ABSTRACT
The ultimate litmus test for any newly designed and developed product is its performance at the hands of customers. This is, of course, true for medical devices as well, but because the consequences of product malfunction or failure for these devices can be severe, performance is a more serious consideration. We have heard from design engineers that if the reliability of a medical device is verified on a bench-top, then there is no need for reliability testing under actual or simulated use conditions. We do agree that it is almost impossible to perform medical device failure testing (taking each medical device to failure) under actual use conditions due to its clinical significance. We also accept that it is extremely difficult to perform failure testing for capital equipment medical devices (CT scans, power generators, and so forth) under simulated conditions (but not stimulated conditions such as HALT) due to the long test time required. However, we emphasize to those engineers that it is necessary to perform functional testing under actual conditions (even if the device is class I or class II) before the device is released for sale, and, where possible, to perform failure testing under simulated conditions (e.g., in animal or tissue laboratories). Both functional and failure testing are part of reliability testing. When we analyze the definition of reliability and design validation, a strong overlap between the two is obvious.
