A PSA prescription can be marked wrong because the medicine is unsuitable, another required element is incorrect, the selected formulation does not match the intended prescription or the automated system does not recognise an otherwise credit-worthy variation. Do not assume either that the computer must be right or that choosing the intended medicine makes the whole prescription correct.
The official PSA explanation, reviewed on 24 September 2026, describes automated matching against a mark scheme and post-assessment review of unrecognised responses in the real assessment. A practice platform's immediate result is therefore not the same thing as a final official assessment decision.
Correct medicine, incomplete prescription
Treat a prescription as a connected set of decisions. Product selection, formulation, route and the other requested fields need to make sense together and answer the task in the stem. Knowing the medicine's name is only part of that work.
As described by the PSA on 24 September 2026, prescription-writing questions allocate ten marks: five for medicine choice and five for dose, route and frequency. Timing and signature fields are populated automatically and do not earn marks. This distinction explains why learners should examine the component feedback rather than reduce the result to "I chose the correct drug".
Before disputing an answer, reconstruct exactly what was submitted. A learner may remember the intended prescription while the saved response shows a different selection. That is an entry error, not necessarily a knowledge deficit or a marking defect. Conversely, a technically complete entry can still be inappropriate for the scenario.
A prescription-error taxonomy
| Error category | What to look for | Useful next learning task |
|---|---|---|
| Task mismatch | The response answers a different question from the one asked | Restate the requested action before selecting any product |
| Medicine-selection error | The medicine does not fit the supplied scenario | Identify the clinical discriminator or overlooked limitation |
| Formulation mismatch | The selected product differs from the intended preparation | Practise reading the full product description |
| Inconsistent fields | Entries do not form a coherent prescription | Check the relationships between the fields, not each in isolation |
| Missing required information | A relevant entry was omitted | Develop a final completeness check |
| Calculation or unit error | The method or unit handling was incorrect | Rebuild the calculation independently before checking the entry |
| Unrecognised response | A potentially valid variation is not matched automatically | Use the official or provider-specific query process |
| Explanation discrepancy | Feedback appears to conflict with its own source | Record the source, context and exact disputed claim |
This is an educational classification, not an official PSA marking scheme. Several categories can coexist. For example, an inappropriate formulation may explain why an intended entry was not available, rather than demonstrate that the interface is defective.
Why formulation matters even when the active ingredient is familiar
Consider an original fictional scenario in which a student intends to select a particular preparation but clicks a different release formulation in a long product list. The student later describes the answer by active ingredient alone and cannot understand the lost marks. The first question is not whether that ingredient can ever be used; it is whether the selected product matches the scenario and the rest of the prescription.
A second fictional learner chooses a route inconsistent with the preparation they selected. Again, listing a recognised medicine is not enough. The prescription should be assessed as a whole.
These examples deliberately avoid treatment instructions. Their purpose is to show the information a learner must preserve when investigating feedback. Record the full product description and the task requirements, rather than using a shortened label that removes the source of the disagreement.
Understand what automated matching can and cannot tell you
An automated marker needs a defined way to recognise acceptable submissions. A mismatch may identify a genuine error, or it may reflect a variation not captured by that implementation. The official process accounts for this possibility through review; commercial practice products may use different processes and response times.
That does not make every rejected answer defensible. "The system did not recognise it" is a hypothesis to investigate, not an explanation that should be accepted whenever the result is disappointing. First determine whether the response satisfies the question and the relevant source, then use the appropriate review route.
Nor does an accepted answer prove that you understood it. A learner can obtain credit after selecting a familiar pattern without noticing why the scenario supports it. For educational purposes, inspect uncertain correct prescriptions as well as rejected ones.
A verification workflow for a disputed practice answer
Start by saving the item identifier, date, your submitted response and the provider's feedback within the service's permitted use. Do not publish the question or redistribute a screenshot containing proprietary material. Write a short account of the disagreement in your own words.
Next, separate the task from the treatment knowledge. What was the question asking you to prescribe or decide? Which patient details constrain the answer? Does your proposed explanation rely on information that was not actually supplied?
Then consult the relevant permitted assessment reference or an appropriate current learning source. Identify the precise section, product, indication and context rather than citing a whole website. A statement about a different population or formulation may not resolve the disputed item.
Finally, submit a focused query. A useful message is: "For item X, my submitted product and fields were Y. The feedback identifies Z as incorrect. The cited reference appears to support this specific interpretation in the stated context. Could you clarify the assumption or review the marking?" Keep confidential material inside the provider's support channel.
What to do while a query remains unresolved
Do not let one disputed item consume the entire revision session. Mark the uncertainty, record the most defensible interpretation and continue with a different example of the underlying skill. A disagreement can reveal a learning need even before the provider replies.
If the issue concerns a product-selection interface, practise reading and checking product descriptions. If it concerns source interpretation, work on identifying population, indication and qualification. If it concerns arithmetic, check the setup independently. Each task remains useful without pretending that the disputed mark has been settled.
For the real assessment, follow the official communication and review arrangements. A response from a commercial bank, a tutor or an online discussion group cannot amend an official result.
Use AI to analyse the disagreement, not to overrule it
A productive AI prompt asks for the assumptions that could explain the discrepancy and the source that would resolve it. An unproductive prompt asks the system to reassure you that your answer should have received full marks. Agreeable language is not adjudication.
Per iatroX product information, September 2026, Ask-iatroX is a free reference service grounded in NICE, CKS, SIGN and SmPC information from emc, with linked sources. Its retrieval and checking methodology supports traceability, not guaranteed correctness or authority over an examining body. It should not be represented as an official PSA marking engine or as an in-examination resource.
The Socratic Tutor, described in iatroX's September 2026 product information, can explore a misconception after an attempted question. For this problem, the useful goal is being able to explain the discrepancy and complete a different task correctly without assistance. The goal is not winning an argument with a practice platform.
Frequently asked questions
Can an unrecognised PSA prescription still receive credit?
The official PSA explains that unrecognised responses are considered during post-assessment review of the real examination. That does not mean every rejected practice answer is correct or will be treated identically by a commercial provider.
Does choosing the correct medicine guarantee full prescribing marks?
No. The prescription must also satisfy the other required elements and the task's context; inspect the component feedback rather than the medicine name alone.
Can iatroX decide whether my official PSA result is wrong?
No. It can support explanation and source-based learning, but only the relevant assessment process can determine official marking and review outcomes.