Attachment E - Recommendations on FDA draft Real-World Evidence guidance

SACHRP Recommendation Approved May 26, 2017
 

 

The HHS Secretary’s Advisory Committee on Human Research Protections (SACHRP) has considered carefully, and in this document provides public comment on, the FDA draft guidance “Use of Real-World Evidence to Support Regulatory Decision-Making for Medical Devices.”   SACHRP addressed this and similar issues in its March 30, 2012 “Recommendation on Applicability of FDA Regulations,” online at http://www.hhs.gov/ohrp/sachrp-committee/recommendations/2012-march-30-letter-attachment-a/.  

SACHRP commends FDA for providing guidance on this important issue, particularly as it pertains to the IRB’s role in overseeing FDA-regulated clinical investigations involving devices.  IRBs have a regulatory mandate under 21 CFR Part 812 to make a determination of the regulatory status of investigational devices in clinical investigations.  The IRB can find that such clinical investigations are Significant Risk (SR) and require an IDE, or are Non-Significant Risk (NSR) and fall under the abbreviated IDE requirements, or are exempt from the IDE requirements under one or more of several criteria in 21 CFR 812.2(c) and associated guidance.  In addition, there are other possible regulatory categories such as Humanitarian Use Devices (HUD) and devices, such as general wellness devices, that fall under FDA enforcement discretion.  Determining the correct regulatory status of a device in an FDA regulated clinical investigation is one of the most difficult issues that IRBs face, due to the complexity of the regulations and of the device approval process, and the lack of definition of key regulatory phrases such as “used or investigated in accordance with the indications in labeling” in 812.2(c)(2).  

Therefore, SACHRP supports, in general, FDA’s intent to clarify “when an Investigational Device Exemption (IDE) may be needed to prospectively collect and use RWD [Real World Data] for purposes of determining the safety and effectiveness of a device.”  (Lines 117 to 119.)  However, and more importantly, SACHRP urges FDA to use this opportunity to clarify the broader question of when the research use of Real World Data is a clinical investigation at all.  SACHRP encourages FDA to provide clear, objective criteria that IRBs and other regulated parties can use to determine when such research is a clinical investigation subject to FDA regulations, and if it is, when it is SR and therefore requires an IDE, when it is NSR and can proceed under the abbreviated IDE requirements, or when it is exempt from the IDE requirements.  This would be of significant benefit to IRBs, sponsors, institutions and investigators.  In short, SACHRP is supportive of an approach that would allow, to the greatest extent and with least transaction costs, industry and others to submit to FDA, and FDA to consider as part of its regulatory processes, RWD and RWD analyses that are taken from records of care administered as treatment and in the course of which all treatment decisions were made only with an intent to assist the patient and not with any intent of conducting medical research.  In this document, using the regulatory terms and principles that apply to research under FDA jurisdiction, SACHRP suggests ways in which FDA might implement that larger goal.  
 

This commentary is divided into two parts.  The first part focuses on the process of determining when a given research study, particularly involving the use of Real World Data (RWD), is an FDA-regulated clinical investigation.  If it is a clinical investigation, we then focus on the process of determining when an FDA regulated clinical investigation is SR and requires an IDE, is NSR and can proceed under an abbreviated IDE, or is a clinical investigation that is exempt from the IDE requirements.  On this determination, we note that in general SACHRP believes that clinical investigations involving the use of RWD should be NSR under an abbreviated IDE or exempt as they generally do not involve the assignment to the use of an investigational device for the purpose of studying that device.  The second part provides commentary on the examples provided in the draft guidance of when an IDE was required for various RWD projects.
 

