Biological Evaluation (10993) & Risk Management

ThatSinc

Involved In Discussions
#1
Hi All,

I'm looking for a more efficient way of documenting the "biological evaluation / material selection / process agent selection" process and how it fits into the 14971 framework.

I'm finding that there is a lot of double handling, and what I would consider "evaluation for the sake of evaluation".
For example, I'm making a disinfection container, with fill line, specifically to be used with a range of devices and disinfecting solutions.
The devices are then surface contacting devices.

currently there are multiple lines in the hazard analysis for toxicity, allergenicity, irritants, etc. for the materials used in the container, processing agents used in the manufacture of the container, incoming materials not meeting specification, leachable substances from materials etc.

I'm finding the evaluation of user exposure to various aspects, toxic materials as an example, somewhat redundant, when it is known* that the product is to be made from HDPE. The "pre-controls" evaluation doesn't particularly add any value, it seems to just serve to show that you have reduced risk from some arbitrary level to the validated level you have following controls.

*whilst I appreciate that your material selection should be dependent on the requirements and therefore risks relating to the product, I don't evaluate the risks relating to the hazard of radioactive materials, I know I'm not going to make this product out of radioactive materials.

I find it doesn't add anything within the hazard analysis to say "toxic substances" - "toxic substances in container leach into solution and onto device" or "toxic substances from manufacturing processes remain on device" and arbitrarily score a P1 and P2, just to say "device is manufactured from HDPE" or "processing agents do not contain toxic substances" as the control measure, show that the risk reduction is validated through the biological evaluation testing and report.

In keeping with both the 10993 process and the 14971 process, the risk file "device characterisation" collates details regarding contact type & contact time.
would it be appropriate to then characterise the material choice within the 10993 documentation and only assess within the 14971 documentation the risks that actually exist based on the material characterisation performed within the 10993 records?

e.g. if the materials selected have evidence that they do not pose a risk, there is no risk to evaluate?

Am I missing anything in this approach?

Thanks,

TS.
 
Elsmar Forum Sponsor

Tidge

Trusted Information Resource
#2
I may have missed the question,,,

I'm looking for a more efficient way of documenting the "biological evaluation / material selection / process agent selection" process and how it fits into the 14971 framework.

I'm finding that there is a lot of double handling, and what I would consider "evaluation for the sake of evaluation".
For example, I'm making a disinfection container, with fill line, specifically to be used with a range of devices and disinfecting solutions.
The devices are then surface contacting devices.
[...]

e.g. if the materials selected have evidence that they do not pose a risk, there is no risk to evaluate?
I don't consider such lines of analysis to be 'double handling"; its more of a blunt acknowledgement of the existence of the risk and a (perhaps understandable) blunt application of a well-understood risk control that is straightforward to implement (specify) and verify (material certificates!). While it is unlikely that HDPE would all of a sudden be discovered to be an ineffective risk controls for the applications which you chose to use it... it is not impossible that state of knowledge could change. The existence of an industry accepted risk control doesn't mean that there was no risk to begin with.
 

ThatSinc

Involved In Discussions
#3
I don't consider such lines of analysis to be 'double handling"; its more of a blunt acknowledgement of the existence of the risk
Do you acknowledge in every risk assessment the risk of radiation, and control it by not using radioactive substances to manufacture the device?
Or do you only assess the risk of radiation when you have a device that by necessity requires a material source of radiation to achieve its intended use?

I would consider the material characterisation as a part of 10993 to cover material selection and identify any pertinent hazards associated with any particular materials that you may need to use, in order to achieve an intended use.

How do you assign a P1 and P2 value (if that's the way you're assigning probability) of a user being exposed to a toxic substance, and that toxic substance causing harm. Logically, you'd have to identify every possible substance you could make the product from, and then determine what portion of those are toxic.
This brings into question the value in assigning probabilities and severities of harm to risks prior to risk control, particularly when you consider the MDD/MDR requires you to reduce all identified risks, not just those that you do not deem acceptable.
It seems the "pre-control" assessment and "post-control" assessment is just there to show that hopefully you've reduced risk from some level that may or may not have been acceptable, to a level that you have defined as acceptable.

The question was ; would you consider it appropriate to *not* assess those lines in a hazard analysis, and leave those to the material characterisation section of your 10993 biological evaluation report.

