How to organize clinical reasoning in the chart

It is two in the morning, and you need to understand why the patient in bed 12 received antibiotics. The progress note has three screens of copied exams, yesterday's physical exam and, at the end, one line: "maintain management". The reasoning that led to the decision is not written anywhere. A review on copy and paste in the electronic health record describes exactly this problem: the practice can bloat the note, create internal contradictions and spread errors (Tsou et al., 2017).

NEXO is a daily medical diagnosis game: one clinical case a day, six progressive clues, and ICD-10 proximity feedback across five levels.

This text shows how to organize clinical reasoning in the medical record so that another person understands what you thought, why you decided and what you expect to happen. At the end, there is an assessment model and a checklist.

Key points

• The medical record serves to communicate with the team and provide continuity of care, and written reasoning is the part that allows this continuity.
• The problem-oriented record and the SOAP method organize the note into narrative, findings, assessment and plan.
• The assessment is the space for reasoning: interpretive summary, prioritized differential and justification of the main hypothesis and the alternatives.
• Long notes are not better notes; copy and paste and the excess of ready-made templates increase the note without adding reasoning.
• Note assessment tools show that it is possible to give objective feedback on the quality of written reasoning.

Summary

• What written reasoning is for
• Problem-oriented record and SOAP
• How to write the assessment
• Long notes, copy and paste and ready-made templates
• An annotated model
• How to train writing reasoning
• Frequently asked questions
• Checklist to organize reasoning in the medical record

What written reasoning is for

Many people learn to fill in the medical record as a list of required fields. The fields matter, but the main purpose of the record is another: to allow the next person to continue care from where you stopped.

What the Brazilian regulation says

CFM Resolution 1.638/2002 defines the medical record as the single document of information about the patient's health and the care provided, of a legal, confidential and scientific nature, which enables communication among team members and continuity of care (CFM, Resolution 1.638/2002). The same resolution considers the medical record valuable also for teaching and research (CFM, Resolution 1.638/2002).

Communication and continuity depend on more than data. A colleague can read the blood count alone; what they need to know from you is what that blood count means for this patient and what you decided because of it.

Written reasoning as a learning tool

For the student, writing the reasoning is also a way of thinking. By putting on paper why one hypothesis is the most likely, you notice when the justification is weak or when a finding was left unexplained. The very origin of the problem-oriented record has this idea in its title: Lawrence Weed's 1968 article is called "Medical records that guide and teach" (Weed, 1968).

Where notes usually fail

In a study that assessed 252 admission notes written by internal medicine residents and fellows, only 53% were considered high quality in the documentation of clinical reasoning (Schaye et al., 2022). The authors also report that residents receive little feedback on this part of the note (Schaye et al., 2022). In other words, the skill is important, little taught and can be improved.

Problem-oriented record and SOAP

The most used format to organize reasoning in the medical record comes from the problem-oriented record. It is the model adopted by the electronic health record of primary care in the SUS.

Weed's idea

Lawrence Weed is known as the creator of the problem-oriented medical record, described as the standard for collecting, organizing and contextualizing patient data (Shukor, 2017). The manual of the e-SUS APS electronic health record notes that this model, conceived by Weed in the 1960s, consolidated the concepts of problem list and problem-oriented medical record and brought the SOAP method to progress notes (Ministry of Health, e-SUS APS manual).

The four letters

The e-SUS APS manual describes SOAP as four items in sequence (Ministry of Health, e-SUS APS manual):

Letter	What goes in it, according to the manual	Question that helps to write
S, subjective	The patient's account, if possible as it was reported	What did the patient bring, in their own words?
O, objective	The findings from observation, physical examination and tests	What did I see, measure and find?
A, assessment	The professional's judgment in light of the complaints, the findings and clinical reasoning; the space for speculation, inferences and conclusions	What do I think is happening, and why?
P, therapeutic plan	The course of action, whether ordering tests, recommending medication or giving guidance	What will I do, and what do I expect to happen?