SACHRP notes that one of the reasons this issue has been important is that the current FDA regulations only allow for a waiver of consent in certain emergent situations, unlike the Common Rule which allows for waivers of consent for minimal risk research for which it is impracticable to obtain consent.  Thus, many study designs that would qualify for a waiver of consent under the Common Rule do not qualify for such a waiver under the current FDA regulations.  This has resulted in a significant barrier to the use of RWD for research purposes when such studies are considered a clinical investigation under the FDA regulations.  However, the 21st Century Cures Act was recently passed and allows FDA to adopt a waiver of consent for certain minimal risk research.  Although we cannot now know the details of how this waiver will be enacted in regulation, when it is enacted it is likely to help make the research use of RWE in clinical investigations more feasible.  SACHRP notes that it has consistently supported the FDA adoption of an FDA waiver of consent in minimal risk studies, in both its March 30, 2012 “Recommendation on Applicability of FDA Regulations, and its October 26, 2016 “Recommendation on Regulatory Issues in Cluster Randomized Trials.”  In those recommendations SACHRP suggested that FDA adopt the waiver of consent that exists in the Common Rule at 45 CFR 46.116(d).  This would serve the goal of creating harmonization between the FDA regulations and the Common Rule.  However, SACHRP also notes that the Common Rule waiver of consent was in part originally drafted for the use of deception in behavioral research, and that this might be an appropriate opportunity for FDA to create a waiver of consent more directly applicable to the use of RWD to support research that provides evidence of the safety and efficacy of devices.  SACHRP will be pleased to provide further input on this issue at FDA’s request. 

 

 

SACHRP has previously addressed the difficult issue of determining when research is a clinical investigation in its March 30, 2012 “Recommendation on Applicability of FDA Regulations,” online at http://www.hhs.gov/ohrp/sachrp-committee/recommendations/2012-march-30-letter-attachment-a/.  In that recommendation, SACHRP first addressed the triage process IRBs apply when reviewing proposed research, the first step of which is to determine whether the activity is research under the Common Rule, and the second step of which is to determine whether the activity is a clinical investigation under the FDA regulations 21 CFR Parts 50 and 56.  
 

The first step follows this triage process:
• Is the activity research?

• Is the activity research involving human subjects?

• Is the institution engaged in research?

• Is the activity research that is exempt under 45 CFR 46.101(b)(1) through (6)?

• Is the activity research that can be reviewed through expedited procedures?

• Is the activity research that requires review by a convened IRB?

The second step follows this triage process:

• Is the activity a clinical investigation?

• Is the activity a clinical investigation involving human subjects?

• Is the activity exempt under 21 CFR 56.104(a) through (d)?

• Is the activity a clinical investigation that can be reviewed through expedited procedures?

• Is the activity a clinical investigation that requires review by a convened IRB?

 

Research studies involving human subjects may fall under the Common Rule, FDA regulations, or both.  If the IRB fails to interpret properly the relevant regulations and apply them correctly, it opens the possibility of observations and enforcement findings against the IRB, sponsors, institutions and investigators.  SACRHP strongly has encouraged FDA to provide clear objective criteria for the regulated community to use in performing this triage process.

 

Among the problems of the regulated community is that there are multiple definitions of “clinical investigation.”  The prior SACHRP recommendation discussed the difficulty in interpreting the four definitions in 21 CFR Part 50, 56, 312, and 812, and offered recommendations on providing clarification of the terms to provide objective criteria.

 

Finally, the prior SACHRP recommendation addressed various study designs and their status as clinical investigations, including among others retrospective chart reviews, drug and device registries, and interviews and questionnaires.  The use of Real World Data overlaps significantly with some of these designs, particularly registries, and therefore many of the prior recommendations apply.

 

The growing availability of big data, which can provide useful information about the safety and efficacy of FDA regulated products, adds additional complexity to the regulatory determinations that IRBs must make, particularly as they are used in clinical settings after device approval.  SACRHP commented on the regulatory issues that arise with big data in its recommendation “Human Subjects Research Implications of ‘Big Data’ Studies.” https://www.hhs.gov/ohrp/sachrp-committee/recommendations/2015-april-24-attachment-a/index.html

 

In this recommendation, SACHRP reiterates many of its prior recommendations in the context of the FDA draft guidance “Use of Real-World Evidence to Support Regulatory Decision-Making for Medical Devices,” and also makes new recommendations reflecting the issues discussed in the draft guidance.

 

SACHRP encourages FDA to provide clear criteria for use to determine when a given research project using RWD is a clinical investigation, and if it is a clinical investigation involving a device, whether it is SR requiring an IDE, is NSR and can be conducted under an abbreviated IDE, or is exempt from the IDE requirements.  These categories are very well laid out and explained in relation to traditional clinical trials in the FDA guidance entitled “The IDE Process,” available online at http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/HowtoMarketYourDevice/InvestigationalDeviceExemptionIDE/ucm046164.htm.  Similar detailed guidance would be extremely useful to the regulated community regarding the criteria to be used in determining if the use of RWE and RWD is a clinical investigation that is SR, NSR, or IDE-exempt.

