Regulatory Open Forum

Recently, I had a colleague recommend design controls be included for a CFR designated 510(k) exempt device to be a part of combination device whereby the therapeutic action by a drug is the PMOA. The device is simple; let's imagine it is a graduated dosing cup for a liquid medicine for the purpose of this conversation. 

I already have my opinion on design controls given the eCFR but I am looking for other rationale and points of view.

What is the point of view/rationale of this community (i.e. yes - design controls because... or no - design control because...)?

------------------------------
Timothy Kline, Ph.D.
Senior Associate Regulatory Affairs

------------------------------

Timothy,

This is a great, but unanswered question that remains open pending FDA addressing this issues when it finalizes the Companion guidance to the Combination Product GMP rule.  I suggest that you review the comments from the many organizations that commented on that guidance at Docket ID: FDA-2015-D-0198.

There are several ways of looking at this. The most common, and simplest, is that development of the device was not subject to design controls, and this use is within the same intended use of the original device, so additional application of design controls is not necessary.  This is the industry's usual position and it makes logical sense.  It is my opinion that this, combined with appropriate suitability testing that the pharmaceutical company would perform on the finished co-packaged product, would be more than enough to ensure the combination was safe and effective. 

However, it is my opinion that FDA has taken a more conservative position that agrees that the development of the device was not subject to Design Controls, but that the Design controls must be applied to the "combination Product" to ensure that the device use as part of that combination meets the "design Input" requirements of the combination.  This usually means establishing that the device safely and effectively meets its intended use with this drug and these user groups under these specified conditions of use. As such, FDA expects that the drug manufacturer, in compliance with 21 CFR Part 4a, will have a Design control system and apply it to this product.

One way FDA may justify this is that a device, when used as part of a combination product, "loses" its device classification, as this is "outside" of its intended use.

From a "quality" not "regulatory" position, the application of design controls, which is (very simplified) the application of a formal method of determining, verifying and validating the product design, would be prudent for any company.  However, from a practical perspective, developing and implementing formal design control system in a pharmaceutical company just because they include an oral syringe or dosing cup in with their drug, is a waste of resources and does not in any way benefit the public health.  If the company has a Design Control system and the ability to approach anything in a simple manner (not easy for any pharma company), it would be prudent to document a simple design history file (DHF) for the combination.  If the company does not have a Design Control System, documenting the Suitability testing and producing a justification as to why this is equivalent to the DHF would be one course of action.  The company can still be cited for not having the system.

FDA has the opportunity to clarify this in the Companion Guidance, once finalized, but I am not hopeful.

  

------------------------------
Lee Leichter RAC
President
P/L Biomedical
Fort Myers FL
United States

Original Message:
Sent: 18-May-2016 21:34
From: Timothy Kline
Subject: Design Control - Class 1 510(k) Exempt

Recently, I had a colleague recommend design controls be included for a CFR designated 510(k) exempt device to be a part of combination device whereby the therapeutic action by a drug is the PMOA. The device is simple; let's imagine it is a graduated dosing cup for a liquid medicine for the purpose of this conversation. 

I already have my opinion on design controls given the eCFR but I am looking for other rationale and points of view.

What is the point of view/rationale of this community (i.e. yes - design controls because... or no - design control because...)?

------------------------------
Timothy Kline, Ph.D.
Senior Associate Regulatory Affairs

------------------------------

Original Message------

Recently, I had a colleague recommend design controls be included for a CFR designated 510(k) exempt device to be a part of combination device whereby the therapeutic action by a drug is the PMOA. The device is simple; let's imagine it is a graduated dosing cup for a liquid medicine for the purpose of this conversation. 

I already have my opinion on design controls given the eCFR but I am looking for other rationale and points of view.

What is the point of view/rationale of this community (i.e. yes - design controls because... or no - design control because...)?

------------------------------
Timothy Kline, Ph.D.
Senior Associate Regulatory Affairs

------------------------------

What is the situation you consider as not including design controls, as opposed to including them?

From my perspective, without design controls, if your combo product ends up with a dosing cup, it was a matter of sheer random luck.  You don't know if a dosing cup was needed, nor why the product has one.  Indeed, your product could have just as easily have a hammer or a stapler instead of a dosing cup.

Design controls are a set of actions.  What action(s) do you think of as "including" design controls?  Are you talking about the action of including design control documents in a regulatory submission?  Or actions taken to actually establish design controls?  Or actions taken to apply them to the specific device?

To clarify, to me, establishing design controls means to develop written procedures that describe your design control process.  I think this is pretty easy to do.  Applying them to your devices means to then follow those procedures for that device.  This can be simple to do or very complicated, depending on the nature of the device and its intended use.