Source of the table: descriptions from the e-SUS APS manual; the question column is from the Editorial.

Note that the manual defines the assessment as the space for clinical reasoning. It is there that the note stops being a list of data and starts explaining the patient.

The problem list

According to the same manual, each item of SOAP develops a problem recorded in the problem list, which helps to locate the history of that problem across the various times it is managed (Ministry of Health, e-SUS APS manual). In practice, this means writing the assessment and the plan per problem, and not in a single block.

A problem can be an established diagnosis, a syndrome, a symptom still without explanation or an abnormal finding. When the diagnosis does not yet exist, record the problem at the level at which you understand it. The guide on how to find the right ICD-10 code shows how to record symptoms and findings and keep the suspicion recorded as a suspicion.

How to write the assessment

The assessment is the hardest and most useful part of the note. A tool created to assess students' admission notes helps to organize what it should contain.

The four elements of the IDEA tool

The IDEA tool assesses admission notes across four elements: interpretive summary, differential diagnosis, explanation of reasoning and alternatives (Baker et al., 2015). The revised version kept the same four domains, with more detailed descriptions of each quality level (Schaye et al., 2022).

Element	What to write	Example sentence
Interpretive summary	One or two sentences that bring together the essential data in clinical language	"68-year-old man, smoker, with progressive dyspnea for three weeks and lower limb edema"
Differential diagnosis	The hypotheses in order of probability, including the serious one that cannot be missed	"Main: decompensated heart failure. Alternatives: acute exacerbation of chronic lung disease, thromboembolism"
Explanation of reasoning	The data that support the main hypothesis	"Favoring: orthopnea, edema, bibasal crackles"
Alternatives	Why the other hypotheses came behind, and what would make you change your mind	"Thromboembolism less likely due to the slow progression; reassess if there is sudden chest pain"

Source of the table: elements of the IDEA tool according to Baker et al., 2015; the examples are from the Editorial, illustrative, and do not replace the assessment of each patient.

The interpretive summary

The interpretive summary condenses the history into a short portrait of the problem. Instead of repeating everything that is in the S and the O, it selects what matters and translates it into clinical terms: "crushing chest pain on exertion" instead of "feels a pain in the chest when climbing the stairs". A good test is to read only the summary and ask whether it already suggests the main hypotheses.

The prioritized differential

The differential in the note is an ordered list, and not an inventory of all possible diseases. The main hypothesis comes first, followed by the most likely alternatives and the serious hypothesis that must be ruled out. The text on differential diagnosis with clues explains how to build this list; in the note, what changes is that each item needs a short justification.

Justify the main one and the alternatives

The explanation of reasoning is what separates a note that communicates from a note that only records. Write which data favor the main hypothesis and which data weaken it. For the alternatives, write why they came behind and what would make you reconsider them.

That last point is the most forgotten and the most useful for whoever will read the note at dawn. A sentence like "if the fever persists after 48 hours of treatment, reconsider an abdominal focus" tells the colleague exactly what to watch for. The text on clinical decision-making goes deeper into how to link each test ordered to a question.

The management linked to the assessment

In the P of Therapeutic plan, each action must answer something that is in the assessment. If you ordered a test, the assessment must say which hypothesis it confirms or rules out. If the management is to observe, write for how long and what changes the path. When the decision was made with doubt, the text on priority and hypothesis in the medical internship shows how to present that doubt clearly.

Admission, daily progress and discharge

The structure of the assessment is the same at all times, but the weight of each part changes.

Moment	What weighs most in the assessment	Question to close the note
Admission	Complete interpretive summary and broad differential, with the severe hypothesis made explicit	What needs to be confirmed or ruled out in the next hours?
Daily progress	What changed since yesterday and how does that shift the order of the hypotheses	Does the main hypothesis still stand, and why?
Discharge	Final diagnosis or open problem, pending items and guidance	Who follows up on what is pending, and when?

