Homeopathic Case Analysis: Documenting the Differential
Educational and safety note: This guide describes professional documentation of homeopathic case analysis. It is not medical, legal or regulatory advice and does not establish clinical efficacy. It does not replace diagnosis, emergency assessment, referral, informed consent, applicable law or practitioner judgment. Workflow software does not diagnose, prescribe automatically or choose a remedy for the practitioner.
Quick answer: what should a homeopathic case analysis contain?
A useful homeopathic case analysis is a dated working record of how a practitioner moved from a clarified case picture to a reasoned comparison—and, if appropriate, to a plan. It should show four things that a repertory chart cannot show by itself: which features were treated as characteristic, which candidates were compared, what named materia medica supported or contradicted each candidate, and what remained uncertain.
Write the analysis so another trained reader, a supervisor or your future self can reconstruct the reasoning without guessing. Preserve the source narrative. Link every important claim back to a session note, rubric path or materia medica edition. Treat the conclusion as provisional until the practitioner accepts it.
Key takeaways
- Analysis begins after the interview has been clarified and the rubric set has been justified, not while the person is still telling the story.
- The characteristic picture used for comparison is a subset of the record, not a rewrite of the person.
- Name the repertory, edition or database behind the candidate set.
- Read leading and close alternative remedies in materia medica before documenting a working decision.
- Record contradictions as carefully as confirmations.
- Keep facts, interpretations, method choices and actions on separate lines.
Why the chart is not the analysis
Repertorization answers a retrieval question: which indexed remedy entries cover the selected rubrics? Case analysis answers a different question: does that shortlist still make sense when read against the whole person, the source notes and fuller remedy texts?
The rubric-selection and source-verification guide covers the translation from patient language to a reviewable rubric set. The case-taking and follow-up documentation guide covers the interview record that must exist before that translation. This guide covers the written artifact that should exist afterwards.
A high score can be produced by overlapping rubrics, a keyword mismatch, an edition difference or an early preferred-remedy bias. A low score can hide a remedy that fits the case picture better once the materia medica is read. The analysis note is where those possibilities become visible.
Hahnemann’s Organon treats the written case as a picture to be compared, not as a pile of disconnected complaints. Later commentary on §153 emphasizes that striking, singular and peculiar features carry more comparative weight than vague generalities shared by many states and many medicines. That is a method for ranking information. It is not a claim that any particular analysis proves a clinical outcome.
What belongs in the analysis note
Keep four layers connected but distinguishable.
| Layer | Store here | Do not let it replace |
|---|---|---|
| Source session | Verbatim phrases, chronology, observations, concurrent care | A later summary that “improves” the wording |
| Rubric set | Paths, repertory/edition, weights, exclusions | The characteristic picture itself |
| Analysis note | Picture used, candidates, materia medica checks, uncertainties, working decision | An automatic prescription or a profile fact |
| Plan and communication | What was agreed, by whom, through which channel, and the next review point | The reasoning that led there |
The practice-management workflow treats “analysis pending” as an operational state with an owner and due date. The analysis note is the required output of that state. Until it exists, the case is not ready for a communicated plan.
The REASON model
Use REASON as a practical documentation sequence. It is a working model, not a school, legal standard or proof of effect.
- R — Restate the characteristic picture
- E — Enumerate the candidate set and its provenance
- A — Assess confirmations in named materia medica
- S — Surface contradictions and missing information
- O — Own a working decision, including a decision to wait or refer
- N — Note review criteria and version the record
1. Restate the characteristic picture
Before opening materia medica, write a short picture of the features actually used for comparison. This is not a second case-taking. It is a declaration of what the analysis is testing.
Include:
- the dated session or sessions used;
- the main complaint in the person’s language;
- the few features treated as characteristic, with a one-line reason for each;
- features deliberately left out of the comparison, and why;
- safety, referral or concurrent-care facts that sit outside the repertory work.