------------------------------
Julie Omohundro, ex-RAC (US, EU), still an MBA
Principal Consultant
Class Three, LLC
Durham, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com

Original Message:
Sent: 18-May-2016 21:34
From: Timothy Kline
Subject: Design Control - Class 1 510(k) Exempt

Recently, I had a colleague recommend design controls be included for a CFR designated 510(k) exempt device to be a part of combination device whereby the therapeutic action by a drug is the PMOA. The device is simple; let's imagine it is a graduated dosing cup for a liquid medicine for the purpose of this conversation. 

I already have my opinion on design controls given the eCFR but I am looking for other rationale and points of view.

What is the point of view/rationale of this community (i.e. yes - design controls because... or no - design control because...)?

------------------------------
Timothy Kline, Ph.D.
Senior Associate Regulatory Affairs

------------------------------

The combination product final rule enhances a QMS based on the starting point. For example, a drug/device product starting with a drug QMS adds some sections from the device QMS. In particular, it adds design control in 820.30. The question is the applicability of design controls.

The requirement for design control is based on the device class, not the pre-market submission method. The requirement for design controls applies to Class 3, Class 2, Class 1 with software, and Class 1 on the list in 820.30(a). Outside of these devices, design control is optional.

The good news is that the manufacturer can conduct only those activities needed for a safe and effective product. For example, there may be value in a usability study, but no need for a formal design review. My recommendation is to pick and choose from the design control elements to select those that add value without the constraint of the whole system.

The conceptual model is that Class 1 devices are of such low risk that they don’t need the whole design control mechanism. I do suggest, however, that you review the TPLC database for MDRs and recalls, identify hazards, and ensure the device has adequate risk reduction.

------------------------------
Dan O'Leary
Swanzey NH
United States

Original Message:
Sent: 18-May-2016 21:34
From: Timothy Kline
Subject: Design Control - Class 1 510(k) Exempt

Recently, I had a colleague recommend design controls be included for a CFR designated 510(k) exempt device to be a part of combination device whereby the therapeutic action by a drug is the PMOA. The device is simple; let's imagine it is a graduated dosing cup for a liquid medicine for the purpose of this conversation. 

I already have my opinion on design controls given the eCFR but I am looking for other rationale and points of view.

What is the point of view/rationale of this community (i.e. yes - design controls because... or no - design control because...)?

------------------------------
Timothy Kline, Ph.D.
Senior Associate Regulatory Affairs

------------------------------

Dan, without design controls, what is the gate through which a changed device cannot pass without assurance that what comes out the other side will still be a Class I device?

------------------------------
Julie Omohundro, ex-RAC (US, EU), still an MBA
Principal Consultant
Class Three, LLC
Durham, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com

Original Message:
Sent: 20-May-2016 11:37
From: Dan O'Leary
Subject: Design Control - Class 1 510(k) Exempt

The combination product final rule enhances a QMS based on the starting point. For example, a drug/device product starting with a drug QMS adds some sections from the device QMS. In particular, it adds design control in 820.30. The question is the applicability of design controls.

The requirement for design control is based on the device class, not the pre-market submission method. The requirement for design controls applies to Class 3, Class 2, Class 1 with software, and Class 1 on the list in 820.30(a). Outside of these devices, design control is optional.

The good news is that the manufacturer can conduct only those activities needed for a safe and effective product. For example, there may be value in a usability study, but no need for a formal design review. My recommendation is to pick and choose from the design control elements to select those that add value without the constraint of the whole system.

The conceptual model is that Class 1 devices are of such low risk that they don’t need the whole design control mechanism. I do suggest, however, that you review the TPLC database for MDRs and recalls, identify hazards, and ensure the device has adequate risk reduction.

------------------------------
Dan O'Leary
Swanzey NH
United States

Julie, what an interesting question. I’m glad you asked.

First let me give you the negative answer. Design control in general, and design changes, 820.30(i) in particular, does not assure the changes leave the device class unchanged. In fact, there is an implicit assumption that the device class stays fixed.

In training and consulting work, I tell people that they need to augment design changes with other activities. These are not design change requirements, but usually happen at the same time.

Determine the linkage between production and process changes, 820.70(b), and design changes 820.30(i). QSIT says they are redundant with each other.

For 510(k) devices, determine if the change is significant by completing the flow charts and associated documentation in the guidance document. (Similar consideration for PMA devices, but I rarely deal with them.)

Determine if the change creates a new version or model under the DI rule or triggers a mandatory new primary DI under the GUDID guidance.

Determine if the change requires an update to the ISO 14971:2007 risk management file.

