A clinical question can begin a valuable learning activity. A record that the question was asked does not, by itself, explain what the clinician learned, whether the answer was checked or what will change in practice.
The expansion of clinical AI makes this distinction increasingly important. Some products record eligible activity for educational credit. Others help users practise a topic and produce a reflective learning record. Both can be useful, but a credit, a transcript and a personal reflection are not interchangeable.
The Doximity-UpToDate announcement creates another reason to check which learning functions belong to the specific product and account being used.
Three records that answer different questions
A search history answers: what did the user ask? It may help someone return to a source, but it can include abandoned questions, repeated prompts or material that was never reviewed.
An educational credit answers a different question: did the activity qualify under a particular provider's scheme and applicable conditions? The existence of a certificate or credit depends on those conditions, not on whether the interface looks educational.
A reflective record explains the learner's need, activity, understanding and intended application. Its value comes from the connection between those elements. A beautifully written paragraph that does not represent the user's learning is weaker evidence than a short, accurate account of a genuine change in understanding.
UpToDate Expert AI already has a CME proposition
Wolters Kluwer announced CME capture within UpToDate Expert AI on 18 March 2026. Eligible questions and dialogue can form part of that workflow. This is an established product capability and should be acknowledged in comparisons with newer learning tools.
However, a function offered in one product configuration does not automatically accompany every partner integration. The October Doximity announcement should not be read as confirmation of the credit arrangements for every embedded interaction.
A clinician relying on credit should check the exact service, account eligibility, completion steps and export process. A hospital should establish whether the intended users can retain the evidence when their institutional access changes. Those are practical entitlement questions, not minor administrative details.
iatroX approaches the task through an observable learning activity
iatroX CPD starts with a selected professional and clinical-content context, followed by a focused question session, Tutor-supported review and a user-confirmed record. Its published workflow includes PDF export and direct FourteenFish export for linked accounts. CPD is included in its all-access subscription rather than sold as a separate add-on.
This is a useful proposition for clinicians who want to show the work behind a reflection. The relevant educational object is not merely the final prose. It is the relationship between the initial gap, the attempted questions, the explanation considered and the user's own conclusion.
The workflow should still be described accurately. A generated draft is not a completed personal reflection until the user has reviewed and made it their own. A platform's framework suggestion is not a professional judgement about what evidence must be accepted.
An example of a worthwhile reflection
Consider a fictional learning session about interpreting research results. The learner initially treats a large relative improvement as sufficient evidence of a large practical benefit. After reviewing a worked example, they recognise that the baseline risk and absolute difference are also needed.
A useful personal record might say: "I was focusing on the relative figure. In this exercise I recalculated the absolute difference and recognised why the same relative effect can have different practical implications. In future, I will look for the baseline event rate before summarising the result."
This example is original and illustrative. It does not claim a patient benefit that has not occurred. It names the misunderstanding, the learning activity and a specific intended action. A later reflection could revisit whether the action was actually adopted.
Honest limitations strengthen a learning record
A record can legitimately state that a question remains unresolved. The activity may have revealed a conflict between sources, an unfamiliar method or a need to discuss the issue with a colleague.
What weakens the record is pretending that exposure equals mastery. "I am now confident in all aspects of this topic" is a large claim after a short activity. "I can explain this distinction but need more practice applying it" is more informative when it accurately describes the learner's position.
The Faculty of Public Health's guidance on reflective notes emphasises relevance and impact rather than lengthy descriptions of an event. Its scheme is profession-specific, but the underlying distinction is useful: the account should explain what the activity meant for the learner, not simply reproduce its contents.
AI should assist the record, not invent the learning
Useful assistance includes organising the activity, prompting the user to identify a change in understanding and suggesting a structure. The user must decide whether the suggested wording is true.
A system should not convert a wrong answer into a claim of demonstrated competence, invent a clinical encounter or state that practice changed merely because the user intends to change it. Those distinctions are especially important when a polished draft makes an unsupported claim sound plausible.
A practical final review asks whether each statement refers to something actually done, understood, checked or planned. Any statement about impact should identify whether the impact has been observed or remains an intention.
Keep the portfolio useful and proportionate
The record should contain enough context to make sense later without accumulating unnecessary clinical detail. A learning need can usually be described at the level of the concept or decision, using a fictional example where an illustration is helpful.
Do not add identifiable patient material to a general learning tool to make a reflection appear more authentic. The educational point is the reasoning and intended improvement, not the reconstruction of a person's record.
Portability also matters. Before making any platform the main home for learning evidence, check how records can be exported and retained. A collection of valuable reflections is less useful when the clinician cannot retrieve it at the point it is needed.
Choose the evidence route that matches the need
A clinician seeking eligible CME credit should examine the provider's specific credit arrangements. A clinician seeking a structured account of a learning gap may value questions, explanation and reflection more directly. Some will benefit from both.
The practical advantage of iatroX is the connection between practice and a record of that practice. It offers a way to make continued learning visible without pretending that a generated document is equivalent to accreditation, revalidation or proof of competence.
Frequently asked questions
Will every UpToDate answer inside Doximity earn CME?
That should not be assumed. UpToDate Expert AI has an announced CME capability, but the applicable eligibility and capture arrangements for a partner deployment need to be confirmed separately.
Is an iatroX CPD record an accredited CME certificate?
A professional learning record and an accredited credit are different things. Use the record to document genuine activity and check the requirements of the relevant professional scheme rather than assuming automatic equivalence.
Can AI write my final personal reflection for me?
It can help organise and prompt the reflection, but the final account must accurately represent your learning. Review the wording, remove invented implications and confirm what you actually did and intend to do next.
Build a focused learning activity and review its evidence record with iatroX CPD.