Dysphagia Notes That Carry the Reasoning Behind the Diet Level

No other part of an SLP caseload gets its documentation read the way dysphagia does. A language note is reviewed by a payer looking at medical necessity. A dysphagia note can be read by a payer, an auditor, a facility’s quality team, and — in the worst case — an attorney reconstructing the days before an aspiration pneumonia admission. Every one of those readers is looking for the same thing: not whether you made the right call, but whether the note shows how you made it. The chart that holds up is the one where each diet-level decision, each strategy, and each referral can be traced backward to an observation that is actually written down.

Swallowing is the place where a therapy decision and a medical outcome sit closest together. When a patient on a modified diet is admitted with pneumonia, the chart is re-read in reverse: what was the diet order, who recommended it, what was it based on, and did anyone document the signs that should have prompted a change. The same chain gets read, more gently, by payers reviewing whether treatment was skilled — per Medicare’s framework, covered therapy must relate directly to a written plan of care, and the record has to show that the service required a therapist’s judgment rather than routine supervision. Both audiences are reading for reasoning, and reasoning is exactly what boilerplate crowds out.

The phrase that fails both readings is the one every SLP has written under time pressure: “Patient tolerated mechanical soft diet well.” It contains no observation, no interpretation, and no decision. “Tolerated” is a conclusion with the evidence deleted — it does not say what you watched for, what you saw, or why what you saw supports the diet the patient ate that day. The fix is not longer notes. It is a small amount of structure, applied at the moments that matter: trials, diet-level decisions, and referrals.

Strip any defensible dysphagia note to its skeleton and the same pattern repeats: an observable finding, an interpretation of that finding, and the action it led to. The finding is what a camera would have captured — wet vocal quality after thin liquid sips, a cough at the third consecutive bolus, oral residue in the lateral sulcus. The interpretation is your clinical read of it — suspected penetration, fatigue effect across the meal, reduced lingual sweep. The action is what changed because of it — a strategy introduced, a consistency trialed, an instrumental study recommended, a diet-level discussion with the physician. Most weak notes contain actions with no findings, or findings with no actions. The chain is what connects the two.

1. 01

   Record the finding as an observation, not a verdict

   Write what happened and when: the consistency, the amount, the bolus number, and the observable response. “Wet vocal quality following the second 5 ml thin-liquid sip” is a finding. “Difficulty with thin liquids” is a verdict wearing a finding’s clothes.

2. 02

   State the interpretation and own its uncertainty

   Say what you think the finding means and how confident the bedside view allows you to be: “suggests possible penetration; bedside exam cannot rule out silent aspiration.” Documented uncertainty is not weakness — it is the justification for the next step.

3. 03

   Tie the action to the finding by name

   The decision line should point back at its evidence: “Given wet vocal quality on thins that cleared with a chin-down posture and slowed rate, continued Level 2 (Mildly Thick) trials with strategy in place; will reassess across a full meal before any texture discussion with the team.”

4. 04

   Close the loop with the re-test condition

   State what would change your mind and when you will look again: the criteria for advancing, the criteria for stepping back, and the plan for instrumental confirmation where indicated. A note with a re-test condition reads as a clinician managing risk; a note without one reads as a snapshot.

Trials are where dysphagia documentation is won or lost, because trials are where the observations live. A trial line needs five parts: the consistency (named in IDDSI terms), what was presented and how much, the strategy or cue in play, the observable response, and what you concluded from it. Five parts sounds heavy; in practice it is one line per trial, and it is the difference between a note that contains evidence and a note that asks the reader to trust you.

Two habits make trial lines dramatically stronger. First, record the denominators — “coughed on 2 of 6 thin-liquid sips” is data, “occasional coughing” is an impression. Second, record the conditions that made a strategy work: volume, pacing, cueing level. A strategy that succeeds on three supervised teaspoon sips has not yet succeeded at an unsupervised meal, and a note that says so is a note showing clinical reasoning.

