Regulatory Open Forum

 View Only
  • 1.  Design Verification and Validation (DV&V)

    This message was posted by a user wishing to remain anonymous
    Posted 26-Aug-2021 09:46
    This message was posted by a user wishing to remain anonymous

    Hi team

    I will try to explain this complicated subject as best I can. This is for an IVD medical device. 

    We have recently identified numerous deficiencies in our DV&V stages, in that there are plenty of design inputs that do not have DV&V completed. After many internal discussions, we've come up with the following approach: 

    1. Since our devices have been on the market around the world for over 10 years, post market surveillance data to indicate if there are failures for any of the design inputs. 
    2. Scientific literature, guidelines etc. search, if any exist. Otherwise opinions from internal subject matter experts. 
    3. Client acceptance / approval documents.
    4. IQ/OQ/PQ as design verification.
    Since this is a retrospective completion of the gaps, it is really hard to come up with the most efficient way to fulfil the requirement and achieve compliance. We know for a fact that our device is fit for purpose, safe and achieves its intended use. 

    I am really looking forward to your thoughts, and if you need more clarity then let me know.


  • 2.  RE: Design Verification and Validation (DV&V)

    Posted 26-Aug-2021 10:09

    You are making it too complicated. Just do the design verification and the design validation as you should have initially. The other things you list are not relevant.

    Document the protocols and reports in the design history file. Each should have an indication that you performed them after release to production and are part of a correction.

    These all fall under corrections of QMS nonconformances.

    Also, open a corrective action to determine the cause of the nonconformance and eliminate it.



    ------------------------------
    Dan O'Leary CQA, CQE
    Swanzey NH
    United States
    ------------------------------



  • 3.  RE: Design Verification and Validation (DV&V)

    This message was posted by a user wishing to remain anonymous
    Posted 27-Aug-2021 08:05
    This message was posted by a user wishing to remain anonymous

    Hi Dan

    Thank you very much. I understand it is complicated.

    How would you retrospectively do DV&V on say things like IFUs, confirmation of orders, labelling, production processes used to manufacture your device, packaging etc.


  • 4.  RE: Design Verification and Validation (DV&V)

    This message was posted by a user wishing to remain anonymous
    Posted 27-Aug-2021 08:05
    This message was posted by a user wishing to remain anonymous

    Also, is it correct to say that you are not validating every single Design input but the final medical device, that is it meets its intended use and is safe?


  • 5.  RE: Design Verification and Validation (DV&V)

    Posted 27-Aug-2021 12:23

    I'm going to address each item in your list individually using FDA QSR. ISO 13485:2016 may use different words, but has the same concepts.

    Design verification confirms that the design output fulfills each design input.

    Design validation confirms that the device meets patient and user needs.

    Design transfer converts design output to production specification specifications. Production specifications may also involve process validation. I think of production settings as information in the production operator's work instruction on the equipment "knob settings".

    Lastly, don't think of this as "retrospective", but rather as completing missed process steps. You are not collecting information that is only available in a post-market setting to close gaps.

    IFUs are a design output based on the design input. The design input should list things that need to be in the IFU. For example, if your ISO 14971:2019 risk management system identified as risk reduction measure as information for safety then the process should have transferred it to the design inputs. Design verification, confirms that the IFU contains the required information. A common method uses a checklist to match design input to design output. The design verification protocol includes the blank checklist. The design verification report, is the completed checklist.

    Confirmation of orders is not part of design control, and is not subject to design verification, design validation, or design transfer. It is important, but it is a different process.

    Labeling is similar to the IFU. However, for labels the design output should include both location and content. Typically, there is a drawing for the label content and a drawing for the label location. Design verification ensures both are correct. Note that a label may have multiple locations.

    Production processes used to manufacture the device are part of design transfer. It follows design verification, since you need to have correct design output to convert to production specifications. Design transfer can include process validation, since it provides the information for the operator to set-up the process and monitor the process inputs.

    Packaging is a design output, so design verification applies. Also, packaging is a production process so design transfer applies.