 

Section IV(B) of the draft guidance addresses the application of Investigational Device Exemption (IDE) requirements in 21 CFR 812 to the collection of real world data.  In this section, FDA describes three factors that can inform the decision of whether an IDE is necessary:

 

Whether the collection of RWD could be subject to the IDE regulations depends in part on whether that collection constitutes a clinical investigation. Several factors can inform this determination, including the purpose for which the data is being gathered, whether the process for gathering the data would influence treatment decisions, and whether the rights, safety and welfare of human subjects are impacted, among other things.  (Lines 325 to 329.)
 

SACHRP recommends that FDA provide further explanation of these three criteria, and/or additional criteria, so that IRBs and other regulated entities can use them to make correct and independent decisions of whether a study constitutes an FDA regulated clinical investigation, and if so, the appropriate regulatory classification of the device study as SR, NSR, or IDE-exempt.  The remainder of this section addresses potential criteria for determining when research is a clinical investigation, and if so the device status of research projects, starting with the three criteria cited above from lines 325 to 329 of the draft guidance.
 

The purpose for which the data are being gathered
 

Data may be collected for any number of purposes.  In many cases, the original intent of the data collection may not be for research purposes, such as data collected during the provision of routine clinical care, but subsequent use of the data may be for research purposes.  Within research there is a wide range of uses, such as student-led studies conducted to fulfil the requirements of an advanced degree, scholarly activities by academic faculty for publication in peer reviewed journals, or industry-sponsored evaluations of the safety and efficacy of a device with the express intent of submitting to the FDA for the purpose of obtaining approvals, clearances, or supporting marketing claims.

In the case of gathering real world data or evidence, the utility and benefit of FDA regulatory oversight varies considerably depending upon not only the purpose of the activity but the circumstances under which the data will be or have been collected.  For instance, if the data are being collected primarily to establish safety and efficacy of a device or to submit to the FDA in support of a marketing claim, then the activity, as currently construed under FDA interpretations, might properly be considered an FDA-regulated clinical investigation and requires a determination as to whether it is an SR study that requires an IDE, an NSR study that can be conducted under the abbreviated IDE requirements, or an exempt study.  In contrast, if the data are collected as part of routine clinical care and the decision whether or not to use the device is not altered by the research activity, in the absence of intent to submit to the FDA to support a marketing claim, the research may not be a clinical investigation and therefore not subject to FDA regulatory oversight.  SACHRP recognizes that there may be instances where the intent of the activity is primarily academic, but subsequent data published as a result of the research may be used to promote off-label uses of a device.  However, whether this concern is sufficient to justify imposing FDA regulatory oversight on the initial research activity is not clear.  In any event, leaving aside historical FDA regulatory interpretations, SACHRP would urge the FDA to consider, insofar as possible, as exempt from IDE requirements data collection and analysis – regardless of the intent to submit those data and analyses to FDA – if the actual treatment rendered to patients was done or is being done as standard of care and with no deviation from the patient’s own best interests, consistent with the physician’s fiduciary duty toward that patient.

Whether the process for gathering the data would influence treatment decisions
 

Whether or not the use of the device is legitimately considered part of the research should depend upon whether and how the patient’s clinical experience is altered by the research activity.  In many cases, data collection on the clinical use of a device occurs independently of the decision of whether or not to use the device in the care of a patient.  In addition, patients may or may not be exposed to additional diagnostic tests or interventions beyond standard clinical care for the purposes of the research.

When the research consists solely of the collection of data collected during the routine care of the patient, and there is no impact on or alteration of standard clinical care, then the only research risks are those related to confidentiality and not those of the device.  However, a determination must still be made as to whether the research is a clinical investigation and therefore subject to FDA regulations.  If determined to be a clinical investigation, then a determination as to whether the study is SR, NSR or exempt would need to be made.  If in fact the research is a clinical investigation, but the use of the device was on-label, then it may be exempt from the requirements for an IDE.  If not exempt and the only risks are determined to be related to confidentiality, then the clinical investigation may be considered NSR.  As noted above, SACHRP believes that it is rare in these cases that the research is SR and would need to be conducted under an IDE.
 

