Labelling can be use as risk control..?

Generally in the field of medical devices, you can identify IFS as a risk control measure but consensus is that you cannot claim a reduction in risk (*1) by using IFS controls. I 100% believe you can do a risk control option analysis and determine that (for some lines of risk analysis) that (those) lines of risk are acceptable by using (sometimes only) IFS.

It used to be the case (somewhat recently, but I stopped tracking this specific edge case) that ALARM SIGNALS (particularly those implemented via software) were occasionally referenced as IFS, and that for those specific implementations... the implication was to verify the implemented ALARM SIGNAL and claim a reduction in risk. I never liked this wording, I believe ALARM SYSTEMS are a protective measure of the design... considering elements of an ALARM SYSTEM to be IFS muddies the waters IMO.

(*1) My argument against claiming an actual reduction in risk via IFS is this: It is easy to demonstrate that IFS has been implemented, but it is incredibly difficult to demonstrate (with any confidence/tolerance) that the IFS is effective.

I would be very interested in knowing where EU has stated that risks cannot be reduced through IFS control measures. EN TR 24971 seems to state just the opposite in H.3.4, where examples of IFS which can be used to control risks are reported...

MrTetris said:

I would be very interested in knowing where EU has stated that risks cannot be reduced through IFS control measures. EN TR 24971 seems to state just the opposite in H.3.4, where examples of IFS which can be used to control risks are reported...

Click to expand...

This was directly called out in Annex ZA of the harmonized version of ISO 14971:2012. That ZA annex had a lot of details that reviewers were really picky about (this was also when ALARP was a bad word...)

I believe that specific requirement and the content deviations in the ZA annex are no longer part of the 2019 version. However, I suspect that since the position was that IFS cannot reduce risk for so long, you will have an uphill battle with any reviewers who still have this mentality. It is practically difficult to demonstrate meaningful risk reduction through IFS, so even if you accept that you can reduce risk I don't know that your verification testing will support it.

EmiliaBedelia said:

It is practically difficult to demonstrate meaningful risk reduction through IFS, so even if you accept that you can reduce risk I don't know that your verification testing will support it.

Click to expand...

I agree strongly. My advice would be that the lines of risk analysis don't show a reduction in risk, but that the Risk Control Option Analysis should conclude that the IFS made the risks acceptable (for documented reasons).

A (hypothetical) example: Perhaps there is a medical device with a risk profile that is unacceptable if exposed to a certain type of cleaning solution. Further suppose it is, for technological reasons, impossible to make the device out of a material that survives that cleaning solution. A warning (IFS) against using that cleaning regimen is appropriate... and depending on the user class I would be very comfortable identifying the IFS as a risk control, not change the ratings, and then declaring "in the context of use for this device, the labeling is expected and appropriate to make this risk acceptable."

My interpretation of this is that informing a user of a risk (i.e. labeling that says there are risks associated with use of a medical device) can not be assigned risk reduction value. However, this is not that same as saying labeling cannot be assigned risk reduction value.

The best illustration of this is for a patient to pick from the following two choices; the surgeon who as read the IFU or the surgeon who has not? So its quite obvious that instructions on safe and proper use can be assigned value as a risk reduction. However, having the IFU inform the surgeon that there is a risk of the patient getting an infection by using the product should be assigned no value in reducing the risk of the patient getting an infection since informing what the potential hazards or harms are doesn't help avoid them.

d_addams said:

My interpretation of this is that informing a user of a risk (i.e. labeling that says there are risks associated with use of a medical device) can not be assigned risk reduction value. However, this is not that same as saying labeling cannot be assigned risk reduction value.

The best illustration of this is for a patient to pick from the following two choices; the surgeon who as read the IFU or the surgeon who has not? So its quite obvious that instructions on safe and proper use can be assigned value as a risk reduction. However, having the IFU inform the surgeon that there is a risk of the patient getting an infection by using the product should be assigned no value in reducing the risk of the patient getting an infection since informing what the potential hazards or harms are doesn't help avoid them.

Click to expand...

Thank you for your contribution d_addams. I get your point, but even when instructions are given to the surgeon through IFU, there cannot be any control on the surgeon reading and understanding the IFU, so that is why Auditors are often reluctant with accpeting ifu as a risk control measure.

MrTetris said:

Thank you for your contribution d_addams. I get your point, but even when instructions are given to the surgeon through IFU, there cannot be any control on the surgeon reading and understanding the IFU, so that is why Auditors are often reluctant with accpeting ifu as a risk control measure.

Click to expand...

the user not reading the training is a foreseeable condition and should be included in your risk assessment, but that does not mean the instructions for proper use cannot be assigned risk reduction value. If failure to read or abide by the IFU presents an unacceptable risk, mandatory training prior to distribution to physicians can be used as well.

Also the point about 'control of the physician understanding the IFU' is incorrect. Human Factors assessments will provide evidence the IFU is understandable and usable by the physicians or patients if necessary.