A compact example of style, not of a complete case:
| Feature used | Source | Why it is treated as characteristic | Left out of comparison |
|---|---|---|---|
| “I wake at 3 a.m. with my thoughts racing.” | Session 12 Mar, client words | Precise time and mental state; not a generic “insomnia” label | Common tiredness after a late meeting |
| Speaks quickly; pauses when describing work | Practitioner observation | Repeated, visible and independent of the complaint wording | A single restless gesture |
If this table cannot be filled, the analysis is premature. Return to clarification rather than forcing a differential from incomplete notes. Unknown is more accurate than a prompted or software-completed detail.
2. Enumerate the candidate set and its provenance
List the remedies under comparison and why each one is present. A candidate may arrive from rubric coverage, a materia medica recollection, a supervisor suggestion or a software retrieval. The source of the candidacy belongs in the note.
Record:
- the repertory or database and its edition or version;
- the rubric set identifier or date;
- whether an unweighted, conservative or alternative set was also viewed;
- any candidate added or removed after seeing the chart, with a reason;
- candidates kept for comparison even though they were not the highest score.
Do not let the display order of a chart become the only explanation. “First on the graph” is not a rationale.
3. Assess confirmations in named materia medica
A repertory entry is an index line. Materia medica is the longer text that the index points to. Read the leading candidates and at least one close alternative in a named source before writing a conclusion.
For each candidate, note:
- which materia medica, author and edition or digital version was opened;
- which passage or heading was relevant;
- whether the match is to the person’s words, to a clarified modality or only to a repertory synonym;
- whether the same confirmation appears in more than one independent source, or only in a condensed keynote list.
Do not invent a quotation to make the note look complete. If you did not read a passage, write that it was not checked. If two authors disagree, record the disagreement instead of averaging them into a smoother story.
Comparative reading is the point. The question is not “does this remedy have many symptoms?” It is “does this text fit the characteristic picture better, worse or differently than the next candidate, and where is that stated?”
4. Surface contradictions and missing information
Actively look for what does not fit. A contradiction may be a general (general symptom / generalia), a modality, a mental state, a thermal pattern or a timing that the case shows and the text does not, or the reverse.
Useful contradiction classes:
- direct opposite: the text emphasizes a modality the case clearly does not show;
- missing expected general: the text is built around a general (general symptom / generalia) that was never elicited or is unknown;
- source mismatch: the apparent fit exists only after the person’s language was stretched into a different meaning;
- care confound: another medicine, life event or concurrent treatment arrived at the same time as the feature being used;
- safety override: a finding that requires referral or conventional assessment regardless of repertory coverage.
Missing information is not a contradiction. Mark it as unknown and decide whether the analysis can proceed, should wait or should return to a specific question at the next contact.
5. Own a working decision
The analysis may conclude in several legitimate ways:
- a reasoned preference among compared candidates;
- two remaining alternatives that need one missing fact;
- no preference because the picture is incomplete or unsafe to interpret;
- a decision to wait, observe or gather documents;
- a referral or a recommendation to seek appropriate conventional care;
- a plan communicated after practitioner review.
Write the decision as a working position, not as proof. Include the rejected near-alternatives and why they were not preferred on the information then available. If no prescription is made, that is still an analysis outcome and should be dated.
Do not document potency, repetition or product choice in a way that turns this guide into dosage instruction. If a plan includes a medicinal product, record the facts required by local professional rules—name, date, communication and responsibility—without treating the note as a tutorial in prescribing.
6. Note review criteria and version the record
End with what would change the analysis:
- the next review date or trigger;
- the domains to compare against today’s baseline;
- the specific unknown that would reopen the differential;
- who owns any outstanding task.
If later information changes the picture, save a new dated version. Do not overwrite the original analysis as though the new fact had been known at the first sitting. Follow-up comparison belongs in the next session note and, where needed, in a new analysis version. See the follow-up table in the case-taking guide.
