A clinical evaluation according to MEDDEV 2.7.1 is carried out in 5 steps.
Clinical Evaluation: Preparation or Step 0
Preparation or step 0 is the creation of the clinical evaluation plan based on the general safety and performance requirements to be addressed.
This plan includes the following:
a) Product description
b) Design features, intended use, indications, contraindications and target groups that require special attention.
c) Data and results of the risk assessment.
d) Preclinical data from verification and validation steps.
e) State of the art in the relevant medical field, standards and guidelines. Information related to the medical aspect, its natural history, benchmark devices, other devices, gold standards and medical alternatives.
f) Data sources and type of data. Relevant data can be generated by the manufacturer and / or drawn from the scientific literature.
g) Information for the assessment of equivalence, if equivalence can possibly be claimed.
h) Clinically relevant product changes (for devices already CE marked).
i) Newly created clinical risks (for devices already CE marked).
j) Aspects from the post-market -surveillance that are to be updated regularly in the clinical assessment report (for devices already CE marked).
k) Who and with what qualifications works on this team?
Clinical Evaluation: Step 1
In step 1, the product to be considered, its clinical and technical status are described and the aspects to be considered and the necessary data to be collated are defined. Data from various sources is collected.
a) Product description and information accompanying the product,
b) Data from your own risk assessment
c) Data from your own market surveillance and product monitoring
d) Vigilance data (own and those from the databases at BfArM, FDA, swissmedic etc.)
e) Preclinical data from verification and validation measures
f) Own data from clinical trials
g) Clinical data from published studies in relevant databases (ATTENTION: equivalence discussion)
Clinical Evaluation: Step 2
In step 2, the existing data are evaluated and their suitability for this clinical evaluation is determined.
Appropriate rating scales should be developed for the assessment so that the assessment can be understood by third parties. You can use a numerical system, for example distribute 1 to 10 points for qualification of the authors, the evidence, the product proximity, etc. or also assign “-” and +++ “. How you do this is up to you. Check here whether all clinical aspects can be substantiated with data. Look out for information gaps and fill the gaps with appropriate data.
Clinical Evaluation: Step 3
In step 3, the data is critically assessed. Positive and negative data are taken into account. Recommendations for measures and the further collection of data are given.
At this point, you should make sure that you really take into account all previously collected and evaluated data and record why the data support your performance statements or relate to possible residual risks. When using clinical data from databases, remember to show that the data is transferable to your product. Have an equivalency discussion and compare the technical, biological and clinical characteristics of your product with the product from the clinical study.
Clinical Evaluation: Step 4
In the last step 4, the above-mentioned steps are summarised as a result and approved with the signatures of those involved.
All aspects should be listed again here, i.e., who prepared this report for to what end and how. If the state of the art is complied with, preclinical data, vigilance data etc. are available the clinical suitability can be confirmed. Discarded data will not be taken into account or logged.
If you can substantiate all clinical aspects with suitable data from the literature, it is not necessary to conduct your own clinical trials. Make sure, however, that you discuss any gaps in information critically and close them in an appropriate manner. Information gaps can be filled with laboratory tests or clinical tests. Negative information must also not be withheld. Rather, these must be evidently invalidated.