However, if the patient’s clinical experience is directly altered by the investigation, then additional risks would need to be considered.  For example, the decision to use one device over another may be determined by the research, or additional diagnostic studies may be obtained solely as a result of participation in the research.  In these cases, the use of the device and/or the risks related to any additional tests or procedures should be considered as part of the research, and the risks associated with these considered in any determinations.  In these cases, it is more likely that a clinical investigation will qualify as SR and require an IDE.  SACHRP agrees here that if treatment decisions are altered in any way by the “research,” then additional FDA oversight is appropriate.

For registries in which all data collection is retrospective, it would seem apparent that no clinical decisions were affected by the use of the device.  Thus, the only risks are those associated with confidentiality.  Furthermore, if there is no intent to submit the data to the FDA, then such research need not be considered a clinical investigation subject to FDA regulations.  If such real world data is to be submitted to the FDA, SACHRP understands that the research is then a clinical investigation.  It would be extremely helpful for the FDA to clarify whether SACHRP understands this issue correctly, as it appears that FDA has at times issued conflicting opinions.
 

Whether the rights, safety and welfare of human subjects are impacted
 

This third criterion from the draft guidance is difficult to interpret.  It could refer to a right to choose whether or not to participate in research, or whether or not to have private information about their medical condition examined for purposes other than clinical care.  It could also apply to the evaluation of the risk and benefit of research participation as safety and welfare issues.  FDA should provide more clarity on the nature of this criterion and what facts would be relevant to its application in determining whether an activity is a clinical investigation, and if so, the regulatory status of a clinical investigation involving devices.
 

We now turn to other potential criteria that FDA may wish to consider adopting, that were not presented in the draft guidance.

Public Health Authority Activities as Opposed to Clinical Investigations
 

SACHRP recommends that FDA clarify, if possible, that certain registries are established for non-research purposes, such as performing surveillance on medical devices captured within the registry.  These can be considered quality improvement registries under FDA’s activities as a public health authority that does not meet the definition of a clinical investigation. FDA should consider whether this can be accomplished under section 522 of the FD&C Act and/or 21 CFR 822, or some other regulatory provision.  Under 21 CFR 822.3(i), “Postmarket surveillance means the active, systematic, scientifically valid collection, analysis, and interpretation of data or other information about a marketed device.”  Of note, HIPAA permits covered entities to disclose protected health information without authorization for specified public health purposes.  See 45 CFR 164.512(b). https://www.hhs.gov/sites/default/files/ocr/civilrights/education/Documents/publichealth.pdf.  It would be very useful for the regulated community for FDA to clarify this distinction.  Section VI (B) of the draft guidance addresses postmarket surveillance studies under Section 522, but does not address these issues.
 

Retrospective versus prospective
 

Another possible criterion FDA could consider is that retrospective studies such as record reviews and chart reviews, in care delivered as standard of care and with treatment dictated only by patients’ best interests, are not clinical investigations.  In contrast, prospective studies, including registries, are more likely to qualify as clinical investigations when other criteria are included in the analysis, if only because prospective assignment of patients to treatment might be unconsciously influenced by the possibility of enrolling patients in a registry study.  This would provide a useful “bright line” criterion for IRBs and other regulated parties to rely upon.  FDA would need to determine the definition of retrospective.  SACHRP recommends that FDA consider the interpretation that OHRP provided for the current HHS exemption at 45 CFR 46.101(b)(4), that the data must be in existence at the time that the exemption decision is made.  With this interpretation, data that are collected after the date of the exemption decision would not qualify as retrospective.  
 

Identifiable versus deidentified data or specimens
 

Another possible criterion that FDA could offer is that if records or tissue samples are anonymous, with no way to link back to the identity of the individuals, then the use of those data would not be a clinical investigation unless the data is to be submitted to the FDA.  FDA has already essentially adopted this position through enforcement discretion in its “Guidance on Informed Consent for In Vitro Diagnostic Device Studies Using Leftover Human Specimens that are Not Individually Identifiable.”  http://www.fda.gov/RegulatoryInformation/Guidances/ucm078384.htm.  For this to be practical, enforcement discretion on the requirement for informed consent would have to be extended to data that is not individually identifiable, such as imaging results and medical records, and not be solely limited to tissues.  Alternatively, FDA could allow a waiver of consent for such research under the waiver provisions for certain minimal risk research in the 21st Century Cures Act.

