Case Analysis
The second prescription: the decision that separates practitioners
By Dr. Harish · October 2, 2026 · 5 min read
Choosing the first remedy is a skill you can study. Knowing what to do at the follow-up is judgement, and it is where most cases are actually lost.
Where cases are actually lost
Students spend nearly all their study time on the first prescription — case-taking, repertorisation, differentials. That is reasonable; it is teachable and examinable.
But most cases that go wrong do not go wrong at the first prescription. They go wrong at the second, when a remedy that was working gets repeated too soon, or a remedy that was working gets abandoned because the patient reported a bad week, or a remedy that never acted at all gets given a higher potency on the theory that it needs more force.
Kent devoted a great deal of attention to this and it remains the most useful part of his lectures. The core insight is simple and counterintuitive:
The second prescription is usually not a prescription.
Most of the time the correct action is to do nothing and to say so with confidence.
First, establish what happened
Before deciding anything, you need three facts. Get them in this order — the order matters, because asking about the complaint first anchors the whole conversation to it.
- The generals. Energy, sleep, appetite, mood, thermal state. How is the person?
- The particulars. What has the named complaint done — better, worse, moved, changed character?
- The timing. When did any change begin, relative to the dose? A change that started before the remedy was taken is not the remedy's doing, and patients routinely compress the chronology.
If you cannot answer all three, you are not ready to prescribe. Ask more questions. "The follow-up was too short" is a much cheaper mistake than a wrong second prescription.
Six situations
1. Clear improvement, still continuing
Do nothing. Do not repeat. Do not raise the potency.
This is the hardest instruction to follow, because doing nothing feels like not practising, and the patient came for something. Say what you are doing and why: the remedy is still acting, and interrupting an action that is working is how a good case gets muddled.
Set the next review by the pace of the illness and stop there.
2. Improvement that has clearly stalled
The patient improved, plateaued, and has been flat for a while.
Repeat the same remedy. Same potency first, if the earlier response was good. This is the classical indication for repetition, and the only one that is unambiguous.
3. Improvement, then clear relapse to the original state
The patient improved substantially, then slid back to roughly where they started.
Repeat, usually higher. A good response followed by relapse is evidence the remedy is right; the question was duration, not selection. This is the situation where a potency increase is genuinely indicated — and note how narrow the indication is compared with how often potency gets raised in practice.
If this happens repeatedly — acts well, relapses, acts well, relapses — something is sustaining the state. Look for a maintaining cause before you keep escalating. That mechanism is set out in Organon §5 and the maintaining cause.
4. Brief aggravation, then improvement
The classical homeopathic aggravation: short, involving the existing symptoms rather than new ones, followed by real improvement in the generals.
Do nothing. Wait. Warn the patient in advance that this can happen, because a patient who is not warned will stop the remedy and, quite reasonably, lose confidence.
A long aggravation, or one accompanied by deterioration in the generals, is a different thing entirely and should not be waved through. Check three things: when it started relative to the dose, how long it has run, and — decisively — whether the generals are worse or better.
5. Nothing has happened at all
No change, in either direction, after a reasonable interval.
Do not raise the potency. This is the commonest error at this juncture. Nothing happening usually means the remedy was not similar enough, and a stronger dose of a wrong remedy is still a wrong remedy.
Instead, work through this order:
- Was it actually taken, as directed? Ask directly and without reproach; non-adherence is ordinary.
- Is there an obstacle to cure — an ongoing exposure, a maintaining cause, a medication?
- Re-take the case. Not re-read your notes — re-take it. Patients disclose different material at a second visit, once they have decided you are worth telling.
- Only then reconsider the remedy.
6. The picture has changed into something else
New symptoms, a different character, a different centre of gravity.
Re-take and re-analyse. Do not force the original prescription onto a state it no longer matches. But first ask the direction question: has something deeper appeared while the surface complaint cleared? That is suppression, not progress, and it is read through the direction of cure.
The quick reference
| Follow-up picture | Action |
|---|---|
| Improving, still moving | Wait |
| Improved, now flat | Repeat, same potency |
| Improved, then relapsed | Repeat, usually higher |
| Short aggravation, then better | Wait |
| Nothing at all | Re-take; do not escalate |
| Picture changed | Re-take; check for suppression |
Make it falsifiable
The habit worth more than any table: before the follow-up, write down what you expect to change and by when.
It takes fifteen seconds and it changes the follow-up completely, because you arrive with a prediction instead of an impression. When the prediction is met, you have learnt something about the remedy. When it is not, you have learnt something about your case-taking. Without the prediction, both visits blur into a vague sense that things are roughly all right, which teaches nothing at all.
Ten cases recorded this way are worth fifty without it.
Dr. Harish
Homeopath and founder of HomeopathyWorld. Writes on materia medica, case-taking, repertory and the Organon for students and practitioners.
Put this into practice
Repertorise these symptoms against the real Kent & Bönninghausen, read the full Materia Medica, and practise the case on an AI patient — free account, no card required.
Create your free accountMore in Case Analysis
Discussion
Sign in to join the discussion.
No comments yet — start the conversation.