After a long period of waiting, several delays, and the MDR deadline without proper guidance hanging above our heads like the Sword of Damocles, as of end of April we have one more year until MDR date of application and a ‘flood’ of guidelines to work with.
So while keeping the clinical work on track as good as possible in the almost ‘fluid’ COVID-19 environment, I thought I’d give the ones regarding clinical evaluation a look.
At first sight there is not a lot of news, but a couple of things stood out from my perspective, especially since after a closer look they appear to give us a taste of the MDCG’s thinking and what we may expect from further guidance. In this blogpost I will address my observations on the guidance for the evaluation of Medical Device SoftWare (MDCG 2020-1), later this week followed by the guidance for equivalence (MDCG 2020-5).
Besides that MDCG 2020-1 addresses MDSW of all medical devices, including IVD’s, it is interesting to see that, although emphasizing that it “solely” applies to MDSW, this guidance uses IMDRF concepts as a starting point, such as N41. With several medical device guidances still pending, this gives us an idea of the MDCG’s thinking, and in the run-up to further publications it seems a good strategy to continue using existing guidelines at our disposal including the ones from the IMDRF.
The guidance includes a helpful table clarifying the clinical/ performance evaluation requirements in relation to intended purpose of the MDSW, indicating that MDSW with an independent medical purpose and MDSW with an intended purpose driving a medical device require an evaluation.
The scope also has a note specifying that
“If a software is driving/ influencing more than one medical device, an independent CLINICAL EVALUATION (MDR) / PERFORMANCE EVALUATION (IVDR) is required for each foreseen and clinically viable software – device combination”.
So in case you are using certain MDSW for multiple medical devices, you are unfortunately not done with just one evaluation, which makes perfect sense since the MDSW should meet the intended purpose in a real-world environment (also refer to below).
Also of interest I find the acknowledgement that
“the final clinical outcome for the patient is dependent on further diagnostic and/or therapeutic options which could be available”,
in other words the clinical benefit for MDSW may be indirect and just lie in providing accurate information.
MDCG 2020-1 nicely (re-)visualizes the general evaluation PROCESS, which is the same regardless whether we are dealing with medical devices, in vitro diagnostics devices, or MDSW, and describes the 6 stages we are (supposedly) familiar with: planning/ scoping, data identification, data appraisal, data analysis within the context of the Essential Safety and Performance Requirements, reporting, and updating throughout the product life cycle.
I am delighted seeing in this guidance we finally get a sense for what will be considered sufficient evidence. Something we have been waiting for for ages by now, and clarifying (at least for MDSW although it reads pretty generic with the exception of one example question) that the assessment of what is sufficient data should be determined based on answers to questions regarding the amount as well as quality. Example questions proposed include:
- Does the data support the intended use, indications, target groups, clinical claims and contra-indications?
- Have the clinical risks and analytical performance/ clinical performance been investigated?
- How big is the body of scientific evidence?
- Were the type and the design of the study/ test appropriate to meet the research objectives?
- Was the statistical approach appropriate to reach a valid conclusion?
- Were all ethical, legal and regulatory considerations/ requirements taken into account?
In fact no big surprises here, since the questions simply reflect aspects that we are supposed to take into account during the general process of appraisal and analysis of clinical data.
REAL WORLD DATA
With respect to the performance analysis, it is interesting to see the emphasis that is placed on the translation of MDSW performance into daily practice: in the section on technical as well as clinical performance it is emphasized that the manufacturer should verify and test that the MDSW meets the intended purpose in real-world usage, in the intended use environment and with the intended users, telling me that Human Factor testing basically is the bare minimum that one needs to have data on.
A little misplaced seems the statement that
“CLINICAL EVALUATION of class III and implantable devices (MDR), shall include data from a CLINICAL INVESTIGATION unless the conditions of Article 61(4), (5) or (6) of the MDR have been fulfilled”.
While no-one will dispute that typically a pre-market clinical study is required for class III medical devices and implantables under the MDR unless it can be scientifically justified to rely on existing data, based on the text as it is right now, I find it puzzling whether this statement is meant to be applicable for MDSW, and, if so, for what type (stand-alone MDSW up-classified to class III and/or MDSW influencing a class III medical device?). I anticipate some clarification in this respect upon revision of this guidance.
To conclude this post, while at first sight this guidance seems a repetition of what we already know, and it indicates to be subject to revision upon the publication of other “horizontal” guidance for other devices that MDSW, it has interesting new elements worthwhile to take note of which are potentially also supportive for evaluations beyond the scope of this guidance.