I think your response suggests, no, you would not consider it appropriate to do that - and they should always form a part of the hazard analysis.

The existence of an industry accepted risk control doesn't mean that there was no risk to begin with.
At what point do you stop?

When you assess the risk of a user not being able to use a device properly, and so you control this by including an insert in the packaging - do you state that you've introduced a new risk of paper cuts from the insert, and control it by ensuring that the paper is flimsy enough that it won't cut you?

And then, with the contentious requirement from the MDR to disclose *all* residual risk, do you need to state on the insert that it may cause paper cuts?
 

Tidge

Trusted Information Resource
#4
Do you acknowledge in every risk assessment the risk of radiation, and control it by not using radioactive substances to manufacture the device?
Or do you only assess the risk of radiation when you have a device that by necessity requires a material source of radiation to achieve its intended use?
I use a predicate questionnaire that asks about all sorts of hazards that don't often appear in the medical devices. If radioactive materials aren't used, then risks from radiation hazards aren't in the HA. I suppose if the design team felt the need to add one, we'd revise the original assessment and put those risks back into the analysis.

If the device is going to contact patients users, there are going to be risks associated with bio-compatibility... even if those risks are low without the implementation of risk controls. There are plenty of patients with acute or chronic injuries because somebody didn't take bio-compatibility even a little bit seriously. As far as hills go, this is not a good one to choose to die on... especially when industry accepted solutions for biocompatibility, cleanability, etc., make it very easy to turn it into a molehill.

If you are truly worried about paper cuts introducing unacceptable, you can always IBD (get rid of the paper). Or you could include the risk and assess that it is acceptable and impossible to reduce. I'm not sure there is as much history with sever paper cuts as there is for bio-compatibility risks, so "state of the art" has probably got all of us with paper IFUs covered.
 

planB

Super Moderator
#5
TS,

ISO 10993-1:2018 comprehensively covers all risk management aspects required for performing a biological evaluation. Figure 1 in this standard gives you a high-level overview how to perform this biological risk management process. A word of caution - quote from section 4: "The biological evaluation shall be planned, carried out, and documented by knowledgeable and experienced professionals."

HTH,
 

ThatSinc

Involved In Discussions
#6
If the device is going to contact patients users, there are going to be risks associated with bio-compatibility... even if those risks are low without the implementation of risk controls.
Once again, you are right and it's not a hill I plan on dying on.
As you say, the control measures for these risks are robust.
It's not difficult to add those lines into the assessment, excepting the fact we're being asked to provide supporting evidence for the P1 and P2 values prior to control measures being implemented, and have just listed the probability of both P1 and P2 at 100%.


TS,
A word of caution - quote from section 4: "The biological evaluation shall be planned, carried out, and documented by knowledgeable and experienced professionals."
Fortunately the actual material aspects and biocompatibility assessment of materials, including all processing parameters and processing agents are being managed by materials scientists and clinical specialists in far more detail than I could ever understand. The documentation of the evaluation and testing is comprehensive within the various 10993 reports and overall 10993 summary.
 

Tidge

Trusted Information Resource
#7
It's not difficult to add those lines into the assessment, excepting the fact we're being asked to provide supporting evidence for the P1 and P2 values prior to control measures being implemented, and have just listed the probability of both P1 and P2 at 100%.
I hesitate to give specific advice about assigning pre-control ratings, either in a HA or something more focused like an FMEA. I agree that simply dropping in '100%' (or 0%, depending on which is 'worst case') can serve as reasonable starting point. I have a short rant to explain my thinking, and I hope it will explain why I think your choice is fine.

The reason why I hesitate to give such advice is because I feel that it is extremely important to be consistent in the meaning behind the ratings (both pre- and post-controls) across the type of document (HA v. FMEA). I don't think I have ever worked with a team within a company, let alone an entire organization, that consistently applied the same meaning to ratings across the risk files. All groups who consider 'ratings' within risk files struggle with this (including external reviewers), most (in my experience) simply shrug with a wink and a nod in the direction of 'it's all qualitative'.

Frankly: I feel like the move away from using 'Detection' ratings in Design FMEA is entirely because people were unwilling and/or unable to ascribe a useful meaning to 'ratings'. My own attitude and approach doesn't usually align with other peoples' thinking , but I can at least apply my own way of thinking about qualitative rankings in a consistent manner. For me, I tend to approach the qualitative ratings as a combination of both expected value and degree-of-belief in that expected value. I am always perfectly comfortable assigning 'pre-control' ratings in a design FMEA for detectability, because certain design solutions are 'well known' to act as effective controls for certain types of risks even in the absence of objective evidence that those risks for the specific design are mitigated. Subsequent testing/certification can improve the risk rating with verification of effectiveness.