Source of the table: didactic organization by the Editorial based on the elements of the IDEA tool (Baker et al., 2015) and on the logic of problem-oriented notes in the e-SUS APS manual.

In the daily progress, the most frequent mistake is rewriting the admission. The useful question is another one: what happened in the last 24 hours that confirms, weakens or changes the main hypothesis? If nothing changed, one sentence is enough, as long as it says why the management continues. At discharge, the note needs to be read by someone who does not know the patient, such as the physician at the health unit who will follow up with them; therefore, pending items and open problems are written explicitly, and not hidden in the middle of the progress.

Language that makes the doubt clear

The way you write probability also communicates. "Probable", "possible" and "ruled out" say different things, and mixing the terms confuses the reader. A simple rule is to always use the same words for the same degrees of certainty and, when the doubt is great, to say which piece of data would resolve it. Thus, the note shows not only what you think, but how much you trust what you think.

Long notes, copy and paste and ready-made templates

With the electronic health record, it became easy to write a lot. It became equally easy to write a lot without thinking.

Copy and paste

The systematic review on copy and paste found that 66% to 90% of clinicians use the resource frequently (Tsou et al., 2017). In a study cited by the review, copy and paste contributed to 2.6% of diagnostic errors in which the patient needed additional unplanned care (Tsou et al., 2017).

The authors acknowledge that the evidence of direct risk to the patient is still scarce and propose four safe practices: make copied text easy to identify, show where it came from, train the team and monitor use (Tsou et al., 2017). For whoever writes, the practical rule is simple: copy data, if needed, but write the day's assessment in your own words.

What shrinks a note

An analysis of health record metadata from more than 130 thousand outpatient physicians identified those who reduced the size of notes or the time spent on them (Apathy et al., 2023). The reduction in size came mainly from less copied and pasted text, with an average relative drop of 18.9%, and from less text from ready-made templates, with a drop of 17.2% (Apathy et al., 2023).

A template that improved the notes

In a study at four university hospitals in the United States, internal medicine residents received a class on documentation and a daily progress template with good practices (Kahn et al., 2018). After the intervention, the notes improved in quality, including recording only relevant data and being concise without losing the essential, and ended up with about 25% fewer lines (Kahn et al., 2018). The study was not randomized, which calls for caution in interpretation.

The lesson is that a ready-made template is not the problem. The problem is the template that pulls data into the note without asking for reasoning. A good template leaves explicit room for the summary, the differential and the justification.

Common mistakes

• Assessment that repeats the data. "Patient with fever, leukocytosis and raised CRP" is a summary of tests, not an assessment. Say what these data mean together.
• Differential without order. A list of six diseases without hierarchy does not tell the colleague what you think is most likely.
• "Maintain management" without a reason. Write why the management continues and what would make you change it.
• Progress note copied without review. A physical exam from yesterday pasted into today's note can contradict what the patient has now.
• Suspicion recorded as diagnosis. While the hypothesis is not confirmed, make it clear that it is a hypothesis.

An annotated template

The template below is a suggestion from the Editorial team for the assessment and management part of a progress note. The clinical data are illustrative and serve only to show the structure; adapt them to your service's rules.

The template

Problem 1: dyspnea with lower limb edema.

Summary: 68-year-old man, smoker, with progressive dyspnea for three weeks, orthopnea and lower limb edema, with bibasilar crackles on examination.

Main hypothesis: decompensated heart failure. Favoring: orthopnea, edema, crackles and progression over weeks. Weakening: no known history of heart disease.

Alternatives: acute exacerbation of chronic lung disease, less likely due to orthopnea and edema; pulmonary thromboembolism, less likely due to the slow progression, but which needs to be reconsidered if there is sudden chest pain or rapid worsening.