    Typically design validation starts with design inputs. Identify the design inputs that involve patient or user needs and conduct design validation on them. You will also conduct design verification on them.

    In summary, just conduct the missed steps. Post-market data won't help, because these are all pre-market activities. Use the current version of the documents. There is no value in performing, say design verification, on an obsolete drawing.



    ------------------------------
    Dan O'Leary CQA, CQE
    Swanzey NH
    United States
    ------------------------------



  • 6.  RE: Design Verification and Validation (DV&V)

    Posted 27-Aug-2021 08:45

    While Dan alluded to it, but did not specifically mention there is no such thing as "DV&V". Design Verification is a separate and distinct activity from Design Validation. They have very different requirements and work with different aspects of the design with different focuses and different requirements. 


    Design Validation is very different from Process Validation, which your statement seems to confuse. I recommend you refer to the "ISO Practical Guide-ISO 13485:2016 Medical devices" for a good discussion of the differences in the three topics. 


    I might mention that successful Design Verification is an input to both Process Validation and Design Validation, so it should be completed first. Design Validation can use product that has been created during the PQ phase of Process Validation. 


    Since it appears you are doing this from a retrospective perspective you must provide a rationale for why each of the three activities you are performing after the fact provides sufficient basis to keep existing distributed product on the market. You should also provide a rationale for product already on the market, if any discrepancies in the product are found during the three activities. 



    ------------------------------
    Edwin Bills MEd, CQA, RAC, BSc, CQE, ASQ
    Principal Consultant
    Overland Park KS
    United States
    elb@edwinbillsconsultant.comPrincipal Consultant
    ------------------------------



  • 7.  RE: Design Verification and Validation (DV&V)

    Posted 27-Aug-2021 10:59

    When discussion verification and validation I recommend using the appropriate adjective. For example, design verification, design validation, process validation, software validation, etc. This avoids a lot of confusion.

    In this post, which I answer in more detail in another response, the topics are design verification and design validation only.

    Bill is correct that DV&V doesn't exist, because it is really two different things. We see this "abuse of language" in many places. The common is CAPA. I have people tell me they will open a CAPA. My first question is whether they intended to conduct corrective action or preventive action. They are two different processes.



    ------------------------------
    Dan O'Leary CQA, CQE
    Swanzey NH
    United States
    ------------------------------



  • 8.  RE: Design Verification and Validation (DV&V)

    Posted 27-Aug-2021 15:55
    Since this is stated to be an IVD medical device, the relevant regulatory markets and classification of the device are probably important to consider. If this is the case of a device that would be submitted for self-certification under IVDD, but is now being considered for marketing under IVDR, you will have a significantly different regulatory burden. Likewise, if this is a laboratory developed test under CLIA in the US, design controls are not required; however, if it is a medical device under FDA (enforcement discretion aside), that should be a factor too.

    The advice given already to complete the necessary verifications and validations is correct.

    With respect to the approach to you lined out, you may consider analyzing the available PMS data during remediation to determine if your product is potentially able to meet your design requirements. Likewise, searching scientific literature and/or guidance should reflect back on the design inputs and design outputs, as well as ensuring that the design meets the current state of the art (special consideration should be made with respect to your company's established policy for establishing risk acceptability). Complaint data/customer satisfaction data should likewise inform your design inputs and user needs. Lastly, IQ/OQ/PQ may not be useful information to consider unless manufacturing is not complete until the device is assembled on-site.  It's worth noting that none of the proposed approaches can demonstrate (to any reasonable certainty) that design outputs meet each design input, nor that representative samples meet user needs and intended uses.

    As mentioned with respect to CAPA, it would seem that this requires corrective action and should probably be accompanied by some level of containment and consideration as to whether or not field action is necessary as part of the correction. The very notion that this device is "fit for purpose, safe and achieves its intended use" is at the very least unsupported if not incorrect at this point.

    ------------------------------
    Christopher Erwin
    Scottsdale AZ
    United States
    ------------------------------