01Define the device before choosing a pathway
Write an intended-use statement that can anchor your evidence file.
Self-paced
Define the device before choosing a pathway
Write an intended-use statement that can anchor your evidence file.
A device pathway depends on its intended use and jurisdiction. A quality certificate, establishment registration and product authorization answer different questions. Use the original regulator record for the precise product and scope.
02Build a standards applicability matrix
Connect each applicable standard to its edition, scope and evidence.
Self-paced
Build a standards applicability matrix
Connect each applicable standard to its edition, scope and evidence.
Check the regulator’s recognition or harmonisation record, including any limits. A standard number without an edition and scope is an incomplete reference. ASCA participation also requires checking the laboratory’s current status.
03Review biological safety of the finished device
Recognise when a materials statement leaves an evidence gap.
Self-paced
Review biological safety of the finished device
Recognise when a materials statement leaves an evidence gap.
Biocompatibility assessment considers the finished device and its contact with the body. Manufacturing, sterilisation and residuals can matter. The evidence must fit the nature and duration of contact; a raw-material certificate alone may not address the finished product.
04Separate cleaning from sterilisation evidence
Spot an unsupported claim in a reusable-device validation package.
Self-paced
Separate cleaning from sterilisation evidence
Spot an unsupported claim in a reusable-device validation package.
Reprocessing must address the particular device and its instructions. Complex channels and joints can retain soil. Cleaning validation should use a relevant challenge and measure suitable residual soil; spore reduction alone is not a substitute for demonstrating cleaning.
05Examine the user, setting and interface
Turn a use-error scenario into a focused evidence question.
Self-paced
Examine the user, setting and interface
Turn a use-error scenario into a focused evidence question.
Human-factors work connects users, use environments and interfaces. Relevant users may include patients, caregivers and staff who set up, maintain or reprocess the device. Labelling and training are part of the interface, not the whole interface.
06Connect the claim to clinical evidence
Frame evidence around the intended population, use and outcomes.
Self-paced
Connect the claim to clinical evidence
Frame evidence around the intended population, use and outcomes.
Clinical investigations assess device safety and performance; IVD performance studies address diagnostic performance. A registry record helps locate a study, but it does not by itself establish a favourable result or authorize a product.
07Ask the right diagnostic-evidence questions
Organise an IVD evidence review without relying on a single accuracy headline.
Self-paced
Ask the right diagnostic-evidence questions
Organise an IVD evidence review without relying on a single accuracy headline.
Use a transparent diagnostic-study report to examine the intended use, participants, test and reference standard. Reporting guidance helps reveal missing information; it does not replace critical appraisal or the applicable IVDR requirements.
08Review software and cybersecurity changes
Link software changes to evidence and regulatory review.
Self-paced
Review software and cybersecurity changes
Link software changes to evidence and regulatory review.
Software evidence depends on the device’s functions and intended use. Use current guidance to examine lifecycle documentation and cybersecurity. An AI authorization list is a starting point: the individual record defines the authorized device and scope.
09Read a safety signal responsibly
Distinguish a report, a signal and a confirmed corrective action.
Self-paced
Read a safety signal responsibly
Distinguish a report, a signal and a confirmed corrective action.
MAUDE reports can be incomplete, unverified or biased. Counts alone cannot establish incidence, compare device event rates or demonstrate causation. Read signals alongside the official alert and other evidence.
10Purchase a usable lifecycle, not just a unit
Turn the purchasing specification into a verifiable acceptance plan.
Self-paced
Purchase a usable lifecycle, not just a unit
Turn the purchasing specification into a verifiable acceptance plan.
Procurement planning should connect the clinical need with technical requirements and the resources needed to use the device. Include installation, training, consumables and service when comparing options. Our cost worksheet is an editable planning model, not a price quotation.
11Connect inventory, service and end of life
Build a usable asset record for clinical engineering.
Self-paced
Connect inventory, service and end of life
Build a usable asset record for clinical engineering.
A maintenance programme includes inspection, preventive work and corrective work. Link records to a unique asset, its use and service history. Decommissioning needs an explicit decision and documented process rather than simply deleting an inventory row.
12Make a literature search reproducible
Create a transparent search and screening log.
Self-paced
Make a literature search reproducible
Create a transparent search and screening log.
Record the database, exact strategy, date and screening decisions. Use reporting guidance appropriate to the review and study design. A search result is a candidate source: read the full study and appraise its relevance and limitations before using its conclusions.
Continue with the source organisations.
EMDN nomenclature & training
Explore the nomenclature, version information and official training materials.
Open training resources ↗Choose a reporting guideline
Match the reporting checklist to the study design, including CONSORT, STROBE and STARD.
Open training resources ↗Your reading list, learning progress and worksheets stay in this browser. They are not synced to your member account. Export a backup before changing devices or clearing browser data. Do not enter patient identifiers or confidential product information.
Editorial review: 3 October 2026. Source publication dates and versions remain those of the original publishers.
From source to practice
Prepare your scientific review with clear steps
16 guides across quality, manufacture, registration, safety and research, with scope, source versions and saveable working checklists.