Your decision to apply the pre-control "100%" in this case is completely understandable to me, as I can think of this as essentially choosing to claim that you have no prior knowledge about how a risk control could possibly work for this case. In Bayesian probability theory this is akin to an 'uninformed prior'. You will accumulate your risk controls and make a new (posterior) assessment of risks that is improved.

I've never met or read about anyone else who thinks about risk assessments this way. I try not to discuss this in person because folks can get really defensive about their rating systems. Ultimately the value of any rating system is in the transformation of the (knowledge about) the state of risk prior to implemented controls into the (knowledge of) the state of risk after the implementation of controls, and not about the ratings themselves.

One last example where I almost never see logical consistency in ratings. Many groups will include a 'roughly quantitative' guideline about ratings (typically in FMEA) that involve 'powers of 10'. For example: an ordinal rating of 4 is 'ten times better' than 5 and 'ten times worse' than 3. It is common that a 'risk analysis' (again, typically in an FMEA) includes a 'risk control' that changes a rating by one or more of these 'powers of 10'... yet it is extremely uncommon that the 'verification of effectiveness' for such a control includes actual data (let alone a study design) that merits a 'power of 10' improvement in the rating. (quotation marks included because I evangelize about how FMEA are not tools for 'risk' analysis relating to patient/user safety)
 
Last edited:

indubioush

Quite Involved in Discussions
#8
Do you acknowledge in every risk assessment the risk of radiation, and control it by not using radioactive substances to manufacture the device?
No. Early on, when you identify characteristics of the device that are related to safety, you will consider radiation and document that the device does not emit radiation. No further action required. However, when you identify characteristics of the device that are related to safety and consider toxic materials, you will document that there is a potential risk of non-biocompatible materials used in the device or packaging. This becomes a line item in the risk analysis, and the control measure is your design input that requires the device and packaging to use only materials that are biocompatible. You don't have to have a crazy number of lines in your analysis for this, but it should be there. The further details regarding material selection can be captured in the biological evaluation.

How do you assign a P1 and P2 value (if that's the way you're assigning probability) of a user being exposed to a toxic substance, and that toxic substance causing harm. Logically, you'd have to identify every possible substance you could make the product from, and then determine what portion of those are toxic.
No, you don't have to do this. If the probability of harm cannot be estimated, you follow the method outlined in your risk management plan. Typically you just assess the risk based on the severity alone and assign the highest possible probability level.
 

ThatSinc

Involved In Discussions
#9
No. Early on, when you identify characteristics of the device that are related to safety, you will consider radiation and document that the device does not emit radiation... you will document that there is a potential risk of non-biocompatible materials used in the device or packaging.
But how do you know the device will not emit radiation if you don't evaluate the risk of it emitting radiation, and therefore choose to not put radioactive materials in it... Just because you don't intend it to emit radiation, doesn't mean you don't love that glowing green colour and think that would be great for the packaging to make it really *pop* on the shelf...

I don't *intend* for my device to emit contain toxic substances, yet I still have to evaluate for them, and document that evaluation.

Yes, the above is intended in jest, and relates to my "where do you stop" comment regarding evaluation of risk.
As above, it's not a hill I intent to die upon, and all of my RMFs so far have always included the evaluation of these risks, and will continue to do so based on this discussion thread.

Typically you just assess the risk based on the severity alone
How do you assess the severity of a harm without knowing the particular toxic substance that you're exposing a user to...
There are many toxic substances that could very happily kill you - so do you give it your highest severity scoring too, because you might have put them in there without assessing it?

Once you've assessed the risk, and put your controls in place to make the product out of 316L stainless, how do you re-assess your post-control risk - reducing the probability and/or severity?

It's rare in a risk file that I'll reduce the severity of a harm through risk controls, typically because I don't reassess what the harm will be following controls, and a defined harm will have a defined severity.

