the voice recognition dictation system had misunderstood DKA (diabetic ketoacidosis) as BKA, and none of the physicians who reviewed the chart had detected the error. It had now become a permanent part of the electronic medical record -- as if written in stone.

This sort of thing happens all the time. We become so enamored of our new technologies, and so keen on taking full advantage of them, that we begin to pay less and less attention to the real purpose behind what we do. It is like pilots navigating by instruments on a bright sunny day, or motorists who follow the instructions of their GPS devices right into a lake. During a clinic visit, we physicians are so preoccupied interacting with our tablet computers that we fail to see the patient.

No one knows yet whether electronic medical records improve the quality of care we provide to patients. When they work well, they make patient information more accessible, up to date, and less prone to misunderstandings due to illegibility of handwriting and the like. But there are problems, and they extend far beyond added costs, incompatibility between systems, and increased physician workloads. Above all, they often short circuit genuine understanding.

Good medicine does not mean grinding diagnoses and treatment plans out of large collections of facts. It means understanding the patient. And there are many facets of the patient that tend not to shine through if the electronic medical record is all we have to go on. A robot can collect and process data, but good medicine requires more than information management. It takes effective communication, deep insight, and genuine compassion.

Today health care is awash in information. We have more data on every patient than ever before. But consider these lines from T.S. Eliot: "Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?" Data alone are insufficient. In fact, data alone are little more than noise. What we need most are not data on but genuine knowledge of our patients. And even the most sophisticated information systems cannot supply the wisdom to tell the difference.