An active clinician is a person who meets a stated professional definition and performs a stated activity within a stated period. Without those three elements, an adoption number may describe reach, registration or eligibility rather than current use. The remedy is a reporting dictionary, not a more impressive headline.
This article is published by iatroX and applies the same questions to iatroX's own supplied figures. The framework below is an original due-diligence method, not a claim that a universal industry definition already exists. Examples are hypothetical unless explicitly attributed to the September 2026 iatroX platform data.
Start with the person, not the account
A registered account is not necessarily a unique person. A verified email address is not necessarily a verified professional identity. A professional identity check is not necessarily confirmation of current licensure in the relevant jurisdiction. A clinician eligible to access software is not necessarily a clinician who has used it.
Ask what verification establishes, when it occurred and how duplicate or changed accounts are handled. An institution may provision accounts before anyone logs in. A clinician may have personal and organisational accounts. A test account may perform thousands of actions without representing professional adoption.
Do not infer misconduct from these possibilities. They are ordinary reasons to specify the metric. A clear definition allows a supplier to explain exactly what its number represents and gives a buyer a fair basis for comparison.
Define activity that relates to the product
For a reference tool, an activity could be a submitted clinical question or an inspected answer. For a simulator, it could be a started case, completed encounter or reviewed debrief. For a learning bank, it could be an attempted question with an explanation opened. These events answer different questions.
An illustrative definition might be: a deduplicated, professionally verified person who completed at least one eligible learning task during the previous calendar month. That is a proposed definition, not iatroX's current reporting policy. A service may choose another sensible threshold, provided it explains it and reports consistently.
Retain supporting measures rather than overloading one label. Report people, tasks, sessions and active days separately. If a user submits many short queries during one sitting, that may indicate engagement, difficulty or repeated reformulation. The event count alone does not establish which explanation is correct.
A reporting dictionary worth requesting
| Metric | Minimum definition to disclose |
|---|---|
| Registered accounts | Creation rule, reporting date and treatment of duplicates |
| Verified clinicians | Verification method, professions, geography and rechecking policy |
| Active clinicians | Qualifying activity and time window |
| Completed tasks | Start, completion and failure definitions |
| Retained users | Entry cohort, return window and required activity |
| Institutional reach | Eligible seats or sites, clearly separated from observed use |
This dictionary should accompany the data, not appear only when someone challenges it. Version it when definitions change. If an improved deduplication process reduces the total, explain the methodological change rather than hiding it inside a revised growth chart.
A useful due-diligence request is a reconciliation: show how the reported headline can be reproduced from the underlying event definitions, after exclusions. The reviewer need not receive identifiable clinical queries to inspect whether the arithmetic and categories are coherent.
Apply the same standard to iatroX
The figures supplied for this article are: 1 million uses, over 50,000 verified clinicians, 40 countries, primarily the UK, and over 15% of UK GPs, per iatroX platform data, September 2026. They describe reported reach. They should not be converted into monthly active users, paying subscribers, students or users of a subsequently launched feature.
The public methodology page, checked on 19 September 2026, uses questions answered in its headline presentation. The relationship between that wording and the supplied term "uses" needs an explicit definition rather than silent substitution.
The publication data note therefore requires: [CONFIRM: the event counted as a use and its relationship to questions answered]; [CONFIRM: the cumulative reporting period and precise cut-off date]; [CONFIRM: clinician verification and deduplication method]; [CONFIRM: country attribution method]; [CONFIRM: the UK GP numerator, denominator source, reporting date and treatment of trainees or other GP categories].
Until those fields are completed, retain the attributed figures as supplied and do not calculate unsupported rates from them. In particular, a UK GP percentage needs a clearly matched workforce denominator. A count of people and a full-time-equivalent workforce estimate would not be interchangeable denominators.
Retention adds the missing time dimension
Two tools may each report a large cumulative audience while serving very different current populations. Ask how many people returned, what they did and how long after their first meaningful task. Separate examination-season use from ongoing reference use rather than treating every inactive account as product failure.
A hypothetical cohort of 1,000 new registrants might contain people who never completed a task, people who used one free resource and people who repeatedly returned. Reporting only the cohort's aggregate question count would conceal those distinctions. A small group could account for much of the activity.
For an institutional deployment, distinguish the contract population from enabled accounts, first use and repeated use. A large rollout announcement establishes an opportunity to serve clinicians; it does not, by itself, establish sustained adoption or clinical benefit.
What the numbers can legitimately support
Reach can support a claim that people have encountered or used a platform under a disclosed definition. Retention can support a claim about continued activity. Neither establishes accurate answers, improved learning or better patient outcomes without appropriate evaluation.
For iatroX, the reference methodology and educational designs should be assessed as designs, while learning and reliability claims require their own evidence. A transparent adoption statement is valuable precisely because it does not ask one number to prove everything.
Frequently asked questions
Is a verified clinician automatically an active user?
No. Verification concerns identity or eligibility, while activity requires a defined event and reporting period.
Can I compare consultations with registered clinician counts?
Not as a league table. Preserve the original units and compare like-defined measures within their own categories.
Why are confirmation fields included with iatroX's figures?
The supplied totals do not define every denominator or event rule. Those details must be confirmed rather than invented before stronger interpretations are published.
Explore transparent clinical AI evaluation with iatroX Insights →