The International Dysphagia Diet Standardisation Initiative (IDDSI) framework exists because the older diet labels meant different things in different buildings — one facility’s “nectar thick” was not another’s, and “mechanical soft” had no standard definition at all. IDDSI replaces that with a single continuum of eight levels, numbered 0 through 7, where drinks span Levels 0–4 and foods span Levels 3–7, each with a name, a number, and published testing methods. When your note says “Level 2 (Mildly Thick)” instead of “nectar,” every downstream reader — the kitchen, nursing, the receiving facility after discharge, the auditor — reads precisely what you assessed.

For documentation, the practical rule is simple: write the level number and the name together on first use, and keep legacy terms out of the assessment and plan except where you are quoting an existing physician order. If the facility still runs on older diet names, document the mapping once (“facility ‘mechanical soft’ order; consistent with IDDSI Level 6 (Soft & Bite-Sized) per facility diet manual”) so the chart stays internally coherent while the building catches up.

A diet-level recommendation is a position between two neighbors, and a complete rationale addresses both of them. Why not the less restrictive level: what did you observe, at what volumes and under what conditions, that makes the step up premature — and what would need to be true to take it. Why not the more restrictive level: what evidence shows the current level is being managed safely, so the restriction is not simply reflexive caution. The second half is the one that goes missing, and it matters more than most notes acknowledge: texture modification carries its own costs — palatability, intake, hydration, and quality of life — so “safer” is not automatically better, and a note that never justifies the restriction reads as defensive rather than reasoned.

Written out, the rationale is usually two or three sentences: “Recommend continuing Level 4 (Pureed) with Level 0 (Thin) liquids by teaspoon under supervision: pureed managed × 2 full meals without overt s/sx and with adequate intake; unthickened liquids at open-cup volumes produced wet vocal quality on 2 of 4 presentations even with chin-down posture. Will advance liquid trials toward Level 0 at increasing volumes as strategy tolerance builds; VFSS requested to characterize physiology before any diet upgrade to Level 5.” Every clause in that is either an observation already in the note or a condition for change. That is the whole trick.

Here is the whole argument compressed into a single session. The case is fictional but ordinary: the kind of visit that gets documented forty times a month, and the kind of note that decides — months later, in front of a reader you never met — whether the chart shows a clinician reasoning or a clinician typing.

ASHA’s position is blunt: instrumental exams — videofluoroscopy (VFSS) and endoscopy (FEES) — are the only assessments that visualize swallowing physiology, and a clinical bedside evaluation cannot detect silent aspiration. That limitation belongs in your documentation, in two specific places. In the assessment, as an explicit boundary on your conclusions: “no overt signs of aspiration observed; bedside exam cannot rule out silent events.” And in the plan, as the reasoning for a referral when one is indicated: name the clinical question the study would answer — characterize physiology, test a strategy under visualization, evaluate readiness for a diet change — rather than writing “VFSS recommended” with no question attached.

The same linkage runs the other direction. When an instrumental study exists, treatment notes should visibly build on it — the strong note above ties each bedside observation back to the VFSS findings it is consistent with. A chart where the bedside notes and the instrumental report never reference each other reads as two clinicians who have not met; a chart where they interlock reads as a plan.

For Medicare Part B, the clinical reasoning above sits inside a documentation frame set by the Medicare Benefit Policy Manual (Pub. 100-02, Chapter 15, §220.3): services must relate directly and specifically to a written plan of care established before treatment begins, treatment notes must document each encounter, and a progress report is due at least once every 10 treatment days. The common dysphagia codes are untimed — CPT 92610 for the clinical swallowing evaluation and CPT 92526 for swallowing treatment each bill one unit for the encounter regardless of duration — so for these codes the note’s job is not minute math but demonstrating that what happened was skilled.

Two coding-adjacent documentation habits are worth building in from ASHA’s guidance. First, the 92610 evaluation already includes trials across a continuum of consistencies, compensatory strategy work during the evaluation, and patient and family education — so when evaluation and treatment are billed the same day, the notes must show them as separate and distinct services, with treatment addressing the established plan of care. Second, the progress-report cadence is a floor, not a formality: for a condition as dynamic as dysphagia, the progress report is where trial data accumulates into a documented trajectory. Commercial payers and Medicaid programs set their own rules — verify each payer’s requirements rather than treating Medicare’s as universal.