Setting Expected Service Life for an an Electrical/Electronic Medical Device

D

dwayneq

I am trying to determine methods to help set an electrical/electronic medical device's expected service life. In reviewing other similar devices the standard appears to be 5 years which is what I have for our device. However, I need to be able to back this up with good rationale and testing for submission to my notified body today and and in the futher. The required testing is well outlined in IEC 60601-1 for requirements up through the expected service life as well as our own performance requirements. What is missing is how to set this for a new product based on industry based rationale or upfront testing. Of course real-time evaluations will happen during the product's lifetime but you can't wait 5 years to set an expected service life.

Any assistance with this is greatly appreciated.
 

Marcelo

Inactive Registered Visitor
Re: Help in Setting Expected Service Life

Using reliability engineering techniques such as accelerated life testing?
 

Mikishots

Trusted Information Resource
I am trying to determine methods to help set an electrical/electronic medical device's expected service life. In reviewing other similar devices the standard appears to be 5 years which is what I have for our device. However, I need to be able to back this up with good rationale and testing for submission to my notified body today and and in the futher. The required testing is well outlined in IEC 60601-1 for requirements up through the expected service life as well as our own performance requirements. What is missing is how to set this for a new product based on industry based rationale or upfront testing. Of course real-time evaluations will happen during the product's lifetime but you can't wait 5 years to set an expected service life.

Any assistance with this is greatly appreciated.

In agreement with Marcelo. HALT analysis will help give you numbers on infant mortality, average expected functional life (with random failures), the point of expiration (wear-out failures).

Bathtub curve.
 

Peter Selvey

Leader
Super Moderator
It is a common mistake that the service lifetime is something derived from the equipment's properties.

Rather, it is simply derived from your sales and marketing, i.e. a lifetime that is competitive with other manufacturers (typically 5 years).

Once the lifetime is specified, it can then be used to influence other design decisions, and in particular safety related decisions. For example, ensuring that the failure rate of a critical protection system is sufficiently low over the 5 years, or if not adding a check point during regular maintenance.

Note also that it is OK for equipment to fail within the lifetime (although if this happens a lot, don't expect good sales). However, the device should remain "safe" (free of unaccepble risk) over the defined lifetime.
 
D

dwayneq

Peter, you are hitting on my suspicions that sometimes people are looking at this incorrectly in saying that you are required to derive your expected service life through all sorts of prior testing. My history had always lead me to believe as you say that you need to look at similar equipment and their stated service life and then also look at your most likely failure mechanisms or shortest component stated lifetime and set your device’s expected service life. Then you need to follow that up with verification that you will be able to meet that life by performing HALT testing, etc.
My plan on this product is to put together a rationale based on what I stated to set the expected service life along with the verification plan for testing. Many thanks.
 

Peter Selvey

Leader
Super Moderator
Just to make it clear: this is not my opinion but it's actually what is in the standard. The definition of "expected service life" is simply the "maximum period of useful life as defined by the MANUFACTURER".

The problem is that Clause 4.4 says "The MANUFACTURER shall state the EXPECTED SERVICE LIFE ... in the RISK MANAGEMENT FILE".

Many people look at this and conclude that the service life must be an output of risk management. However if the clause is read carefully this is not the requirement. Rather, the service life is an input to risk management.

In particular, Clause 3.117 states that "SINGLE FAULT SAFE" means:
characteristic of ME EQUIPMENT or its parts whereby it remains free of unacceptable RISK during its EXPECTED SERVICE LIFE under SINGLE FAULT CONDITIONS"

Since the service life is part of the definition of safety, it has to be defined before any decisions can be made.​
 

Peter Selvey

Leader
Super Moderator
This is not opinion, it is directly from the standard. The standard does not say "The service life shall be defined in the risk management file and based on an assessment of component reliability". It just says service life shall be defined.

And, reliability assessments are notoriously complicated and unreliable (no pun intended). If required almost certainly it would be a fudge job with serious risks if the derived service life was used for critical safety decisions.

In contrast, for critical safety systems, knowing the service life as an input is a useful specification, in the same way that you might define the usable temperature range (10-40°C) as an input to engineering decisions on component selection, error calculations and so on.
 

Marcelo

Inactive Registered Visitor
On the other hand, the standard does require that the equipment maintains basic safety and essential performance thru the expected service life, so taking into consideration only sales, marketing, or things like that may be problematic.

ISO 14969 does list a few things that you may need to consider when defining device lifetime:

The basis of the defined lifetime of the medical device should be documented. To assist in determining the lifetime of the medical device, the rationale for the determination should be recorded and may involve consideration of the following:
a) shelf life of the medical device;
b) expiry date for medical devices or components which are subject to degradation over time;
c) number of cycles or periods of use of the medical device, based on life testing of the medical device;
d) anticipated material degradation;
e) stability of packaging material;
f) for implantable devices, the residual risk that results from the entire period of residence of the device inside the patient’s body;
g) for sterile medical devices, the ability to maintain sterility;
h) organization’s ability/willingness or contractual or regulatory obligation to support service;
i) spare parts cost and availability;
j) legal considerations including liability.
 

Peter Selvey

Leader
Super Moderator
Hi Marcelo (long time! - I just found the site is active again)

It is a good reference but highlights the classic active / non-active divide and how this can sometimes mess up the thinking.

Non active devices (non electrical) tend to be more simple and often the materials used are well known with lots of reliability data available. So, for example deriving a shelf life from the packing material is easy (well, easier).

Active (electrical) devices are extremely complicated with thousands of components, no reliable reliability data exists, and testing would be impossible (you would need thousands of samples to get even a rough idea).

Either way, I would still insist that service life is something which from a legal or regulatory perspective must be defined not derived. There is no doubt the lawyers would throw any attempt to put the ISO 14969 text into law or an EU standard.

This is a case where the guide is simply going off what sounds like a good idea for a niche area (e.g. packaging for sterility) and then assuming it can work for all fields.
 
Top Bottom