Clinical search used to disappear. A doctor looked something up, acted on it, and moved on. The learning often genuinely happened — the doctor knew something they did not know before — but it was not captured. No record. No reflection. No appraisal evidence. No CPD credit. The knowledge entered working memory and either stuck through subsequent reinforcement or faded when the question did not recur.
AI changes this — not because AI is smarter, but because AI clinical search creates a structured trace that can become documented, reflectable, appraisal-ready professional development.
Why AI Changes the Learning Trace
When a clinician uses an AI clinical search tool, the interaction creates a natural learning record. The question asked — what the clinician needed to know. The source used — which guideline, formulary, or evidence base. The answer found — what was learned. The clinical uncertainty — what prompted the question (a knowledge gap, an unfamiliar presentation, a guideline change). The learning point — how this changed understanding or confidence. The practice impact — whether the answer affected a clinical decision or will affect future management.
These elements map directly to the GMC's CPD framework. The GMC's guidance on continuing professional development covers planning, carrying out, evaluating, and recording the impact of CPD activities. The supporting information guidance directs doctors to CPD guidance for planning, reflecting, and recording. Good CPD should relate to the clinician's scope of practice, address identified learning needs, use credible sources, and lead to a reflectable change in knowledge, confidence, or behaviour.
A clinical search that addresses a real learning need, uses an authoritative source, and changes clinical understanding or confidence is — by definition — a CPD activity. The challenge has always been capturing it in a format that serves appraisal and revalidation.
When a Clinical Question Becomes CPD
Not every search is CPD. Checking a drug dose the clinician already knows is confirmation. But when a search addresses a genuine learning need, it crosses the CPD threshold.
The criteria: it addressed a learning need (the clinician did not already know the answer or was uncertain). It related to patient care or professional practice. It used a credible source (NICE, CKS, BNF, peer-reviewed evidence, SmPC). It changed knowledge, confidence, or behaviour. The clinician can reflect on impact.
Examples of CPD-Ready Clinical Searches
"What is the current guidance on HRT in migraine with aura?" — checking NICE guidance after a consultation that prompted uncertainty about contraindications. Learning: updated understanding of absolute vs relative contraindications, risk-benefit discussion framework, impact on future consultations with similar patients.
"How should I safety-net a child with fever and no focus?" — reviewing safety-netting best practice after a consultation where the clinician felt uncertain. Learning: specific red flags (non-blanching rash, prolonged capillary refill, reduced consciousness, inconsolable crying), time-bound return criteria, documentation standards, parent communication techniques.
"When should I refer suspected inflammatory back pain?" — confirming NICE referral criteria for axial spondyloarthritis. Learning: inflammatory vs mechanical features (morning stiffness >30 mins, improvement with exercise, age of onset <45), referral threshold, appropriate initial investigations (ESR, CRP, HLA-B27).
"What monitoring is needed for methotrexate?" — checking SmPC and BNF before initiating shared care. Learning: FBC, LFTs, U&Es before starting, then every 2 weeks until dose stabilised, then monthly. Shared care responsibility transition. When to withhold.
"Which calculator should I use for VTE risk?" — clarifying whether YEARS or Wells applies. Learning: YEARS criteria (clinical signs of DVT, haemoptysis, PE most likely diagnosis), simplified approach, D-dimer thresholds, imaging pathway decisions.
Each is a real question from real practice, using an authoritative source, producing a genuine change in knowledge. Each qualifies as CPD.
What a Good CPD Reflection Should Include
Date. Clinical question. Learning need (why this arose). Source consulted. Key learning. Impact on patient care. Further action needed. Time spent. Domain (clinical, professional, academic, quality improvement).
The GMC does not prescribe a specific format. But a structured reflection linking a clinical question to an authoritative source, a learning outcome, and a practice impact meets revalidation requirements — and is substantially more valuable than a generic "attended a webinar" entry.
How iatroX Makes This Easier
iatroX is designed so clinical questions do not disappear. The workflow: ask a question → receive a cited answer → reflect on learning → save as a CPD record with source citations automatically included → link to relevant Q-bank topics for retrieval practice.
The citation becomes the source evidence for the reflection. The learning point becomes the reflection content. The practice change becomes the impact statement. The process — from question to appraisal-ready CPD — happens in one platform, in minutes, without requiring a separate write-up days later when details have faded.
Source citations strengthen the appraisal evidence. The appraiser sees not just "I learned about X" but "I learned about X from NICE guideline NG28, which changed my approach to Y in patients with Z." That is CPD with provenance — substantially more credible and more useful than undocumented learning.
The Bigger Shift
The questions doctors ask during real clinical practice are the most authentic learning needs they have. Capturing those questions, linking them to authoritative answers, and recording the learning as CPD transforms clinical search from a disposable act into professional development that serves appraisal, revalidation, and — most importantly — better patient care.