Management: tests to confirm the main hypothesis and assess cardiac function, treatment according to the service protocol and daily clinical reassessment. Change the hypothesis if there is no improvement in dyspnea with treatment.

Problem 2: smoking. Address at discharge.

Why it works

Each problem has its own assessment. The summary selects the data that matter. The main hypothesis comes with what favors it and what weakens it. The alternatives have a reason to be behind and a trigger to come forward. The management responds to the assessment and says what changes the path. Whoever reads the note at two in the morning knows what you thought and what to watch for.

What it does not do

The template does not define specific tests or treatments, which depend on the patient and the service protocols. It organizes the reasoning so that these decisions are visible and can be discussed.

How to train writing the reasoning

Writing the assessment well requires two skills: reasoning well and writing the reasoning in a few clear sentences. Both improve with practice and feedback.

Ask for feedback with criteria

The revised version of the IDEA tool was created precisely to give feedback on the documentation of reasoning with criteria shared between the evaluator and the person being evaluated (Schaye et al., 2022). Even without the tool, you can ask the preceptor to comment on four points: the summary, the order of the differential, the justification of the main one and that of the alternatives.

There is also research on automated feedback. A study at two American institutions developed artificial intelligence models to assess the documentation of reasoning in residents' admission notes, and the authors conclude that this type of tool can expand feedback on clinical reasoning (Schaye et al., 2025). Performance varied between models, and some had performance insufficient for implementation (Schaye et al., 2025).

Train the summary and the differential with short cases

The interpretive summary and the prioritized differential are the parts that benefit most from volume. A simple exercise:

1. Read a short case and write the summary in one sentence.
2. List three hypotheses in order, with one piece of data in favor of each.
3. Write what would make you switch the first for the second.
4. Compare with the answer and see if your summary already pointed to it.

Train on NEXO

In NEXO, the case arrives in up to six clues, one at a time, and you give your guess by the diagnosis name or by the ICD-10 code. Each wrong guess shows the proximity in five levels, from different chapter to exact ICD, and opens the next clue. This trains the core of written assessment: summarizing what you know, prioritizing a hypothesis and adjusting it with the next piece of data. A good habit is to write your summary in one sentence before each guess.

The Diary brings one case per day, and the Archive gathers the previous daily cases. Multimedia brings images, audio and other exams, similar to the data that go into the objective item of the score. Adrenaline adds time, and Continents changes the background epidemiology. How each mode is unlocked is on the NEXO economy page.

Saving and discussing

When a case shows that your summary left out the decisive piece of data, the Anki button creates the card straight from NEXO and opens it in the Anki app. The error notebook method helps separate knowledge errors from reasoning errors. If you coordinate a medical school or a residency program and want to use cases like these in teaching clinical writing, check out NEXO Closer. Researchers interested in documentation and clinical reasoning can propose studies through NEXO Researcher, and physicians who want to review cases in their specialty can take part through NEXO Curator, and those who produce medical content can check out NEXO Creators.

Frequently asked questions

What is the SOAP method?

It is a standardized way of organizing the progress note into four items: subjective, objective, assessment and the P, for Plan, which brings the management. The e-SUS APS manual describes the assessment as the professional's judgment in light of the complaints, the findings and the clinical reasoning (Ministry of Health, e-SUS APS manual).

Where does clinical reasoning enter the medical record?

Mainly in the assessment. There you write the interpretive summary, the differential in order, the justification for the main hypothesis and the reason the alternatives fall behind, the four elements of the IDEA tool (Baker et al., 2015).

What is problem-oriented recording?

It is the model in which the note is organized by problem, each one linked to a problem list. It was devised by Lawrence Weed in the 1960s and is the model adopted by the e-SUS APS electronic medical record (Ministry of Health, e-SUS APS manual).

Is a longer note a better note?

Quality depends on the reasoning, not on the length. In a study with a daily progress note template, the notes improved in quality and ended up with about 25% fewer lines (Kahn et al., 2018).