So whilst on a practical level regarding the product the "risk" is negligible regardless of whether I have listed the pre-control severity as causing death, it seems strange to consider there being a residual risk of death relating to toxic substances and have a risk profile for a benign device that includes the risk of death.
Unless it's considered acceptable to state that the controls have eliminated the risk entirely and therefore there is no residual risk, however I've never worked with a system where probability of hazardous situation occurring can be scored as 0 and then discounted as residual risk.


I have a short rant to explain my thinking, and I hope it will explain why I think your choice is fine.
This reply made me smile, as I am in complete agreement with everything in it and have had various discussions around risk scoring lately and the use of FMEA.

Working with Process FMEAs that will consider a failure mode as having the potential to lead to harm and then rating that failure mode with a severity of the harm that it could pose, despite a whole host of other failures needing to also occur for harm to happen.
And then assessing risk acceptability from PMS data on reports of that failure mode and severity of harm that could happen, despite no harm occurring, as other risk control measures were happily in place protecting the user.
 

Tidge

Trusted Information Resource
#10
But how do you know the device will not emit radiation if you don't evaluate the risk of it emitting radiation, and therefore choose to not put radioactive materials in it... Just because you don't intend it to emit radiation, doesn't mean you don't love that glowing green colour and think that would be great for the packaging to make it really *pop* on the shelf...
I think there is a subtle point to be made, specifically about this quirky example. You absolutely could do a 'biological evaluation' (or some other sort of test) and surprisingly discover that you have used an emitter as part of your materials and proceed with risk analysis if you intend to keep that as part of your design. The subtle difference that @indubioush is making is that his design team planned to use radioactive materials as an energy source (or possible a planned treatment). A parallel example is "steam", either as a power source (steam-punk medicine anyone?) or as a treatment (e.g. an autoclave, less exciting than steam-powered infusion pumps but nevertheless a real thing). If the fundamental design doesn't include steam, there is no point in adding lines to a risk analysis to study possible controls to make those risks acceptable.

Working with Process FMEAs that will consider a failure mode as having the potential to lead to harm and then rating that failure mode with a severity of the harm that it could pose, despite a whole host of other failures needing to also occur for harm to happen.
And then assessing risk acceptability from PMS data on reports of that failure mode and severity of harm that could happen, despite no harm occurring, as other risk control measures were happily in place protecting the user.
I have seen what you describe. This is one of the areas where I encourage cross-referencing between Design FMEA and Production Process FMEA. Some folks like to let them both trickle up to a line in the Hazard Analysis to let the HA sort things out... and I have seen that the HAs typically don't make a distinction between potential contributors to risk (once past the initial use case and P1, P2 assessment) and controls for risks (from that initial assessment).

More rambling follows: I believe I have advocated (on this forum) about how an individual risk control option analysis at the HA line level could help clarify things where both D and P FMEA trickle up into the HA... and I have vocally made my complaint that FMEA are not tools for analyzing risks... yet: I still encounter auditors from NBs that INSIST that if you have documents subordinate to a Hazard Analysis, that each one of those documents ALSO includes "individual risk control option analyses". In the case of a process adding a risk control to avoid a potential risk to patient/user, it is very rare (but not impossible) that the production process has nothing to do with the safety or effectiveness of a device, so performing an RCOA at the FMEA level is a meaningless exercise. Due the RCOA at the level where the risks are actually analyzed, and not where failure modes are described.
 
