• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Medical Devices

  • Print
  • Share
  • E-mail

Summary of Breakout Sessions for Public Workshop - Using Scientific Research Data to Support Pediatric Medical Device Claims, December 5, 2011

February 28, 2012

Background

There is a substantial unmet need for safe and effective pediatric medical devices. Pediatric medical devices prevent, treat or diagnose diseases and conditions that occur in children across the pediatric age spectrum, from birth through the 21 st year of life. Some medical devices are designed specifically for pediatric use, while others are adapted from specific adult device applications or produced for more general use.

In 2007, Congress passed the Pediatric Medical Device Safety and Improvement Act (PMDSIA), which permits: extrapolation of adult effectiveness data to support a pediatric indication if the disease course or the effect of the device in adults is likely to be the same in children and extrapolation of effectiveness from one pediatric subpopulation to another subpopulation. Extrapolation of effectiveness is limited to FDA-approved medical devices (i.e. premarket approval (PMA) devices and humanitarian device exemption (HDE) devices). The PMDSIA does not authorize extrapolation of safety data, but does allow longer prospective surveillance, beyond 36 months, if needed to assess the impact of the device on growth, development activity level, or other factors related to the safety or efficacy of the device.

Extrapolation of effectiveness simplifies the requirements for establishing a pediatric intended use claim, reduces the need for and complexity of clinical efficacy studies to establish pediatric claims, and facilitates efforts to address unmet medical device needs for children by making optimal use of what is already known to increase efficiency and reduce regulatory burden.

What Happened at the Workshop

To support establishment of a framework for extrapolation of effectiveness data to support pediatric medical device claims (and thus increase availability of approved medical devices that are labeled for pediatric use), the FDA held a public workshop on December 5 th, 2011 at its White Oak campus in Silver Spring, Maryland. The workshop participants included representatives from industry, health care, government, and pediatric academic associations, patient organizations and other pediatric device stakeholders.

The Workshop program consisted of three brief plenary sessions, during which FDA employees provided an orientation to specific topics that were discussed in more depth at small group breakout sessions immediately after each plenary. At each breakout session, each group brainstormed together to generate ideas and issues relevant to a specific focus question. Each breakout group then organized their issues and topics according to larger themes; these thematic summaries were shared with participants after each breakout.

A transcript and the webcast of the workshop plenary sessions are available on FDA’s webpage for this event. However the recording process was not able to adequately capture the work effort of the breakout groups, which is provided here. The ideas and issues developed and grouped in each breakout session are presented in Appendix I in raw form, as they were captured and laid out during the workshop itself (without further grouping or editing). This summary includes a range of suggestions, ideas and comments. None have been verified or assessed for feasibility, ethics, scientific rigor, legal or regulatory compliance. None of these ideas represent FDA policy guidance, or regulation.

Next Steps

FDA appreciated the investment of time and thought give by workshop attendees. The stakeholder input on from this workshop will be considered as the Center establishes a framework for the use of scientific research data to support pediatric medical device claims . Of note, many of the ideas and issues generated by the breakout groups went beyond the specific topic at hand, addressing other pediatric device development issues. FDA will consider and revisit the issues and topics raised at this workshop in an ongoing way, as it continues to establish and develop efforts to support and foster the development of medical devices approved for pediatric use.


Appendix I: Ideas Generated During FDA Pediatric Medical Device Workshop Breakout Sessions Organized by Breakout Topic and Theme

Topic 1: Defining the Useful Research Data Landscape

Focus Question: How might the available research data sources be used for extrapolation of effectiveness for pediatric use of devices?

1) Incentives
  • Closed FDA panel meeting
  • Pediatric Research Equity Act (PREA)
  • Additional funding
    • Fund collaborative registries like the American College of Cardiology Foundation's IMPACT Registry (IMproving Pediatric and Adult Congenital Treatment)
    • More federal funding from National Institutes of Health (NIH) in all their centers
  • Incentives to develop pediatric/innovations
  • Capture information about pediatric off-label use
    • anonymous or “de-indentified” reporting
    • Third Party administration of registry of Pediatric use
    • Use insurance company/Medicare as source of information on actual pediatric device use
    • Pediatric Medical Records
    • OUS experience – where approved indications are broader
  • Successful device use in pediatrics off-label to advertisement incentives
  • Recognition of the role of educational safety support that does not promote off-label pediatric use (goal is labeled use)
  • Need an incentive “carrot”
2) Data Sources
  • Standard data collection at the source
  • Natural History for Control on high priority needs for pediatric medical devices
  • Rest of world experience – where approved indications are broader
  • Off-label registry (potential reporter liability, but important efficacy data source)
  • Use registry for basis of studies
  • Develop a repository of all devices uses (not only a compilations database e.g. MAUDE)
  • Animal research to inform outcomes
  • Use published studies
    • as case reports small studies
    • to develop endpoints
  • Adult data
  • Pediatric Medical Records
  • Databases for Children’s Hospital
  • Charge data master
  • Collaborative data sets
    • I.e. Regional consortiums of healthcare organizations
  • Already available data sources
    • State reporting from Joint Commission
    • Insurance/Payer data on device use
    • Office-based data sources
  • Link clinical and device data

