During my second simulation report, the purpose of the simulation was to find a solution and develop a test plan to verify and validate the efficacy of the solution. I expected researching and coming up with a solution to be the hardest part of the process. However, coming up with the test plan was the hardest part it is essential to develop a testing plan that can prove the efficacy of the product as fast as possible and with the lowest cost possible. Even though every project needs its customized test plan, What are some common strategists to develop testing plans?

It is very important to understand the product ecosystem. The tests have to start to be designed as product specs are being developed. The tests have to be treated like a subproject. While defining the acceptance criteria ensuring there is no ambiguity is important too. It is great to have test engineers in charge of testing, and with time test engineers would develop the experience that is required to foresee how everything would go. It is also good practice to have responses planned for tests in case there will be failed tests. Some tests might require high sample lead times or high test run times and planning accordingly is important. 

During the second simulation, I found it helpful to first brainstorm all the things we thought could be the problem. From there, we were able to evaluate which were more probable as well as which would be the easiest and quickest to test first. Then, while the first round of testing was occurring and we were waiting on results we were able to start making tentative for the next round to have a head start once the results came in. By prioritizing which components were most likely the problem and then creating test plans in a timely manner, we were able to efficiently run tests to solve the problem. It was also useful to plan independent tests that could be run in parallel. This made it possible to test multiple components while being more efficient by running them at the same time. 

For the second simulation my group found it helpful to develop a list of the potential issues, but not develop a new test plan. This is because we need to first make sure we completed the existing test of just overnight storage. Instead we thought of issues and then created a procedure and validation criteria that would point out what the issue was with the initial design. Overall, my go to strategy for making testing plans is to try and make them as simple as possible, while also giving detailed instructions in the procedure so that repeatability is possible.

When I like to tackle a challenge, I find it very helpful to alter one variable at a time. In this simulation, by changing the packaging and medicine at one time, a concrete solution could not be determined if this were to fail. What exactly caused the failure? Think of this approach as introducing new foods to a child. At the moment, you don’t know what they’re allergic to. But giving them peanuts, milk, or wheat in one day, makes it very difficult to determine what they should not eat if an outbreak occurs. After you have determined one variable is not the issue, continue introducing more, and eventually the one that is causing the problem will be discovered.

My team's approach to the simulation was to first list all the possible roots of the problem. Then, we would list them in decreasing order of how probable they would to occur. After that, we would conduct some research on available tests that would allow us to test our theory. As others have mentioned, it is important to change one variable at a time to better target the problem. In some cases, there might be one factor causing the problem, and in others there might be more than one. While you're waiting for the response, it is good to plan the next step. Look into solution that might help your chosen variable, and at the same time look into other factors that might've caused the problem too.

@nr473 I agree, nice explanation and mapping out the process. 

From working in a testing group, one common strategy I see for developing fast and low cost testing is by taking advantage of compendial test methods. Compendial test methods are test methods that have been used in the past and are accepted as a "standard". ASTM has a wide variety of useful test methods ( ASTM TEST METHODS » SGS-IPS Testing ). In the medical device industry, you won't always find a compendial method that overlaps with exactly what you are trying to test. Despite that, these methods can still be very useful. If there is a method that is similar, you can use is as a basis for developing a medical device specific method leveraging as much as you can from the other methods. This can save time and money.

Has anyone had any experiences working with compendial test methods?

Thanks,

Matt

I agree Naglaa, the hardest part of these simulations thus far has been setting up the test procedure. I'd say this comes with both our lack of experience setting up these kinds of tests and our ignorance with the root of the failure of these scenarios. I think depending on the scenario you want to tackle the situation differently. For example, in the second sim, I tried to take that approach of researching the different materials that could be substituted for what was used because I thought the flaw came from the components used in the system as well as the actual syringe. However with sim 3, I tried taking the same approach, and based on the professor's feedback I should've also considered the different toxicology lab results, which is something I didn't even think to tackle on the fly.