Regulatory Open Forum

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

I would consider it a hazardous situation that can lead to different harms depending on the type of treatment.  For example, prolonged exposure to x-rays can be a harm caused by delayed treatment. I rarely see harms resulting from "delayed treatment" documented in risk management documents.  This can impact other documentation such as the IFU and clinical evaluation.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Hello Jasmin - thank you for sharing your approach. In a recent warning letter to an IVD manufacturer, FDA highlights this point - when an assay is used for therapeutic drug monitoring, delay in results can lead to harm(s). In this case, severity assignment was not done correctly, leading to underestimation of risk. 

I think there is a need to stary paying more attention to harm(s) resulting from delay in treatment. I am concerned to read your comment that you rarely see these harms properly documented in risk management documents.

Here is the link to the warning letter referenced above.

Best regards

I would consider it a hazardous situation that can lead to different harms depending on the type of treatment.  For example, prolonged exposure to x-rays can be a harm caused by delayed treatment. I rarely see harms resulting from "delayed treatment" documented in risk management documents.  This can impact other documentation such as the IFU and clinical evaluation.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Hi

Risk management may have been around for years but is a real weak spot for IVD manufacturers and gives rise to NB questions under the IVDR. Many do not consider harm to the patient at all but stop at false positive or negative and do not consider what impact this could have on the patient. The guidance in ISO TR 24971 describing how to use P1 + P2 to describe the likelihood of a false result or a delay in treatment causing harm to the patient is very useful and being a RM geek a good read.

Hello Jasmin - thank you for sharing your approach. In a recent warning letter to an IVD manufacturer, FDA highlights this point - when an assay is used for therapeutic drug monitoring, delay in results can lead to harm(s). In this case, severity assignment was not done correctly, leading to underestimation of risk. 

I think there is a need to stary paying more attention to harm(s) resulting from delay in treatment. I am concerned to read your comment that you rarely see these harms properly documented in risk management documents.

Here is the link to the warning letter referenced above.

Best regards

I would consider it a hazardous situation that can lead to different harms depending on the type of treatment.  For example, prolonged exposure to x-rays can be a harm caused by delayed treatment. I rarely see harms resulting from "delayed treatment" documented in risk management documents.  This can impact other documentation such as the IFU and clinical evaluation.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Hello Naveen,

As someone who works in clinical operations I think this really depends on the device. In many cases even if there is a device fault the physicians can use a different device or alternative method so this may not necessarily result in treatment delay or even if you have a delay it may not cause harm. Of course, there maybe a cases where the delay could cause a harm so I guess it really depends on the device type.

Have a great week,

Olga 

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Hello Olga - you are right; there are many situations where delay in treatment may not matter. I think we can figure this out by building different sequence(s) of events following a device related delay. By doing this, we can consider only those situations where harm can occur. Would you agree with that approach? 

Hello Naveen,

As someone who works in clinical operations I think this really depends on the device. In many cases even if there is a device fault the physicians can use a different device or alternative method so this may not necessarily result in treatment delay or even if you have a delay it may not cause harm. Of course, there maybe a cases where the delay could cause a harm so I guess it really depends on the device type.

Have a great week,

Olga 

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Annexes A, F and H of ISO/TR 24971:2020 generally and specifically categorize delay in treatment as a hazardous situation.

Hello Olga - you are right; there are many situations where delay in treatment may not matter. I think we can figure this out by building different sequence(s) of events following a device related delay. By doing this, we can consider only those situations where harm can occur. Would you agree with that approach? 

Hello Naveen,

As someone who works in clinical operations I think this really depends on the device. In many cases even if there is a device fault the physicians can use a different device or alternative method so this may not necessarily result in treatment delay or even if you have a delay it may not cause harm. Of course, there maybe a cases where the delay could cause a harm so I guess it really depends on the device type.

Have a great week,

Olga 

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Hello Kevin - thanks for your guidance.

I am curious to learn from your experience about why there is so much inconsistency in applying this guidance when it comes to analyzing risks associated with delay in treatment/therapy.

In my LinkedIn poll for example, 38% of the over 200 respondents considered this a harm, and another 13% a hazard. 

Is it because there is still a lot of confusion about terminology (hazard, hazardous situation, harm etc.)? At least that has been my experience so far, but I would love to hear your perspective. 

Annexes A, F and H of ISO/TR 24971:2020 generally and specifically categorize delay in treatment as a hazardous situation.

Hello Olga - you are right; there are many situations where delay in treatment may not matter. I think we can figure this out by building different sequence(s) of events following a device related delay. By doing this, we can consider only those situations where harm can occur. Would you agree with that approach? 

Hello Naveen,

As someone who works in clinical operations I think this really depends on the device. In many cases even if there is a device fault the physicians can use a different device or alternative method so this may not necessarily result in treatment delay or even if you have a delay it may not cause harm. Of course, there maybe a cases where the delay could cause a harm so I guess it really depends on the device type.

Have a great week,

Olga 

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Hey Naveen.  Honestly, I'm not sure of the cause for the various conflicting interpretations.  Each divergent case could have its own respective root cause(s) and would require proper investigation to flush out.  But I can say that proper medical device 14971 risk management is an intrinsically difficult undertaking that requires sufficient expertise and careful attention to 14971's many critical parameters.  Design/development teams often seem to be under-staffed or under-trained to keep real-time pace and synchrony with 14971's various moving parts.  Thus it can be helpful to have at least one person, even if a consultant, who knows the way and can help a team stay on target.

Hello Kevin - thanks for your guidance.

I am curious to learn from your experience about why there is so much inconsistency in applying this guidance when it comes to analyzing risks associated with delay in treatment/therapy.

