I have been practicing medicine since 1985. With good fortune, I hope to continue practicing for several more years.

When younger physicians ask what medicine was like when I began, I am careful not to romanticize the past. The pre-computer era had many shortcomings. Paper charts were often unavailable. Records were fragmented. Laboratory results could be misplaced. Communication was slower. Finding information could be frustrating. Many aspects of modern medicine are unquestionably better.

Yet after four decades in practice, I find myself reflecting on something that is difficult to measure. We are very good at measuring what technology has allowed us to gain. We are less skilled at measuring what may have been lost.

One of the books that has influenced me most is Technopoly by Neil Postman. Published in 1992, it raised a question that seems even more relevant today than when it was written. Every technological advance creates benefits, but it also creates costs. The gains are usually obvious and celebrated. The losses are often subtle and difficult to see.

Medicine has experienced extraordinary technological growth during my career. I practiced nearly twenty years before computers became a routine part of the examination room. Now I practice another twenty-plus years as medicine became increasingly computer-mediated. That experience feels like a natural experiment. I have worked in two different worlds.

The first world had no electronic inbox, no patient portal messages, no quality dashboards, no electronic alerts, no mandatory acknowledgments, no electronic refill queues, and no endless streams of notifications. The second world has all of those things. Again, this is not an argument that the old world was superior. It is simply an observation that the arrival of computers did not merely make existing work easier. Computers enabled entirely new categories of work.

The promise was often that technology would improve efficiency. In many ways it has. But efficiency has frequently been accompanied by expansion. As new capabilities appeared, new expectations appeared alongside them. Once a task became possible, it often became required.

Computers did not only help us perform the work of medicine. They allowed us to create additional layers of work around medicine. Perhaps this is not surprising. Institutions naturally seek measurement. Regulators seek documentation. Payers seek verification. Lawyers seek records. Once computers made extensive documentation possible, extensive documentation gradually became expected.

The challenge was not that any single requirement was unreasonable. Most were created for understandable reasons and often with good intentions.

The challenge was accumulation.

Computers made new forms of documentation, measurement, communication, oversight, and reporting possible. Over time, each was added, while few were removed. Portal messages became expected. Quality dashboards became expected. Electronic reminders became expected. Documentation expanded because electronic storage made expansion easy. Data that once would have been unavailable became continuously available, and therefore continuously reviewable.

Viewed individually, each change may seem modest. Viewed collectively, they consume a growing share of clinicians’ attention. And attention is not an unlimited resource. Every minute spent managing the electronic representation of a patient is a minute not spent with the patient.

The challenge, as I see it, is not to reject technology. That would be foolish. The challenge is to ask whether we are using computers for the things computers do best. Computers are extraordinary at storing information, retrieving information, transmitting information, and making information available where it is needed. Human beings remain better at listening, better at understanding context, better at navigating uncertainty, and better at helping another person make sense of illness, fear, hope, and suffering.

I sometimes wonder whether we should be more willing to push back, not against technology itself, but against the assumption that every measurable process deserves measurement, and every measurable activity deserves documentation. Recently, I was trying to find a single biopsy result in an electronic record. I opened four separate online documents, each more than fifteen pages long. Despite all that documentation, I still struggled to locate the actual result. This would have been unimaginable in 1995. We had less information, but sometimes the signal-to-noise ratio was better.

We have gained capabilities that would have seemed miraculous when I graduated from medical school. But we have also accumulated layers of complexity, documentation, measurement, and process that consume attention. The purpose of medicine is not the creation of records. The purpose of medicine is the care of patients. Documentation, measurement, compliance, and reporting are means to an end, not the end itself.

As I grow older, I find myself valuing institutional memory. Organizations need people who remember what came before, not because the past should be restored, but because memory helps us recognize what has changed.

I am not arguing for a return to 1995. But if I am honest, there are parts of that era I miss. I miss shorter notes. I miss narrative thinking. I miss an environment in which technology supported clinical care without so often competing for clinical attention.

Perhaps every generation believes it lived through a period of transition. Perhaps that is true. My purpose in writing this is not to complain, and certainly not to suggest that medicine was once perfect. It was not. Rather, I simply hope to bear witness.

I have practiced long enough to experience two distinct eras of medicine. I have seen what was gained. I have also seen what was lost. Both deserve to be remembered.

Mark McConnell is an internal medicine physician.