The numbers on physician documentation time have become a familiar refrain in healthcare technology discussions: physicians spend somewhere between 1 and 2 hours per 8-hour clinic day on documentation, with a significant portion happening outside clinic hours. The range varies by specialty, practice type, EHR, and how you measure it, but the direction is consistent across surveys and time studies going back to the widespread adoption of electronic health records in the 2010s.
This article is less about the aggregate statistics and more about the specific ways that documentation time is lost - because understanding where the time actually goes is the prerequisite to knowing whether a given tool will recover any of it.
Between-patient documentation
In an ideal clinic day, a physician completes a note between seeing patients - patient finishes, note is typed while the next patient is being roomed, physician calls up next patient with a clean slate. This happens. It also often doesn't happen, for predictable reasons.
In primary care, "completing a note" includes not just the narrative SOAP documentation but also order entry, referral letters, medication changes, preventive care documentation, and any required quality measure documentation. A 15-minute visit can generate 8 to 10 minutes of post-visit administrative work. If the physician is scheduled at 15-minute intervals, this math doesn't work - there isn't time between appointments to complete everything before the next patient is ready.
The documentation debt from overbooked schedules is one of the primary drivers of end-of-day and after-hours charting. It's not that the physician deferred documentation because they were unmotivated; they deferred it because completing it in the available window was not possible.
The false efficiency of the EHR
Electronic health records were supposed to make documentation faster. In many respects, they made it more thorough but not faster. The move to EHRs added structured data capture requirements that paper charts didn't have: required fields, drop-down selections, coded diagnoses, order sets, and payer-specific documentation elements. Each of these individually takes seconds. Collectively, across 20 encounters in a day, they add up to a non-trivial documentation overhead that didn't exist in the paper era.
EHRs also created the conditions for pajama charting - physicians completing notes on laptops at home after the clinic day ends. Paper documentation required being at the practice to access the chart; EHRs are accessible from anywhere, which is convenient and also removes the natural stopping point that physical location used to impose. When the chart is accessible from home, the temptation (and often the expectation) is to finish it from home.
Cognitive fragmentation during documentation
Documentation time is often underestimated because it doesn't present as a single continuous block. A physician might spend 3 minutes charting after Patient 4, get interrupted by a nurse question about Patient 2's medication, spend 90 seconds on that, return to Patient 4's note, spend another 4 minutes, get a lab result notification for Patient 7, review it, add an addendum to that note, then come back to Patient 4. The total documentation time for Patient 4 is accurate but the experience is fragmented.
Cognitive fragmentation has a cost beyond the raw time spent. Each interruption requires a brief re-orientation - remembering where in the note you were, what you were about to write. These transitions are short individually but they compound, and they leave physicians feeling at the end of the day that they spent the entire day documenting even when the actual documentation time was not as large as it felt.
This is part of why physician time diary studies often show lower documentation time than physician self-report surveys. The experienced burden of fragmented documentation is larger than the measured time would suggest.
Documentation that doesn't count toward patient care
Some portion of physician documentation time goes not to clinical documentation but to administrative documentation: prior authorization letters, peer-to-peer reviews, appeal letters for denied claims, and payer-specific forms that require physician attestation. This documentation is often burdensome precisely because it is disconnected from clinical care - it is paperwork generated by the insurance system, not by the physician-patient relationship.
Ambient documentation tools do not directly address prior authorization work. They address the SOAP note and clinical documentation that records what happened in the visit. Practices experiencing high documentation burden from administrative and payer-related work will not see a proportional reduction from ambient documentation alone; that problem requires a different category of solution.
The late-day concentration effect
Documentation burden tends to concentrate at the end of the clinic day, not distribute evenly through it. This happens for structural reasons: the morning is easier to stay current because there is less accumulated backlog; as the day progresses, documentation debt accumulates and the afternoon becomes harder to stay current in. By late afternoon, many physicians are running one or more notes behind their current patient.
The experience of facing 5 or 6 incomplete notes at the end of clinic - knowing that they need to be finished before the day ends, and that completing them will take 45 minutes to an hour - is qualitatively different from spending 3 to 4 minutes on a note immediately after each visit. The concentration effect amplifies the psychological burden of documentation beyond what the raw time would suggest.
This is why the primary promise of ambient documentation is not just fewer hours spent but a different distribution of documentation time - smaller, immediate review moments at the point of care rather than a long deferred block at the end of the day. Practices that have implemented ambient tools consistently report that this pattern change is one of the most valued aspects of the tool, independent of total time saved.