FDA may wish to consider that registries that collect data about all patients who are treated for a given disease or condition, regardless of which product or procedure is used, are not clinical investigations, even if they provide valuable information about the safety and efficacy of a given product.   Within this classification of “all comers registries” there will be a spectrum.  For instance, in some cases a registry will be specific to one individual product, such as an artificial knee.  Further down the spectrum, a registry might collect data about a class of devices, such as all artificial hips.   At the other end of the spectrum, a registry might collect information about a wide variety of interventions, including devices, drugs, surgery, and watchful waiting.  For example, a registry could gather information about all individuals treated for progressive heart failure.  In this case, the wide spectrum of information collected about device and non-device interventions might be the basis for deciding that the activity is not a clinical investigation, even if the resultant data is of value in assessing the safety and efficacy of certain devices.
 

On label versus off label 
 

For prospective clinical investigations, FDA regulations and guidance already include a distinction between the regulatory status of on label versus off label investigations.  21 CFR 812.2(c)(2) provides that the following studies are exempted investigations:  
 

(2) A device, other than a transitional device, introduced into commercial distribution on or after May 28, 1976, that FDA has determined to be substantially equivalent to a device in commercial distribution immediately before May 28, 1976, and that is used or investigated in accordance with the indications in the labeling FDA reviewed under subpart E of part 807 in determining substantial equivalence.

In addition, FDA has clarified in guidance that this is also true for devices approved through the PMA process.  The FDA Information Sheet “Frequently Asked Questions About Medical Devices” states, “Similarly, studies of a PMA approved device are exempt from the IDE requirements if the device is being studied for the indications in the approved labeling.”  SACHRP agrees that whether the use of the device is on label or off label is an important consideration in determining whether the research is a clinical investigation and whether it is SR, NSR or exempt.  SACHRP further encourages FDA to provide explicit guidance on how to apply this when considering both retrospective and prospective record review, registry, and observational study designs. 

 

Observational Design
 

If the fact that a study uses an observational design is of relevance to whether a study is a clinical investigation, and if so is of relevance to whether the study is SR, NSR or exempt, FDA should provide details on those factors.  

SACHRP supports that prospective data collection on an off-label use of a device with the purpose of gathering data about the safety and efficacy of the off label use of device is a clinical investigation of a device, and therefore is subject to all FDA regulations, if only because, as stated above, there is a conscious or unconscious temptation for an investigator’s treatment decisions to be influenced by the availability of enrollment into such a prospective study.  SACHRP recommends that FDA makes it clear that all such studies are subject to FDA regulatory oversight and are not exempt from the IDE requirements, and must therefore be conducted under an IDE or abbreviated IDE.
 

Risk analysis of extra procedures conducted only for research purposes
 

FDA should clarify that for prospective registries or observational studies that are clinical investigations, as long as the decision to use the device was made clinically and is completely independent of the existence of the registry or observational study, that in addition to confidentiality risks, the only risks related to the device that should be considered for an SR/NSR decision is the risks of any extra procedures or interventions that are conducted expressly for the research.  Such procedures might include, for example, additional x-rays, MRIs, blood draws or questionnaires.
 

Other Possible Criteria SACHRP Does Not Support
 

SACHRP would also like to address other possible criteria that have been discussed as objective criteria that IRBs, sponsors, investigators and other entities could use to determine the regulatory status of these activities, but which SACHRP does not support.
 

Intent to commercialize the product
 

A possible criterion is whether the intent of the research is to commercialize the product or otherwise use the research results to generate revenues.  One way to apply this would be for FDA to adopt the position that if the intent is to use the data in the commercialization of the product, then the activity is a clinical investigation, even if the research is a retrospective records review or an interview with investigators or patients.  The decision of SR, NSR or IDE-exempt would then have to be made.  Conversely, some types of research such as on-label studies of approved devices would not be clinical investigations if there were no intent to commercialize the product or otherwise directly financially profit from the results.  Many research projects are conducted out of clinical interest, to publish, or to complete requirements for advanced degrees.  It is worth noting that for drugs, FDA has determined that this criterion is not applicable.  The FDA guidance “Determining Whether Human Research Studies Can Be Conducted Without an IND,” states that “whether the IND regulations apply to a planned clinical investigation does not depend on whether the intent of the clinical investigation is commercial or noncommercial.”  SACHRP recommends that FDA issue guidance stating that, similarly, whether the IDE regulations apply to a planned clinical investigation of a device does not depend on whether the intent of the clinical investigation is commercial or noncommercial.  This has been an issue of some confusion lately.
 

