Where Your Documentation Hours Actually Go: A One-Week Audit

Ask a therapist how much time documentation takes and you get a number — an evening hour, a working lunch, a lost Sunday. Ask where that time goes and the answer gets vague. The total is easy to feel. The composition is invisible. And the composition, not the total, is what decides the fix.

“Documentation time” is really four different kinds of time: writing, searching, switching, and re-reading. A better template shortens writing but does nothing for the minutes spent hunting for the last progress report. A quieter schedule cuts switching cost but will not fix a note that takes twenty minutes to compose. The audit below separates the four in one ordinary week, using nothing more than a paper tally card — so that whatever you change next is aimed at the category that is actually consuming the hours.

Writing time is the part everyone pictures: composing the subjective line, the data, the interpretation, the plan. It is real work and it is the smallest surprise in most audits. Searching time is everything the note needs that is not in your head — the active goal, last session’s cueing level, the authorization visit count, the evaluation report — retrieved from wherever it lives. It is not writing; it is fetching.

Switching time is the re-orientation cost paid every time documentation starts, stops, and restarts: the note opened between sessions and abandoned, the chart reopened at 8 p.m., the thirty seconds of “where was I” that precede every resumed draft. Each individual switch is small, which is why it never gets counted and why its weekly total is so often the audit’s surprise. Re-reading time is reconstruction: when a note is written hours after the session, the details have decayed, so you re-read prior notes, your own half-finished drafts, and the schedule just to remember what happened. It is the interest paid on deferred writing.

None of this is a personal failing, and the burden is not imagined — reducing clinical documentation load is now an explicit national goal. What the federal strategy and the informatics profession cannot do is tell you where your particular hours go. That takes a week of your own data.

The protocol is deliberately low-tech: a paper card and tally marks. Time-tracking apps fail here for the same reason evening notes fail — they add a switching cost to the thing being measured, and they get abandoned by Wednesday. A card next to the keyboard survives a real clinical week.

1. 01

   Pick an ordinary week

   No holidays, no PTO, no evaluation pile-up you already know is unusual. If the week turns strange midway — two sick days, a crisis — finish the audit anyway and note it; a strange week still produces a real split, just a wider error bar. Do not wait for a perfect week, because there is none.

2. 02

   Rule the card into four rows

   Writing, searching, switching, re-reading. Define them before Monday, in your own words, with one example each — the definition argument you have with yourself on Tuesday is a data-quality problem. Add a fifth row at the bottom: after-hours, tracked by location and clock, not by category.

3. 03

   Tally in the moment, in minutes

   Each time you finish a documentation touch, jot the minutes in the right row — a ten-second act. Estimating a morning at lunch is already reconstruction, and reconstruction flatters the writing row: searching and switching evaporate from memory precisely because they feel like nothing. Round to the nearest five minutes; precision theater kills compliance.

4. 04

   Count every switch as its own event

   A note opened, abandoned for a session, and reopened is two switching entries, not zero. Tally the re-orientation minutes — finding the chart again, re-reading the half-draft, remembering the plan — under switching, and keep the writing minutes that follow under writing. The split between those two rows is the audit’s most common surprise.

5. 05

   Log after-hours time honestly

   The 8 p.m. hour on the couch counts, and so does Sunday. This row is the one most likely to be under-reported and the one that matters most, because after-hours documentation is where burnout and note quality problems concentrate. Whatever it totals, that number is the audit’s claim to being taken seriously.

6. 06

   Close each day in two minutes

   Before leaving — or before closing the laptop at home — total the rows for the day. This is not analysis; it is preservation. A card totaled nightly survives to Friday. A card left for Friday gets totaled from memory, which defeats the point.

7. 07

   Compute the split on Friday

   Sum each row for the week. Then compute three numbers: each category’s share of the total, the after-hours share, and the total divided by your visit count — your true per-visit documentation cost. The dominant category, not the total, is the finding.

The last line is not decoration. One observation per day — “spent ten minutes looking for the eval report,” “rewrote the same plan line three times” — is what turns Friday’s percentages into a specific fix instead of a general resolve to do better.

On Friday, one category usually holds a clear plurality of the minutes. That category — not the total, and not whatever a vendor’s demo happens to fix — picks the intervention.

Two of these fixes are covered in depth elsewhere on this blog: the end-of-day documentation workflow shows the three-moment structure that shrinks writing and re-reading time, and the caseload scheduling template shows where documentation blocks belong on a week so switching stops being the default. If searching dominates, the problem is usually the system of record itself — worth knowing before any software renewal.

“

“The total tells you documentation is expensive. Only the split tells you why.”

The audit earns its cost only if it ends in a single deliberate change. Change one thing — the plan-line format, the capture block, the retrieval setup — and run the card again a month later. Two changes at once produce a result that cannot be attributed; five changes produce a workflow nobody follows by Thursday. Against your own baseline, even a modest shift in the dominant category is real, visible progress — which is more than any benchmark can offer, because there is no defensible universal number for how long a therapy note should take.