The Drift
Most problems in healthcare do not begin with error. They begin with momentum.
A patient presents with symptoms that fit a reasonable working diagnosis. Treatment begins. There is partial response. A new symptom emerges. A medication is adjusted. Another is added. A referral is placed. At each point, the decision makes sense. It is defensible. It is often thoughtful. No single step is reckless.
And yet, over time, something subtle changes. The original diagnostic frame is rarely rebuilt from the ground up. It is carried forward.
I first became acutely aware of this in psychiatry, but the pattern is not confined to mental health. It appears wherever care is delivered in short intervals under structural constraint. Follow-up visits are compressed. Documentation is optimized for continuity rather than synthesis. Language from prior notes is copied forward because there is not enough time to reconstruct the narrative from scratch. Each visit advances the plan slightly, but few visits revisit the foundation.
The result is not catastrophe. It is drift.
What makes drift difficult to name is that it does not feel like failure. Symptoms are not entirely untreated. There is usually a medication regimen in place — sometimes a complex one. There may have been periods of stability. The problem is not that nothing has been done. The problem is that what has been done has rarely been interrogated systematically. Psychiatric medication management in this context does not resemble a controlled trial. It resembles accumulated improvisation. Each individual adjustment was reasonable. The cumulative picture is harder to interpret.
Any physician who has tried to reconstruct a longitudinal treatment history from a fragmented chart recognizes the experience. You open a record that spans years. Medications have been started, stopped, restarted, augmented. Diagnoses appear stable in the header but feel provisional when you read closely. Response to prior trials is inconsistently documented. It is not clear what truly worked, what partially worked, and what failed.
You are responsible for the next decision, but the prior reasoning is opaque.
That opacity is rarely the result of indifference. It is the byproduct of how care is structured. We practice inside systems that reward throughput, continuity, and activity. Adjusting a dose fits within a fifteen-minute visit. Adding a medication is measurable. Referring out demonstrates action. Rebuilding the entire diagnostic formulation does not fit neatly into most outpatient schedules.
Even in inpatient settings, where time is theoretically protected, reconstructing years of accumulated decisions can feel like clinical archaeology. Charts are thick but incomplete. Assessments have hardened into labels. Medication lists reflect layers of prior attempts, but the logic behind those attempts is not always retrievable. If clarity is difficult to achieve even during hospitalization, it suggests that the opportunity for deliberate reassessment was deferred for too long.
Drift does not mean clinicians are careless. It means we are working within architectures that do not routinely protect time for reconsideration. Over time, incremental decisions compound. What began as a working hypothesis becomes assumed fact. What began as a trial becomes permanent. Escalation begins to substitute for clarity, not because we prefer it, but because it fits within the available structure.
For patients, the consequence is often subtle but persistent: partial improvement without full understanding. Treatment plans that expand but are not fully synthesized. Diagnoses that no one has explicitly revisited in years.
For physicians, the consequence is quieter but familiar: the cognitive strain of making consequential decisions without complete visibility into what has already been learned.
Adequate reassessment requires something the standard outpatient model rarely allocates: protected time with the explicit purpose of synthesis. Not a follow-up appointment that advances the plan by fifteen minutes — a deliberate reconstruction. Reviewing what was tried, what the response pattern actually was, whether the working diagnosis still accounts for the current presentation, and what a rational next step looks like given the full history rather than just the most recent interval. That kind of diagnostic clarity in psychiatry is seldom formalized outside of academic centers or inpatient settings. And yet it is not exotic. It is simply inconvenient for systems optimized for throughput.
If reassessment requires crisis to justify it, then we have misplaced its timing. Drift is not inevitable. It is predictable. And predictable patterns in medicine are rarely solved by individual effort alone. They require design.
The question is not whether we care enough to prevent drift.
The question is whether our systems are built to interrupt it before it becomes entrenched.
Building a practice organized around that interruption is not a departure from medicine. It is a correction. The diagnostic pause — the deliberate reconstruction of what has been tried, what has worked, what has been assumed without adequate verification — has always belonged to careful clinical thinking. It has been structurally excluded from most outpatient models, not because it is impractical, but because it is slow. Designing a setting where slowness is the point changes what is possible. A practice built for reconstruction rather than continuation, outside the pressures of institutional throughput, creates room for the kind of diagnostic clarity that many patients with complex psychiatric histories have never actually received. For patients in Virginia working with a telehealth psychiatrist or seeking a more deliberate outpatient model, that clarity is not a luxury. For most, it is simply what has been missing.
