Content methodology and AI assistance
DosePractice has been developed with substantial AI assistance. AI has helped draft code, question templates, explanations, tests, documentation and website content. DosePractice remains responsible for deciding what is published.
AI output is not treated as authoritative. Human error, drafting mistakes and technical defects can also occur. The expanded bank is automatically checked for arithmetic consistency, but automated verification does not establish educational quality, regulatory status, clinical suitability or freedom from every wording error. Layered controls are used to reduce—not eliminate—the risk of error.
How the question bank is produced
- Question structures are limited to defined educational arithmetic domains.
- Values are generated deterministically so each question has a known answer and reproducible working.
- Automated tests verify identifiers, answer calculations, topic counts and privacy safeguards.
- Questions use fictional values and deliberately exclude medicine choice, contraindications, maximum doses, routes, patient factors and treatment recommendations.
- Every question has a stable identifier and reporting route so wording, source or calculation concerns can be investigated and corrected. Any future external review will be described only within its genuine recorded scope.
Authorship and editorial responsibility
AI is not presented as the author or professional reviewer. DosePractice is independently operated and remains responsible for publication decisions, corrections and accurate descriptions of how content has been checked. It does not imply that an independent review team exists.
Why this disclosure is available
AI assistance does not remove the publisher's responsibility for accuracy, originality, transparency and useful educational content. Nursing-related educational material warrants clear authorship, source and review information. This page is linked from the Trust, FAQ, Terms and Content review pages rather than repeated beside every question.
What will not be done
- AI-generated material will not be published solely to create hundreds of thin search pages.
- AI will not be described as a nurse, doctor, regulator or independent reviewer.
- The disclosure will remain accessible through the Trust, FAQ, Terms and Content review pages.
- Automated checks will not be presented as legal, MHRA, NMC or university approval.
Controls used to limit error
- Formulae and worked steps are displayed so learners and reviewers can inspect the arithmetic.
- Deterministic question generation and automated tests check answers, identifiers, topic counts and regressions.
- Arithmetic methods, content versions and material correction decisions are recorded for traceability.
- Wording, ambiguity, units, rounding and educational boundaries are considered through testing, user reporting and correction controls; any future external review will be described only within its actual scope.
- Users can report a problem without creating an account, and material corrections must be logged, regression-tested and deployed through the rollback process.
No combination of human review or automated testing can guarantee that every error has been removed. Users must independently verify formulae, units and results using an appropriate trusted source.
Trust and transparency · FAQ · Content-review process · Privacy by design · Report a problem