Can I copy and paste yesterday's progress note?

With care. The review on the topic linked the practice to inflated notes, contradictions and the spread of errors, and recommends that copied text be easy to identify and have a visible source (Tsou et al., 2017). Write the day's assessment in your own words.

How do you record a case with no defined diagnosis?

Record the problem at the level you understand it, as a symptom, syndrome or finding, and write the hypothesis as a hypothesis. The guide on ICD-10 shows how to record symptoms.

How does NEXO help train writing your reasoning?

In NEXO, each case arrives in up to six clues and you adjust your guess with the proximity in ICD-10, which trains summarizing data and prioritizing hypotheses. To get to know the game, read what NEXO is.

Checklist for organizing your reasoning in the medical record

1. Write by problem, linked to the problem list.
2. Separate report, findings, assessment and management.
3. Open the assessment with an interpretive summary of one or two sentences.
4. Order the differential and include the serious hypothesis that cannot be missed.
5. Justify the main hypothesis with data for and against.
6. Say why the alternatives fell behind and what would make you change.
7. Link each management step to a question from the assessment.
8. Write what changes the path when the management is to observe.
9. Review what was copied and write the day's assessment in your own words.
10. Ask for feedback with criteria on the summary, differential and justification.

Want to train the summary and the differential today? Play the case of the day on NEXO.

References

1. Weed LL. Medical records that guide and teach. N Engl J Med. 1968. https://pubmed.ncbi.nlm.nih.gov/5637250/
2. Shukor AR. An Alternative Paradigm for Evidence-Based Medicine: Revisiting Lawrence Weed, MD's Systems Approach. Perm J. 2017. https://pubmed.ncbi.nlm.nih.gov/28488985/
3. Ministry of Health. e-SUS APS Electronic Health Record: conceptual basis of the system. https://sisaps.saude.gov.br/sistemas/esusaps/docs/manual/PEC/PEC_00_base_conceitual/
4. Federal Council of Medicine. CFM Resolution No. 1,638/2002. https://sistemas.cfm.org.br/normas/arquivos/resolucoes/BR/2002/1638_2002.pdf
5. Baker EA, Ledford CH, Fogg L, Way DP, Park YS. The IDEA Assessment Tool: Assessing the Reporting, Diagnostic Reasoning, and Decision-Making Skills Demonstrated in Medical Students' Hospital Admission Notes. Teach Learn Med. 2015. https://pubmed.ncbi.nlm.nih.gov/25893938/
6. Schaye V, Miller L, Kudlowitz D, et al. Development of a Clinical Reasoning Documentation Assessment Tool for Resident and Fellow Admission Notes: a Shared Mental Model for Feedback. J Gen Intern Med. 2022. https://pubmed.ncbi.nlm.nih.gov/33945113/
7. Schaye V, DiTullio D, Guzman BV, et al. Large Language Model-Based Assessment of Clinical Reasoning Documentation in the Electronic Health Record Across Two Institutions: Development and Validation Study. J Med Internet Res. 2025. https://pubmed.ncbi.nlm.nih.gov/40117575/
8. Tsou AY, Lehmann CU, Michel J, et al. Safe Practices for Copy and Paste in the EHR. Systematic Review, Recommendations, and Novel Model for Health IT Collaboration. Appl Clin Inform. 2017. https://pubmed.ncbi.nlm.nih.gov/28074211/
9. Apathy NC, Hare AJ, Fendrich S, Cross DA. I had not time to make it shorter: an exploratory analysis of how physicians reduce note length and time in notes. J Am Med Inform Assoc. 2023. https://pubmed.ncbi.nlm.nih.gov/36323282/
10. Kahn D, Stewart E, Duncan M, et al. A Prescription for Note Bloat: An Effective Progress Note Template. J Hosp Med. 2018. https://pubmed.ncbi.nlm.nih.gov/29350222/