Our AI feature is grounded in four key sources of information. Together, these help ensure its responses are clinically accurate, relevant to the topic, aligned with the questions learners are answering, and personalised to their performance.
1. NICE Guidelines — Critical Clinical Source
NICE Guidelines are a particularly important source for the AI.
They provide authoritative, evidence-based guidance on clinical practice in the UK and should be used to ground clinical recommendations, management pathways and other guidance-sensitive information.
Where textbook or question content differs from current NICE guidance, this should be identified and handled carefully rather than allowing older educational content to override current clinical guidance.
Purpose: Helps ensure the AI's clinical information reflects current, authoritative UK guidance.
2. Textbook Topic Content
The AI uses the relevant Pastest textbook topic as its core educational context. But it's not just any old content - at Pastest we have many decades of experience in understanding what topics and themes commonly appear in medical exams, so if the AI references it, there's a good chance it will feature in the exam in some way.
This gives it:
- The clinical knowledge associated with the topic
- Key concepts, terminology and explanations
- The appropriate level and scope of information for the learner
Purpose: Keeps the AI focused on the curriculum and the specific subject being studied.
3. Question Content
We also provide context from our highly accurate question bank, including the questions and associated educational content.
This helps the AI understand:
- What learners are expected to know
- How knowledge is assessed
- Common clinical scenarios and decision points
- The reasoning and learning points associated with questions
Purpose: Connects the AI's support directly to the learning and assessment experience.
4. User Performance Data
The AI can use relevant learner performance data to understand where an individual may need additional support.
For example, performance data can indicate:
- Topics or concepts a learner is struggling with
- Areas of relative strength
- Patterns in correct and incorrect answers
- Where further explanation or revision may be useful
Purpose: Enables more relevant and personalised learning support rather than giving every learner the same experience.
In Simple Terms
The AI combines:
NICE Guidelines → What is the current authoritative clinical guidance?
Textbook content → What does the learner need to understand?
Question content → How is that knowledge tested and applied?
Performance data → What does this learner need help with?
Together, these sources allow us to create an AI learning experience that is clinically grounded, curriculum-relevant, assessment-aware and personalised.
Important Principle
The AI should not simply generate answers from general knowledge. Its responses should be grounded in the content and evidence we provide, with current NICE guidance acting as a key clinical authority wherever applicable.