SOAP notes must work for the clinician who wasn't there

A colleague opens your note before the next visit. They need to know what changed and what you decided. SOAP notes give that account four parts: Subjective, Objective, Assessment and Plan. The StatPearls reference on SOAP notes describes the basic structure:

The format gives learners somewhere to start. In December 2020, Saumil Chudgar described teaching SOAP notes to first-year Duke medical students. He was excited by their curiosity about something clinicians do so often it can feel automatic.

Keep that curiosity. A template gives you somewhere to put the findings, but the reader still needs to understand their significance. Explain which information shaped the assessment and why the plan follows. Four completed headings can leave those connections missing.

In October 2021, Amy Francis asked what O actually meant. She used O for what she had done and measured, but had seen others use it for the session's goal or objective. She asked whether she should do the same.

In SOAP, Objective refers to findings and results; it does not mean the goal of the encounter. The distinction between Subjective and Objective concerns where information comes from. A number alone does not settle which section it belongs in.

Consider two hypothetical entries: a patient says their home device recorded a particular reading, or the clinician reviews a dated result in the device record. Those are different sources of information. Label the first as patient-reported. For the second, identify the record reviewed and its date. Do not turn a recollection into a verified result simply by moving it under O.

Use the same care when a report occurs during an observed activity. You may observe someone stop walking; discomfort is what they report. Both details can appear in the account of that activity, with their sources clear. Make that distinction in the sentence itself.

Fictional example: this abbreviated physiotherapy follow-up is dated 14 September 2026. It illustrates documentation structure and is not a complete clinical record or treatment guide. The patient already has a clinician-agreed rehabilitation plan.

Subjective: The patient reports walking for 10 minutes at home before knee discomfort, compared with a reported five minutes at the previous visit.

Objective: During today's observed clinic walk, the patient stopped after eight minutes and reported knee discomfort.

Assessment: The patient reports improved walking tolerance at home. Observed walking tolerance in the clinic remains limited. The home and clinic durations describe different conditions and are not interchangeable measurements.

Plan: Review progress against the existing rehabilitation goals at the agreed follow-up on 21 September 2026.

The distinction matters. The assessment records reported improvement while keeping the clinic result separate. Writing only “improving” would lose that qualification. Writing “walked 10 minutes” under Objective would turn the patient's account into something the clinician appeared to have observed.

The plan gives the review a purpose and a date. No new exercise or intervention has been invented to fill the box. If the clinician changes the plan, the real note should record that decision and its rationale. Those details must come from the encounter; a template cannot supply them.

Joseph Miller described changing his preferred note order after moving from paper charts to Cerner at UAB and from paper to Epic at Erlanger. He wanted the assessment and plan at the top, where a colleague could find the interpretation and intended next steps immediately.

APSO puts Assessment and Plan before Subjective and Objective. A colleague reaches the decision first. It changes the presentation without removing the need to record the information behind the decision.

A 2017 study by Belden and colleagues tested four displays containing the same note information with 16 primary-care physicians. APSO prototypes performed better overall on information retrieval and usability. But the displays also varied in columns and collapsible sections. Those results can't tell us what changing the order alone would achieve, and they don't establish better patient outcomes.

Follow your workplace's documentation format. If the system lets you change the display, consider how easily a colleague can find the current assessment and plan. In a required SOAP template, clear headings still help. Keep the relevant history and examination findings in either layout. The reader needs both the decision and enough supporting information to understand it.

Robert Oubre, a physician who teaches documentation, warns that relative dates become inaccurate when carried into later notes. Write “review tomorrow” and the meaning depends on when you wrote it. Copy those words into a later note and the appointment appears to move.

For example, if the agreed review is on 21 September 2026, record that date. In a later note, verify whether the review is still planned, has happened or has been rescheduled. A precise date helps the reader understand the timeline; it does not prove the information is still correct.

The same check applies to the assessment and plan. Read carried text against the current encounter. Confirm which findings remain relevant, what has changed and whether the previous decision still applies. A new timestamp won't do that work.

In the UK, GMC guidance, paragraphs 69–70, calls for accurate, timely records with proportionate detail, including decisions, agreed actions and when review is needed. The record should also say who created it and when.

Keep today's account distinct from the historical record. Do not silently overwrite a signed earlier note to make it current. Use your organisation's amendment process when an earlier entry needs correction.

AI scribes can help with documentation, but results differ by product. In a 2025 UCLA randomized trial involving 238 outpatient physicians, Nabla reduced time spent writing in the EHR note by 9.5% versus control. DAX showed no statistically significant change. That measure excluded editing time inside the vendors' platforms, so it does not capture all the work involved in producing a finished note.

Accuracy needs a separate check. In Biro and colleagues' 2025 simulation study, two commercial scribes generated 44 drafts from repeated readings of 11 encounter scripts. Researchers found errors in 31 drafts, with omissions the most common type. This small simulation does not estimate the error rate in everyday practice. A missing detail can leave a perfectly fluent paragraph. There's nothing awkward to catch your attention.

Compare the draft with the encounter and the records you actually reviewed. Check who supplied each fact, when it applied and whether a negative became a positive. Look for a proposed investigation presented as completed, or an assessment that no longer supports the plan. Finally, look for relevant findings and decisions that never made it into the draft.

Use your organisation's approved system to review, correct and sign the note. Keep identifiable patient information out of unapproved tools. If the draft leaves you uncertain whether a decision was discussed or agreed, resolve that uncertainty before recording it as the plan.