Determine whether or not the change is a performance or quality improvement under the corrections and removal rule and, consequently, its reportability

Each of these requirements are in addition to design changes, and need to be implemented independently of design controls. It is just easier to combine the activities with design changes.

------------------------------
Dan O'Leary
Swanzey NH
United States

Original Message:
Sent: 20-May-2016 12:06
From: Julie Omohundro
Subject: Design Control - Class 1 510(k) Exempt

Dan, without design controls, what is the gate through which a changed device cannot pass without assurance that what comes out the other side will still be a Class I device?

------------------------------
Julie Omohundro, ex-RAC (US, EU), still an MBA
Principal Consultant
Class Three, LLC
Durham, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com

I'm inclined to think that, when it comes to assumptions, the line between convenient and implicit can be very thinly drawn.

I agree that 820.30(I) does not inherently assure that the changes to the device leave the device class unchanged.  However, it does provide for review and approval of the change prior to implementation.  It seems to me that this allows the company to establish design review requirements of its own, and therefore creates at least the opportunity for the company to decide that the review must include a review of the potential impact of the change on regulatory classification.

I know of no other regulatory-induced control point that offers a similar opportunity to prevent a device from being placed on the market without a review of how the changes might have impacted its classification, and therefore, an opportunity to prevent a device from being placed on the market that might not meet the regulatory requirements applicable its (new) class.  Do you?

------------------------------
Julie Omohundro, ex-RAC (US, EU), still an MBA
Principal Consultant
Class Three, LLC
Durham, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com

Original Message:
Sent: 20-May-2016 14:18
From: Dan O'Leary
Subject: Design Control - Class 1 510(k) Exempt

Julie, what an interesting question. I’m glad you asked.

First let me give you the negative answer. Design control in general, and design changes, 820.30(i) in particular, does not assure the changes leave the device class unchanged. In fact, there is an implicit assumption that the device class stays fixed.

In training and consulting work, I tell people that they need to augment design changes with other activities. These are not design change requirements, but usually happen at the same time.

Determine the linkage between production and process changes, 820.70(b), and design changes 820.30(i). QSIT says they are redundant with each other.

For 510(k) devices, determine if the change is significant by completing the flow charts and associated documentation in the guidance document. (Similar consideration for PMA devices, but I rarely deal with them.)

Determine if the change creates a new version or model under the DI rule or triggers a mandatory new primary DI under the GUDID guidance.

Determine if the change requires an update to the ISO 14971:2007 risk management file.

Determine whether or not the change is a performance or quality improvement under the corrections and removal rule and, consequently, its reportability

Each of these requirements are in addition to design changes, and need to be implemented independently of design controls. It is just easier to combine the activities with design changes.

------------------------------
Dan O'Leary
Swanzey NH
United States

IMO,

Design Control methodology is arguably "Best Practice" for all design activities, whether or not it is a "regulatory" requirement or not. .

The level of detail included in the design activity should be commensurate with the level of risk the device encompasses, but the process used should consider all the steps. Meaning, don't overcomplicate the process with activities that don't add value.

------------------------------
Michael Chellson
RAC

Original Message:
Sent: 18-May-2016 21:34
From: Timothy Kline
Subject: Design Control - Class 1 510(k) Exempt

Recently, I had a colleague recommend design controls be included for a CFR designated 510(k) exempt device to be a part of combination device whereby the therapeutic action by a drug is the PMOA. The device is simple; let's imagine it is a graduated dosing cup for a liquid medicine for the purpose of this conversation. 

I already have my opinion on design controls given the eCFR but I am looking for other rationale and points of view.

What is the point of view/rationale of this community (i.e. yes - design controls because... or no - design control because...)?

------------------------------
Timothy Kline, Ph.D.
Senior Associate Regulatory Affairs

------------------------------

Yes, as far as I know, design controls have long been recognized as best practices for all product design activities, and for all types of products, not just for medical devices or other regulated products.  To my mind, they are a simply description of the product design process.

------------------------------
Julie Omohundro, ex-RAC (US, EU), still an MBA
Principal Consultant
Class Three, LLC
Durham, North Carolina, USA
919-544-3366 (T)
434-964-1614 (C)
julie@class3devices.com

Original Message:
Sent: 23-May-2016 08:29
From: Michael Chellson
Subject: Design Control - Class 1 510(k) Exempt

IMO,

Design Control methodology is arguably "Best Practice" for all design activities, whether or not it is a "regulatory" requirement or not. .

The level of detail included in the design activity should be commensurate with the level of risk the device encompasses, but the process used should consider all the steps. Meaning, don't overcomplicate the process with activities that don't add value.

------------------------------
Michael Chellson
RAC