A reusable analysis-note table
| Field | What to record | Quality check |
|---|---|---|
| Date and author | Who wrote the analysis and when | Not an undated software export |
| Sessions used | Source notes included and excluded | Later sessions are not silently mixed in |
| Characteristic picture | Features used and features withheld | Each used feature has a source |
| Candidate set | Names plus why each is present | Provenance is not only “high score” |
| Repertory reference | Work, edition/version, rubric-set date | Another person can reopen the same set |
| Materia medica checks | Source, heading or passage, result | Unread sources are marked unread |
| Confirmations | What agreed, in whose language | Match is not only a synonym leap |
| Contradictions | What disagreed or could not be checked | Negatives are visible |
| Uncertainties | Unknowns and alternative readings | Uncertainty is not deleted for polish |
| Working decision | Preference, wait, refer or further question | Action is separated from interpretation |
| Review criteria | What would reopen the analysis | Next step has an owner and time |
Confirmation-versus-contradiction matrix
Use one row per candidate actually read. Empty rows are a warning that the chart was accepted without materia medica.
| Candidate | Why listed | Materia medica read | Confirmation | Contradiction or unknown | Status after review |
|---|---|---|---|---|---|
| Remedy A | Broad coverage of the selected set | Named source and heading | Matches the verbatim 3 a.m. mental state | Thermal general never asked | Open, needs one question |
| Remedy B | Close alternative on a conservative set | Named source and heading | Matches hurry observed in session | Key modality opposite to the case | Not preferred on present data |
| Remedy C | Recalled from earlier study, not from the chart | Not yet read | — | Check pending | Not decided |
Status language should stay modest: open, not preferred on present data, not checked, preferred working hypothesis. Avoid “proved”, “confirmed cure” or “the simillimum is”.
Common analysis-documentation failures
| Failure | Why it weakens the record | Better practice |
|---|---|---|
| Pasting the chart as the analysis | Retrieval is mistaken for reasoning | Write the picture, checks and decision separately |
| Reading only the first-ranked remedy | Near alternatives never get a fair test | Read at least one close alternative |
| Unnamed “materia medica” | No one can reopen the same text | Record author, work and edition or version |
| Synonym inflation | A stretched label looks like a match | Keep the person’s words beside the passage |
| Silent preferred-remedy bias | Rubrics and reading are steered after the fact | Freeze the picture and set before reading |
| Deleting contradictions | The note becomes a sales argument | Keep mismatches in the same table |
| Overwriting after follow-up | Hindsight rewrites what was known | Version the later analysis |
| Saving an AI comparison as fact | Fluent text hides unread sources | Keep drafts provisional until review |
| Turning analysis into a prescription tutorial | The record starts to give treatment advice | Record the decision; keep dosage teaching out |
| Ignoring safety findings | A repertory fit delays appropriate care | Document referral or urgent routing first |
CARE, HOM-CASE and private records
The CARE guideline is a reporting standard for published case reports. It asks for a timeline, diagnostic reasoning, interventions, follow-up and consent. HOM-CASE adds homeopathy-specific items, including the homeopathic history, symptoms used in the decision, repertorization or analysis, the rationale for the prescription and precise medicine information when a case is written up.
A private consultation record is not a journal article. Not every session note needs an abstract, keywords or a causality score. The value of these tools for daily work is diagnostic of gaps: missing picture, unnamed sources, no rationale, no alternatives, no follow-up baseline and no consent for any later teaching use.
Teut and colleagues also summarize the WissHom documentation standard, which treats case analysis as a distinct area beside history, interview, prescription and follow-up. That separation is the same distinction this guide uses: analysis is a written step, not an unspoken moment between the chart and the plan.
Do not import research causality inventories into routine notes as if they measured efficacy. They are optional scientific instruments with their own limits. For private records, transparency of reasoning is enough.
How software should support analysis
Software is useful when it keeps the analysis attached to the case. It is harmful when it produces a polished conclusion that can no longer be inspected.