Use of the criteria
 

If FDA were to provide such criteria, then the criteria could be used to form rules.  Some examples of such rules could be:
 

• Retrospective chart reviews of treatment rendered as standard of care are not clinical investigations.

• Prospective registries that are established with the purpose of gathering data about the safety and efficacy of the on label use of devices are clinical investigations but exempt from IDE requirements.  If the clinical decision to use the product is made independent of the existence of the registry, and there are no additional tests or procedures performed for research outside of those that would be performed for routine clinical care, the registry will be exempt as use of the product on label.

• Prospective registries established with the purpose of gathering data about the safety and efficacy of the on-label use of devices where the clinical decision to use the device is influenced by the research are clinical investigations and may be SR, NSR, or exempt.  The risks of the device itself and any additional testing or interventions that are performed for research purposes should be considered in making this determination.   

• Prospective registries that are established with the purpose of gathering data about the safety and efficacy of the off label use of devices are clinical investigations.  Even if the clinical decision to use the product is made independent of the existence of the registry, the research is not IDE exempt and the risks associated with the use of the device must be considered in determining whether the registry is SR or NSR.

Clear criteria and rules such as these would be of great value to the regulated community.  The examples given in the draft guidance of when an IDE is necessary for the collection of RWD do not provide sufficient information to allow such criteria or rules to be identified, and as such unfortunately do not provide the regulated community with tools to utilize to determine when an IDE is necessary.  
 

Part 2: Analysis of Examples of When an IDE is Necessary for Clinical Investigations Involving RWD
 

In the draft guidance document, there are several cases provided regarding studies that need an IDE and studies that do not need an IDE.   However, it is difficult from the examples provided to determine why these studies either require or do not require an IDE.   We discuss several of the examples below.

 

Section IV (B) of the draft guidance addresses the application of Investigational Device Exemption (IDE) requirements in 21 CFR 812 to the collection of real world data.  In this section, FDA describes three factors that can inform the decision of whether an IDE is necessary:

 

Whether the collection of RWD could be subject to the IDE regulations depends in part on whether that collection constitutes a clinical investigation. Several factors can inform this determination, including the purpose for which the data is being gathered, whether the process for gathering the data would influence treatment decisions, and whether the rights, safety and welfare of human subjects are impacted, among other things.  (Lines 325 to 329.)

As discussed above, SACHRP believes it would be very helpful for FDA to provide further explanation of these three factors, so that they can be used by IRBs and other regulated entities to make correct and independent decisions regarding the need for an IDE.  

Section IV (B) of the draft guidance also states:

Because the gathering of RWD is unique from traditional investigations, we believe that the determination of whether an IDE is required should be made on a case-by-case basis, and we recommend that you contact FDA about whether an IDE is required in cases where RWD collection is initiated for purposes of determining the safety and effectiveness of a device.  (Lines 335 to 338.)

Registries are a very common research methodology, and are used by a wide variety of investigators in many different settings.  They are often conducted with the intent to publish an article or satisfy academic degree requirements.  SACHRP recommends that FDA provide clear criteria that IRBs and investigators can use to determine the regulatory status of such studies rather than FDA making the decision on a case-by-case basis.  To require individual review of all such study designs would be burdensome to the FDA, investigators, sponsors and IRBs.  

The draft guidance in IV (B) also addresses the use RWD collected for purposes other than establishing safety and efficacy:

However, FDA does not regulate the practice of medicine, and recognizes that some RWD is collected for purposes other than establishing the premarket safety and effectiveness of a device, such as the collection of information related to the actual use by clinicians of an approved or cleared device and/or treatment approaches for a particular disease or condition. Such observations may include RWD from a use of a medical device that was not within the cleared or approved indications for use. When such RWD collection is not intended to determine the safety and effectiveness of the device for purposes of supporting a marketing application to FDA, it would likely not meet the definition of a clinical investigation, and the IDE regulations would not necessarily apply. (emphasis added)  (Lines 340 to 348.)