In my LinkedIn poll for example, 38% of the over 200 respondents considered this a harm, and another 13% a hazard. 

Is it because there is still a lot of confusion about terminology (hazard, hazardous situation, harm etc.)? At least that has been my experience so far, but I would love to hear your perspective. 

Annexes A, F and H of ISO/TR 24971:2020 generally and specifically categorize delay in treatment as a hazardous situation.

Hello Olga - you are right; there are many situations where delay in treatment may not matter. I think we can figure this out by building different sequence(s) of events following a device related delay. By doing this, we can consider only those situations where harm can occur. Would you agree with that approach? 

Hello Naveen,

As someone who works in clinical operations I think this really depends on the device. In many cases even if there is a device fault the physicians can use a different device or alternative method so this may not necessarily result in treatment delay or even if you have a delay it may not cause harm. Of course, there maybe a cases where the delay could cause a harm so I guess it really depends on the device type.

Have a great week,

Olga 

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Delay of treatment is a hazardous situation that could be associated with various hazards. One problem I find with ISO 14971:2019 is that it doesn't provide enough emphasis on the sequence of events which converts the hazard to the hazardous situation. In addition, effective risk control measures block at least one step in the sequence of events. Better risk control measures block multiple steps.

The Swiss Cheese Model, developed by James Reason provides the analysis. Blocking one step may still lead to patient or user harm. Blocking multiple steps, often termed "defense in depth" is the more effective choice.

Your question was "How do you characterize 'Delay in treatment' due to a device malfunction in your risk analysis?"

A device malfunction means, in the language of the standard, that the device is in a fault condition. Therefore, I would expect the sequence of events to include something like, "The device is in a fault condition", "The device has malfunctioned", "The required device functionality is not available", etc. (Note that this includes issues of reliability.)

This sequence of events leads to a specific hazardous situation. However, one must use caution here because the hazardous situation implies, after exposure, patient or user harm. Some companies include patient irritation as a patient harm.

Consider, for example, a routine dental appointment for a cleaning and a periodic x-ray. The hygienist tells me that the dental x-ray is down and they will take the x-ray at my next routine visit. I would not characterize this as patient harm.

In a more severe case, a patient is undergoing an operation, some problem occurs, the patient heart stops, and requires an internal defibrillator. However the defibrillator won't charge and the patient dies.

The step in the sequence of events could be the same in both cases, "device malfunction" but the harm and its severity would be very different.

Consequently, the risk reduction measures would be different.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

There are some good responses in this thread.

I would characterise 'delay in treatment' as a hazardous situation. Like with pretty much all other hazardous situations, context is important (which links to how well you define/characterise the events leading up to the hazardous situation).

I would also see it, like other hazardous situations, as a point where multiple risks cross-over. What I mean by that is that there will usually be more than one set of circumstances/events/causal factors that create that hazardous situation, and as has been pointed out in these responses, there will often be more than one harm arising from the hazardous situation. And with a slight adjustment of the sequence of events (e.g. a last minute detection of a fault condition by the user, use on a specific patient population), the hazardous situation and the subsequent harm can change. The possible harms (and their severity) can help us understand criticality and priority, and to help determine the appropriate level of effectiveness of control (and digging - see last line), but our control measures will target the events leading up to the hazardous situation (and then it becomes a decision on how well you understand and characterise the events leading up to the hazardous situation). The deeper you dig, the more understanding you should have of what is happening and what to control.

I've seen 'delay in treatment' included as above, but have also seen it cited as a harm (which to me is an indicator to look for other framing issues within the risk assessment).

As always, thanks @Naveen Agarwal for initiating useful discussions for everyone to air and share their views.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

Thank you Ed for your insightful comment.

This exercise of clearly articulating a hazardous situation proves to be challenging in practice. I have been thinking about it for many years, and I will be the first to admit that the concept of a hazardous situation as outlined in ISO 14971 is extremely confusing! I am not surprised that most practitioners struggle with this. There are so many scenarios that one could think of following an initial event!

It is important to not be stuck in a perpetual "what if", rather only consider events and scenarios that are probable. This is easier said than done, because it requires a considerable amount of experience. 

What best practices have you seen that help practitioners to articulate hazardous situations clearly without getting lost?

Thanks again for contributing to this discussion.

There are some good responses in this thread.

I would characterise 'delay in treatment' as a hazardous situation. Like with pretty much all other hazardous situations, context is important (which links to how well you define/characterise the events leading up to the hazardous situation).

I would also see it, like other hazardous situations, as a point where multiple risks cross-over. What I mean by that is that there will usually be more than one set of circumstances/events/causal factors that create that hazardous situation, and as has been pointed out in these responses, there will often be more than one harm arising from the hazardous situation. And with a slight adjustment of the sequence of events (e.g. a last minute detection of a fault condition by the user, use on a specific patient population), the hazardous situation and the subsequent harm can change. The possible harms (and their severity) can help us understand criticality and priority, and to help determine the appropriate level of effectiveness of control (and digging - see last line), but our control measures will target the events leading up to the hazardous situation (and then it becomes a decision on how well you understand and characterise the events leading up to the hazardous situation). The deeper you dig, the more understanding you should have of what is happening and what to control.

I've seen 'delay in treatment' included as above, but have also seen it cited as a harm (which to me is an indicator to look for other framing issues within the risk assessment).

As always, thanks @Naveen Agarwal for initiating useful discussions for everyone to air and share their views.

👋Dear colleagues, hello!

📊Recently I posted a poll question on LinkedIn that generated a lot of good discussion.

------------------------------
Naveen Agarwal, Ph.D.
Problem Solver | Knowledge Sharer.
Let's Talk Risk!
@https://naveenagarwalphd.substack.com/
------------------------------