One really important test is a fatigue tests and heat test. A the fatigue test is essentially a device that will put a certain amount of cyclic loading to the device and will determine how many of those cycles it takes to fatigue. That number is documented and a safety factor is created. Another common test is a heat test which is supposed to simulate age. There's a specific science which is able to understand and extrapolate the time it takes to reach failure based on a specific heat test.
I actually have experience working as a test method development engineer for a medical device company (Becton Dickinson), working as this role your main job objectives are to develop and validate test methods that are used for design verification testing. Another part of this job role also includes identification and validation of the equipment that is needed for the DV (Design Verification) test methods. From my knowledge the FDA does not require any specific type of testing or give guidelines to how to perform any testing. There are other regulating bodies such as ISO that do require conformance to specific product performances based on the type of product you are manufacturing. For example if you are creating a medical device with a luer lock connection, ISO requires that the luer design that you are developing must meet certain assembly and removal force requirements. The high level parameters of the test methods are usually outlined by ISO but there are certain parts that are open for interpretation. It is up to the medical device company to identify the test equipment, develop the test method, and show that the product they are making conforms to all of the relevant ISO standards that are required. Other non ISO related project requirements are generally derived from the DID / DSD (Design Input Document / Design Specification Document). All of the product requirements that are outlined in either the DID or DSD need to have some type of an evaluation method so that it is possible to prove that the product meets those requirements. Not all of the requirements listed on the DID / DSD will require test methods, some may be rationalized or verified in other ways.
For Capstone II, I had to make up a test method to validate the product. I don't remember having to use a specific set of criteria for validation. When I had to do it along with my team, we thought of what could be done by ourselves to validate our design. For example, I was responsible for the mechanical properties of the product. So what I did was I suggested Instron testing. Another property I had to validate was its stability when worn, I thought to myself how to do this and I came up with a test method where I had to wear the device and do certain movements to see if the device had moved from its original location. So, I believe there is no real criteria for it because of the nature the feature/characteristic that needs to be validated.
I fully agree with this. There are various test that can be done to physically test a device. Instron testing to test mechanical properties, environmental testing to test how a product reacts in certain temperature conditions and for shelf life, packaging testing to test how a product that is packaged reacts to wear and tear as it is delivered. To expand on the statistics part, any test on a medical device should have a statistical analysis or test backing it. This is highly important since it helps support whatever test method is created and provides proof that the test is valid. A Gage R&R is a good way to prove repeatability and reproducability in a process and identifying if variation is within part-to-part or within the different inspectors or operators. In addition, doing a capability analysis is a very useful way to detect whether your system is able to meet certain specs and requirements. This is extremely using if you are testing whether a measurement for a part of a device falls within the min or max of a spec and tolerance. Both of these statistical analysis tools are powerful for demonstrating the extent of improvements made within a process or system.
I did some research on FDA website for this class and before in capstone class and the one thing I found out is that FDA doesn't care what type of tests you perform as long as you consider all possible tests and perform them. I think it makes sense because as you mentioned it, every product should have different test because of their biology, chemistry and physical requirements. For my capstone class, I was responsible for mechanical properties and did ANSYS analysis the product we invented which was made out of ABS. I never really performed test, but got the data for the product from online sources and used them for ANSYS.
It is true that the engineer is supposed to "make up a test"; however, the test needs to reflect the section of the design requirements that the test will be verifying. In the Capstone class in NJIT, when we were designing our test plan, we had to specify which requirements in our requirements document the test was aiming to satisfy. Numerous tests had to be performed in order to verify all of the requirements we put out based on the customer needs. With that being said, according to the others in this forum, the FDA does not have specific tests that have to be performed for medical devices; however, there are guidelines and standards set by the FDA and other organizations that must be satisfied in the tests crafted by the device designers.
I agree with @gaberuiz13 that the tests have to make sense. Specifically, some products already have a set of "gold-standard" validation protocols to follow, such as blood pressure cuff. There are myriads of different kinds of blood pressure cuffs in the market right now. Internationally, these BP products are validated based on protocols established by the British Hypertension Society (BHS), the Association for the Advancement of Medical Instrumentation (AAMI) and, the Working Group on Blood Pressure Monitoring of the European Society of Hypertension (ESH)... and the BHS protocol has been established as early as 1993. So one of these protocols should be a minimum process for a validation protocol if your product has other new features to test (in terms of BP cuffs).