Thread starter Similar threads Forum Replies Date
J Biological Evaluation requirements per ISO 10993 - Medical Device Testing ISO 13485:2016 - Medical Device Quality Management Systems 2
W ISO 10993 (Biological Evaluation of Medical Devices) for Suppliers Other ISO and International Standards and European Regulations 8
B Biological evaluation plan and report Other Medical Device Related Standards 7
M Informational Work in progress at the FDA for biological evaluation – Color Hazard and RISk calculator (CHRIS) Medical Device and FDA Regulations and Standards News 0
Gamula CFDA Biological Evaluation Report China Medical Device Regulations 2
S IV Catheter Biological Evaluation - EtO Residuals - Permanent Contact Definition Qualification and Validation (including 21 CFR Part 11) 2
T Biological equivalency - Dermal filler EU Medical Device Regulations 4
K Biological indicator system classification under MDR EU Medical Device Regulations 2
L Biological Assessment (ISO 10993-1) Other Medical Device Related Standards 1
A Demonstration of Equivalence - Need for comparing biological characteristics for an SamD EU Medical Device Regulations 1
M Informational US FDA paper – Biological Responses to Metal Implants Medical Device and FDA Regulations and Standards News 0
M Informational MHRA guidance updated to reflect the mDR – Clinical investigations of medical devices – Biological safety assessment Medical Device and FDA Regulations and Standards News 1
D Biological and Chemical Risks to the user in the Hazard Analysis ISO 14971 - Medical Device Risk Management 1
B Self Contained Biological Indicators as internal PCD Other Medical Device Related Standards 2
M Biological Monitoring Program Environmental Bioburden Test Method Needed ISO 13485:2016 - Medical Device Quality Management Systems 1
B Number of Biological Indicators for Routine Monitoring of ETO Sterilization Cycle Other Medical Device Regulations World-Wide 8
J ISO 10993 (Biological Compatibility) Approved Materials US Food and Drug Administration (FDA) 23
S Requirements of USP24 NF18 1995 for the Biological Test for Plastics Other Medical Device and Orthopedic Related Topics 3
M Biological Indicator Ethylene Oxide Sterilization Other Medical Device and Orthopedic Related Topics 11
D MDD 2007/ - Essential Requirement Checklist - Physical and Biological Properties EU Medical Device Regulations 4
I Glass Packaging Material (Jars) - Chemical, Physical and Biological Controls Manufacturing and Related Processes 1
T How to treat something (biological hazard) you know is there... Food Safety - ISO 22000, HACCP (21 CFR 120) 7
J Effects of substrate on biological indicator organisms US Food and Drug Administration (FDA) 2
S Clinical Evaluation of equivalent device EU Medical Device Regulations 4
L Clinical evaluation plan EU Medical Device Regulations 10
C Biologic Evaluation based on ISO 10993-1 EU Medical Device Regulations 2
M Selling non-CE marked devices for evaluation EU Medical Device Regulations 4
M Procedure for clinical evaluation according to new MDR EU Medical Device Regulations 1
Q Process map Evaluation and Analysis Method Process Maps, Process Mapping and Turtle Diagrams 5
M Supplier evaluation Supplier Quality Assurance and other Supplier Issues 5
I QMS monitoring, measurement, analysis and evaluation requirement - Template ISO 9000, ISO 9001, and ISO 9004 Quality Management Systems Standards 9
shimonv Clinical evaluation report for class I device EU Medical Device Regulations 3
A Applicability of Photobiological Safety Evaluation for LED used in medical devices Reliability Analysis - Predictions, Testing and Standards 2
P GSPRs / Clinical Evaluation EU Medical Device Regulations 3
silentmonkey Overall Benefit/Risk Analysis - Risk Management VS Clinical Evaluation ISO 14971 - Medical Device Risk Management 3
M NICE Medical Technology Evaluation Programme - Recommendations Service Industry Specific Topics 0
Q Summative Usability Evaluation Testing: prior or during Clinical Investigation? Human Factors and Ergonomics in Engineering 9
S Regular updates of clinical evaluation report EU Medical Device Regulations 6
K Surface finish (Evaluation Length) Inspection, Prints (Drawings), Testing, Sampling and Related Topics 3
M Clinical Evaluation Benchmark vs. Equivalent EU Medical Device Regulations 2
M IVDR and Performance Evaluation Plan CE Marking (Conformité Européene) / CB Scheme 2
S User evaluation for self monitoring blood glucose test systems US Food and Drug Administration (FDA) 4
S Australia TGA Clinical Evaluation Report (CER) Other Medical Device Regulations World-Wide 0
S API Spec. Q1 clause 5.6.1.2 On site evaluation Oil and Gas Industry Standards and Regulations 10
F Biocompatibility evaluation for Hardware/Interface Components? Other Medical Device Related Standards 12
B Clinical Evaluation Expert Panels - MDR EU Medical Device Regulations 1
T Clinical evaluation of a new medical device EU Medical Device Regulations 0
dgrainger Informational MDCG 2020:13 - Clinical evaluation assessment report template EU Medical Device Regulations 0
S A clinical performance evaluation study with an IVD product as Investagional Use product - Clinical Monitor requirements 21 CFR Part 820 - US FDA Quality System Regulations (QSR) 4
B AS9100 8.4.1 Supplier Selection/Evaluation criteria and reevaluations AS9100, IAQG, NADCAP and Aerospace related Standards and Requirements 13

Similar threads

Top Bottom