Posted by Rob Packard on September 23, 2014
This article reviews FDA QSIT inspection requirements of design validation and is specific to devices that do not contain software.
In the FDA QSIT Manual (http://bit.ly/QSITManual), the word “validation” appears in the QSR 78 times. This exceeds the frequency of the name “verification,” “production,” “corrective” and the acronym “CAPA.” The word “validation” is almost as frequent as the word “management”–which appears 80 times in the QSIT Manual. The section of the QSIT Manual specific to design validation is pages 35-40.
The FDA selects only one product or product family when they are inspecting design controls. Therefore, if you keep track of which products have already been inspected by the agency, you can often predict the most likely product for the investigator to select during the next inspection. The number of MDRs and recalls reported will have an impact on the investigator’s selection. Class, I devices are not selected.
The QSIT Manual instructs inspectors to verify that acceptance criteria were specified before conducting design validation activities and that the validation meets the user needs and intended uses. There should also be no remaining discrepancies from the design validation. Inspectors must verify that all validation activities were performed using initial production devices or production equivalents. The last item to verify is that design changes were controlled–including performing design validation of the changes.
Risk analysis is seldom reviewed in great detail–except software risk analysis. However, when a nonconforming product is reworked, it is required to review the adverse effects of rework. QSIT inspectors will expect you to document this review of risks. Investigators will also expect risks to be reviewed and updated in accordance with trend analysis of complaints, service reports, and nonconformities. Finally, when companies assess the need to report recalls, the FDA expects to see a health hazard evaluation to be completed (http://bit.ly/HHE-Form). A detailed review of risk analysis is uncommon in QSIT inspections but receives greater emphasis in the review of CE marking applications.
Predetermined Acceptance Criteria
Investigators reviewing your design validation protocols will specifically look at the acceptance criteria for testing you perform. Investigators are looking for two things. First, were the acceptance criteria met without deviation? Second, was the protocol approved before knowing the results (i.e., was this a prospective design validation protocol?). In certain areas, there are also known risks associated with products that the investigators will look for. For example, in sterilization validation, the investigator will verify that the validation was performed to the most current version of the standard and that the validation has addressed the most common pitfalls of sterilization. For example:
- Have the most challenging devices been identified?
- Has performance been validated at the maximum sterilization dose?
User Needs & Intended Uses are Met
In the area of user needs and intended uses, there are few problems with the initial launch of devices for the intended use. Problems typically arise when companies expand the intended use to new patient populations and new intended uses. When this occurs, there may be unique user needs and risks that need to be evaluated. Therefore, the FDA periodically reviews claims made by companies in marketing communications to ensure that claims do not stray beyond the cleared intended use of the device. This will sometimes be identified as a 483 inspection observation. In some instances, the FDA will issue a warning letter to a company that continues to market a device for uncleared indications.
Initial Production Devices or Production Equivalents
When investigators review validation protocols and reports, the documentation must include traceability to the production lot(s) of the device. Investigators may even request a copy of the Device History Record (DHR) for the production lot used for validation. If a production lot is not used, then the design validation documentation must disclose how the product differs from production lots, and why the results are acceptable. The samples used should be subjected to the final test/inspection requirements. If final test/inspection requirements are not yet established, samples should be retained, so that they can be inspected at a later date. Without this traceability, you may have to repeat your design validation with a production lot.
Validation of Design Changes
Far too many hours are wasted writing justifications for why re-validation is not required. I recommend that re-validation of design be performed for any design change if all three of the following criterion are not met:
- a sound scientific rationale can be provided with references
- the logic does not require a subject matter expert to understand it
- quantitative analysis is possible to analyze the risk impact
Many design validations require simulated use with a physician. Companies should obtain as much user feedback as possible before launching a device. Therefore, any re-validation that requires simulated use and user feedback should be a priority over writing a rationale for not conducting re-validation.