Testing of component from Third party

Do we need to test a component for 100 million cycles even though the manufacturer has specified that it will function when we qualify it for a medical device? If the auditor asks if we have tested component for 100 million cycle in our system, should we say that the supplier evaluation is sufficient? Let's take the proportional valve as an example. The datasheet for the proportional valve states that the valve has a life of 100 million cycles, and it was purchased from a reputable manufacturer from Parker.

There is not enough information provided to answer your question. What is your medical device's intended use? What is the risk classification? What is your product design input requirement that is verified through the 100 million cycle test? What kind of testing is this?

Our medical device intended use is to provide noninvasive respiratory therapy. and class is IIb, if I'm selecting a component for my device , and the datasheet of the component says that it works for 100 million cycles can i put the component in the device directly by trusting the datasheet, or do i have to do a test for 100 million cycles in our device and then qualify the component? and if we just show the datasheet to the auditor and tell him that we have selected the component based on the datasheet will he accept? or can we justify him telling we have done supplier qualification? or do we have to provide a proof of the component working in our system for 100 million cycles

Minimally: I would reach out to the manufacturer of the component and ask to get some clarification of what their 100M cycle number represents. It will be important to know (for examples) if 100M represents some median value (and then find out what the corresponding uncertainty of the distribution is), or if 100M represents the point at which 95% of the vales will fail.

Depending on the engineering experience of the team making the medical device, a conversation with the manufacturer might also reveal what a "failure" of a 100M cycle valve component actually looks like. Is there an increase in power draw as the valves approach failure? Is the failure related to leaks? Is it breakage of mechanical elements like gears? The nature of the end-of-life for a component is (in my mind) somewhat important, because it might not take much testing (by the medical device designer) to establish their own confidence limits around what the 100M cycle number represents.

Goutham prasad said:

...if I'm selecting a component for my device , and the datasheet of the component says that it works for 100 million cycles can i put the component in the device directly by trusting the datasheet, or do i have to do a test for 100 million cycles in our device and then qualify the component? and if we just show the datasheet to the auditor and tell him that we have selected the component based on the datasheet will he accept? or can we justify him telling we have done supplier qualification? or do we have to provide a proof of the component working in our system for 100 million cycles

Click to expand...

As written above: I lean towards generating your own evidence, but I don't think you ought to be repeating design verification of the individual component's claims. I feel like it is better that the design team understand how their (finished) device works when assembled from discrete elements. Would a third party cause grief relative to a specific component? Of course it is possible... but different third parties would get to that point via different paths: risk management, customer complaints, claims, essential performance, basic safety, etc.