Test a tool with synthetic data before sending real notes.
| Capability | Test | Warning sign |
|---|---|---|
| Linked picture | Jump from an analysis field to the source phrase | Only a remedy name is stored |
| Named sources | See repertory edition and materia medica title | “AI match” with no work or version |
| Alternative candidates | Keep a second remedy visible beside the first | Interface shows a single winner |
| Contradiction field | Save a mismatch without deleting the candidate | Only positive keynotes can be stored |
| Versioning | Edit yesterday’s analysis and inspect history | Latest text overwrites the original |
| Human review | Reject a generated comparison | Draft becomes the record automatically |
| Language fidelity | Check that a quoted phrase is unchanged | The tool “corrects” the client’s words |
| Export | Reopen the analysis with links intact | PDF screenshot of a chart only |
AI may draft a comparison, retrieve a passage or propose a heading to read. Apply the same REASON steps to that draft. The AI repertorization software checklist covers vendor questions for retrieval tools. The digital records and privacy guide covers whether client text should be sent to a model at all. Neither replaces the analysis note.
Reusable case-analysis checklist
Before writing
- Safety, scope and concurrent care are recorded outside the repertory work
- Source notes distinguish client words, observation and interpretation
- The rubric set is justified and named
- The analysis has an owner and a due time
While comparing
- The characteristic picture is stated in a few lines or a short table
- Features withheld from comparison are named
- Each candidate has a reason for being on the list
- Leading and close alternative remedies were actually read
- Materia medica sources are identified
- Confirmations and contradictions sit in the same view
- Unknowns are marked instead of being completed by software
Before closing the analysis
- The working decision is dated and modestly worded
- Near-alternatives remain visible
- Review criteria and next owner are set
- Generated drafts are accepted, edited or rejected
- The original analysis will not be overwritten at follow-up
Frequently asked questions
Is a repertorization the same as a case analysis?
No. Repertorization is a structured search and comparison of indexed entries. Case analysis is the reasoned reading of that result against the whole case and materia medica. One can exist without a useful version of the other, and the record should not pretend they are identical.
How many remedies should be compared?
There is no universal number. Compare the candidates that the justified rubric set and your reading actually put in play, including at least one serious alternative. A long list that nobody read is weaker than a short list with named sources.
Must every consultation produce a prescription?
No. An analysis can conclude that information is missing, that referral is required, that the person should wait, or that two alternatives cannot yet be separated. Those outcomes still need a dated note.
Can I rely on keynotes instead of full materia medica?
Condensed keynotes can help you remember what to look up. They are a poor sole source for a differential because they omit context, contradictions and grading. If only a keynote list was used, say so.
Should the analysis be shown to the client?
Share the plan and the information the person needs in language appropriate to the consultation. The internal analysis may contain provisional comparisons that would be misleading if presented as a finished result. Follow local consent, access and professional rules.
How does follow-up change the analysis?
Follow-up first compares change with the previous baseline. If the picture used for analysis has changed enough to require a new comparison, write a new version. Do not revise the original note so that it appears to have anticipated the later course.
Keep the reasoning attached to the case
The purpose of case-analysis documentation is not to produce a more impressive chart. It is to keep the practitioner’s reasoning inspectable: what was considered characteristic, what was read, what did not fit and what will be reviewed next. That is also what makes supervision, later follow-up and software support safer.
Sources and further reading
- Samuel Hahnemann, Organon of Medicine, §§83–92 and §§101–104 — written case-taking, the case picture and later comparison
- Samuel Hahnemann, Organon of Medicine, §§152–156 — selection according to the more striking, singular and peculiar features
- American Institute of Homeopathy, The Organon, Section 153 — historical discussion of characteristic versus general symptoms
- Teut M. et al. (2022), Case Reporting in Homeopathy—An Overview of Guidelines and Scientific Tools — CARE, HOM-CASE and the WissHom documentation standard, including case analysis as a distinct documentation area
- CARE Case Report Guidelines and checklist
- EQUATOR Network, HOM-CASE extension to CARE
- Kent J. T., Repertory of the Homoeopathic Materia Medica (1897 digitized edition), Internet Archive / Wellcome Library — an historical repertory used here as an example of an index, not as a substitute for materia medica or judgment
This HomeoStudio knowledge-base article supports professional education and documentation quality. It makes no claim about the clinical efficacy of homeopathy and does not encourage self-diagnosis, self-prescribing or changes to prescribed care.