BHS protocol: http://www.eoinobrien.org/wp-content/uploads/2008/08/x.BHS-Protocol.Revision.J-Hypertens-1993.df_.pdf
Most of the time, each medical device has a standard guideline for testing which is followed by the engineers to design the appropriate test for the appropriate device. These guidelines are known as ASTM (American Society for Testing and Materials) standards. These guidelines contain technical standards for a wide range of devices, materials, implants, or services. These guidelines contain what type of material should be used to test the device with and on what type of testing machine. This document basically lays out everything you need in order to create a Test Method to test a medical device. This testing has to be designed and performed before the V&V phase a good number of times until we can prove that this test produces consistent results. This is where accuracy and precision comes into play when perfecting a test method. We have to make sure the test method is accurately testing what we need it to test. In regards to precision, we have to design this test method to show that it outputs consistent and precise results. Once we have accurate and precise results over a number of tests of the test method itself, then we can move onto V&V.
The FDA does give some guides to how the test should be run but the general rule of thumb is to have a test for every claim that your product proposes will happen to the human using the medical device. This way, there are no unfounded claims in any device to deceive the public. In every aspect of biomedical devices, there are a series of tests used by other testers which are published in journals. The best way to test a device is to know that the tests found in publications will be repeated again on your device, so using those tests for the clinical trials is a safe way to know that you're checking all the boxes without explicit guidelines for tests.
Tests used for design verification must be able to demonstrate that a particular characteristic of the medical device is conforming to the specification. As mentioned in other responses, the FDA does not specifically tell you which tests need to be done. However, the FDA does expect you to be complete in your V&V and will request that more be done if they do not believe that sufficient testing has been done.
It is not always necessary to come up with a brand new test for your device. ISO Guidelines exist for many categories of devices and outlines that type of testing that you should be able to complete. There also exist many ASTM (American Society for Testing and Materials) methods which are already defined and validated. If a standardized test method does not exist, a test method will need to be created and a test method validation will need to be performed. A test method validation demonstrates that a test method is suitable for its intended use and can distinguish between conforming and non-conforming samples.
Since a lot of ppl touched on the FDA portion of the question, I will try to help you understand the other portion of your question. In theory, there are unlimited ways to test if the product is safe by doing any possible test you can think about. Therefore, I believe it is important for the designer to think about the most important tasks. Each task will vary by product hence you should be able to be flexible. It is important to breakdown the task you are planning to perform into multiple big categories. From there you should be able to gauge on more specific tasks. At the end of the day, you are doing multiple tasks to check that your product is able to perform constantly for a specific amount of time or counts with producing specific errors which you set per the environment the device is exposed to.
Another way to choose which test to perform is to look at previous studies' design specifications and question how they obtained those specifications. For my senior capstone project, we did a lot of research into the design specifications listed for scanning acoustic microscopes by various vendors and would see commonly reported specs, which allowed us to wisely choose the correct tests. For example, in the case of microscopes, resolution is an important specification, which can be tested via phantoms, to see the smallest possible unit of detection. Even though you could perform a chemical test on a microscope, it is probably unnecessary and many vendors do not report any chemical/biological tests for microscopes, but rather mechanical strength or spatial resolution are of more importance. Therefore, to choose which test to perform, one could reverse-engineer the process by questioning what specifications they would want to end up with, and then choosing the appropriate tests for those specifications.
Verification and validation testing are performed to ensure that the project results align with the correct user needs and specifications. Rather than looking at it from the perspective of satisfying the FDA, validation and verification testing should be viewed from an internal perspective in which you are checking project success. Verification testing is performed throughout various stages of product development and are crucial in determining whether each phase of the project is heading in the correct direction. On the other hand, validation deals more with confirming that the overall design of the product meets the actual needs of the user. To clarify, verification is used to confirm design specifications throughout the project development life cycle, whereas validation is used to confirm user needs closer to the end of the project development life cycle.
@vnd4 If verification test usually stay in house, do you think there is a specific department or procedures that ensure that this verification methods are suffice? I'm aware that inputs and outputs must equal but how can we confirm that the results are correct. Or is this when clinical trials come into play?