We encourage FDA to clarify the criteria that apply to the determination that a given research project would likely not meet the definition of a clinical investigation, and the IDE regulations would not necessarily apply, so that sponsors and IRBs can make accurate and compliant decisions regarding the need for an IDE without going to FDA for a formal opinion each time.  

In Section VI (A) the draft guidance provides examples of the use of RWD for expanded indications for use.  The first example is the National Cardiovascular Data Registry (NCDR).  FDA states that an IDE was not necessary for this registry:

As a registry collecting data on consecutive patients and focused on quality assessment/performance improvement data related to real-world procedures and device use outcomes, an IDE is not required for routine data collection operations, even though a substantial volume of data is generated from use of a device, including data on use outside of the cleared or approved indications for use.  (Lines 549 to 554.)

It is not clear from this description what criteria are being used to make the determination that an IDE was not necessary.

The draft guidance in the same section then provides a second example of the use of RWD for an expanded indication:

Another example is a Class III device with a narrowly defined indications for use that over time, has seen an expansion in clinically accepted use that is outside of the approved indications for use. In this example, recent technological advances in the design of these devices have also led to their rapid and widespread use for a new set of clinical applications that are not described in the approved labeling.  (Lines 555 to 559.)

In this case, the draft guidance states that an IDE was necessary, or at least was applied for and granted:

To address the lack of data to support new indications for use for this device, relevant medical societies have established a national registry to collect safety and effectiveness information for all patients implanted with this specific Class III device at participating institutions. A study using the registry data collection and analysis infrastructure was initiated with an approved IDE application since the study focused on a use of this device that was not within the approved indications for use and imposed collection of specific follow-up data that might not otherwise be performed as part of standard medical care.  (Lines 562 to 569.)

It is not clear from this description what criteria were used to make the determination that an IDE was necessary, particularly in contrast to the registry described directly above in the same section.  It would be very helpful to IRBs and other regulated entities if FDA would provide the criteria used in reaching these contrasting decisions regarding the necessity for an IDE.

Section VI (B) of the draft guidance addresses postmarket surveillance studies under Section 522.  The draft guidance does not indicate whether an IDE was required for these studies.  It would be helpful if FDA stated whether or not an IDE is required for these studies, and the criteria by which the decision is reached.

Section VI(C) of the draft guidance addresses post-approval device surveillance as condition of approval.  An example is provided of a new breakthrough Class III medical device that was recently approved based on prospective randomized controlled clinical trial data, and a registry was launched that provided data to support FDA and CMS data requirements and national quality assessment programs, in addition to the primary clinical quality assurance purpose desired by the clinical community.  In regard to this registry, the draft guidance states:

 

No IDE is necessary for the general data collection activities of the registry, as it collects data on all uses of otherwise approved medical devices. The retrospective analysis of data from uses that are outside the approved indications for use did not require an IDE, but was reviewed by an IRB for human subject protection issues. However, prospective enrollment of new patients into a clinical trial using the registry infrastructure meets the definition of a Clinical Investigation and is subject to 21 CFR 50 (Informed Consent) and 21 CFR 56 (IRB Review).
Additionally, if the prospective enrollment is considered significant risk and is being used to determine safety and effectiveness of a medical device, an IDE approval will be required. (Lines 616 to 624.)

It is not clear from this description what criteria were used to make the determination that an IDE was not necessary for this registry.  Also, the fourth sentence says that if the prospective enrollment is considered Significant Risk, then an IDE will be required.  It would be helpful to clarify the criteria FDA uses to determine whether or not a given prospective enrollment registry or study is SR.  It could be assumed, but it is not stated, that if the decision to use the device is only for clinical purposes, and not due to the existence of the registry, then it is NSR.  If these are the criteria, then the guidance should clearly so state.

Section VI (D) addresses the use of a registry as a control group in a clinical study of a new device.  The draft guidance says:

The manufacturer designed a clinical study that compared the use of the new device to a non-randomized concurrent control group derived from the registry. The existing registry was evaluated by FDA and the manufacturer according to the factors cited in this guidance and was found to provide sufficient data on the control population, such that the manufacturer did not have to collect additional data from these patients or influence the course of their clinical care in any way.  (Lines 630 to 635.)

It would be helpful if the draft guidance noted whether an IDE was required in this case, and whether the registry arm was considered to fall under the IDE.  Also, because the FDA definition of a human subject in the FDA regulations includes controls, presumably consent was required from the control arm subjects.  It would be useful to clarify this, as it is commonly an issue of contention between sponsors and IRBs. 

SACHRP makes the following specific recommendations regarding the issues discussed above:

SACHRP has recommended in the past that FDA adopt the waiver of consent that exists in the Common Rule at 45 CFR 46.116(d).  This would serve the goal of creating harmonization between the FDA regulations and the Common Rule.  However, in light of the passage of the 21st Century Cures Act and its provisions for an FDA waiver of consent for certain minimal risk research, SACHRP also notes that the Common Rule waiver of consent was in part originally drafted for the use of deception in behavioral research, and that this might be an appropriate opportunity for FDA to create a waiver of consent more directly applicable to the use of RWD to support research that provides evidence of the safety and efficacy of devices.  SACHRP will be pleased to provide further input on this issue at FDA’s request.

 

FDA should include a section on the Role of Institutional Review Boards in the guidance, “Use of Real-World Evidence to Support Regulatory Decision-Making for Medical Devices.”  FDA used this approach regarding in the FDA guidance “Guidance for HDE Holders, Institutional Review Boards (IRBs), Clinical Investigators, and FDA Staff - Humanitarian Device Exemption (HDE) Regulation: Questions and Answers,” and it is extremely useful to IRBs.  It is available online at, http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm110194.htm.

FDA should clarify that retrospective chart reviews are not clinical investigations, but that FDA can consider the evidence produced from them for purposes of establishing safety and efficacy of devices.

FDA should clarify that prospective registries established with the purpose of gathering data about the safety and efficacy of the on label use of devices are clinical investigations but are exempt from IDE requirements.  Further, FDA should clarify that if the clinical decision to use the product is made independent of the existence of the registry, then the risks associated with the use of the device need not be considered in determining whether the registry will be SR, NSR or IDE-exempt under 21 CFR 812.2(c).  If the registry does not alter clinical care in any way, then the registry should be IDE-exempt as use of the product on label.  If there are additional tests or procedures performed for research purposes, then the risks of these must be considered in determining whether the registry is NSR or SR.

FDA should clarify that prospective registries that are established with the purpose of gathering data about the safety and efficacy of the off label use of devices are clinical investigations.  Even if the clinical decision to use the product is made independent of the existence of the registry, the risks of the device must be considered and the registry determined to be SR or NSR.

FDA should clarify that if data are gathered for purposes other than establishing the safety and efficacy of devices, for instance if data are collected for benchmarking purposes, or as a quality improvement initiative by a professional association, or for insurance purposes, that the collection of the data is not a clinical investigation.  FDA should also clarify that it can use this data for consideration of the safety and efficacy of a device.  Another way to frame this consideration is that if the collection of the data is incidental to the purpose of establishing safety and efficacy rather than planned to establish safety and efficacy, it is not a clinical investigation.

FDA should clarify that the position regarding commercialization as a criterion for whether an activity is a clinical investigation, as stated in the FDA guidance “Determining Whether Human Research Studies Can Be Conducted Without an IND,” applies to devices as well as drugs.  That guidance states that “whether the IND regulations apply to a planned clinical investigation does not depend on whether the intent of the clinical investigation is commercial or noncommercial.”  SACHRP recommends that FDA issue guidance stating that, similarly, whether the IDE regulations apply to a planned clinical investigation of a device does not depend on whether the intent of the clinical investigation is commercial or noncommercial.  It should also be clear that this applies to all three categories of clinical investigations of devices, SR, NSR, and IDE exempt.

The draft FDA guidance “Use of Real-World Evidence to Support Regulatory Decision-Making for Medical Devices” addresses very difficult and important issues in device development and IRB review.  It also provides an opportunity for FDA to provide the regulated community with clear objective criteria that IRBs and other regulated parties can use to determine when research is a clinical investigation, and if it is, when it is SR and needs an IDE, when it is NSR and can proceed under the abbreviated IDE requirements, and when it is exempt from the IDE requirements.  FDA should take this opportunity to establish such criteria.  The current version of the draft guidance does not provide sufficient information to be of utility to sponsors, investigators and IRBs, and because of the limited fact patterns does not provide objective criteria for determining whether a given clinical investigation needs an IDE.  Such guidance would be of significant benefit to the regulated community.  SACHRP thanks FDA and OHRP for this opportunity to provide this commentary.