Data Quality Concerns/Relevance

  • Off label use (N =1)
    • Minimal data for treatment failures
    • Lack of uniform standard of care
    • Data sources lack info on techniques used
    • Changes in indications for device
  • Data mining for extrapolation for device failure
  • How to weigh data from different sources
  • Prevent Data case duplication
  • Data complete?
  • Data Accuracy?
  • Show similar disease process
  • Factor incidence of disease
  • Normative data to determine what is normal
  • Interpret function
  • Factor indication
  • Factor age group
  • Limited availability of studies
  • Human testing of device for pediatric population
  • Cannot conduct study with control population
  • Difficulty recruiting pediatric patients to substantiate extrapolation
  • Ethical considerations re: pediatric studies
  • Limited understanding for parental consent
  • Assessment of device with non-verbal patients

Endpoint Development

  • Natural History of the disease
  • Establish animal models
  • Innovative clinical trails
  • Review existing endpoints
  • Utilize expert opinion to derive definitions
  • Comparative effectiveness
  • Intended use of device
  • Physiological differences
  • Different metabolism for pediatric patients
3) Barriers
  • Challenges to conducting randomized clinical trials for pediatric devices
    • Open-label study
    • small sample size
    • PMA applications require/request Pediatric registries for effectiveness
    • Concerns about delayed treatment when off-label in standards of care in pediatrics
    • Lack of controlled (reliable) data
    • Lack of reproducible data
    • Need to address all pediatric subgroups for FDA clearance/approval
  • Concern registry reporting liability compliance risk
  • Effectiveness – What to use for controls with a registry?
  • Timely consolidation of available raw data
  • Database for pediatrics – combines multiple sources
  • Reliability of post market
  • FDA upload/standardize post market data
  • Priorities data sources
  • Usability of data
  • Global and US data integration
  • Informed consent
  • Barriers have more impact on marginally profitable companies
  • Ability to access NIH data and other federal agency data
  • Awareness of available databases
4) Improve access to viable, available data
  • Central Repository for Data Analysis
    • Improve access
    • On-line access
    • How/no cost
  • Combine Registries (Government/Industry/ Research)
  • Evidenced-based outcome data (published literature or academic)
5) New idea for data generation
  • Computer simulations/physiologic modeling
  • Extrapolate to most similar age group
6) Creating New Collaborations
  • Mining registry data for patient populations
  • Mining clinical databases online from institutions
  • Data shopping list –piggyback on data (A comprehensive list from Industry and researchers that includes data needed for extrapolation)
  • Reach out to professional societies of prescribers and users
  • Create collaborations with innovative and premier pediatric clinical centers
  • Identify smaller registries by creating a centralized location
  • Organizations taking the lead
  • Input from specialists
  • Government/Industry/ Research partnerships
7) Regulatory
  • What are the priorities for development?
  • Database content
  • Regulations (require pediatric related data)
8) Education
  • Education regarding possibilities
  • Good public relations about possible pathways

Topic 2: Defining scientific and regulatory challenges and limitations with the use of existing research data and published literature

Focus Question: Considering the landscape of data sources identified in Breakout Session #1, what specific scientific and regulatory issues must FDA consider when using these data to extrapolate or establish pediatric effectiveness for various medical devices?

1) Regulatory flexibility premarket
  • Requirements for Medicaid need to be similar to Medicare
  • FDA needs to provide rational to industry to support devices
  • Stepwise to collection of data through registry to lower population age
  • Do not understand FDA’s concerns
  • Clinicians will want to do sham constraint
  • Industry cannot meet regulatory requirements
  • Use baseline data rather than sham
  • Appeal process when premarket requirements are impossible to meet for pediatrics.
  • Regulatory lenience risk/benefit pediatric
  • Expand Orphan Product regulatory benefit to pediatric device industry
  • Financial incentives – How to pro rate?
  • Consistent appropriate definition of pediatrics
  • Safety and effectiveness intertwined
  • Disconnect between FDA Office of Orphan Products Development (OOPD) and CDRH
  • Level of evidence for “probable” for humanitarian device exemptions (HDE)
  • Not device specific
  • Development of device specific guidance from FDA
  • Consistent regulation authorities (e.g. OSHA, FDA, NIH…)
  • Allow access to “proprietary” industry-sponsored raw data if pooled blinded
    • Who will accomplish this? Need a driving force.
  • Specific laws and regulatory protecting children
  • Validation of study design
  • Applicability of study design
  • Accounting for children/adult difference
  • Lack of test methods specific to pediatric population
  • Early thinking about looking at pediatrics. When and where it makes sense.
  • Track off-label use
  • Lack of pediatric norms
2) Effective Collaboration/Experts
  • FDA collaborate with Payers (United HealthCare Consortium (UHC), Health Canada) for collection of information patients knowledge base building
  • Data to support pediatrics Reduce Regulatory Control
  • European Union (EU) data globalization Canada/EU Bench marking
  • Industry-Government partnership grants to support cost
  • National recognized independent Institutional Review Board (IRB) instead of hospital IRB
  • Consider approach to randomized control trials.
  • Creating equivalent of Centers for Medicare and Medicaid Services (CMS) model for data collection clinical evidence development
  • Payer data may not capture effectiveness
  • Partner with Medicaid for pediatric data collection
  • Use of network of experts
  • Increase use of advisory panels
  • Adequate pediatric expertise at FDA/CDRH
  • Requirement for mandatory pediatric clinical consultation
  • Long-term follow up for unexplained adverse outcomes
  • Standardize pediatric practice through out CDRH similar to CDER
3) Risk/Benefits Pediatrics
  • Regulatory lenience risk/benefit pediatric
    • Outgrow device - failure
    • Smaller data points even in young adults
  • Recognition of pediatric/devices failure
  • Instruction for use which devices needs to change
  • Recognize device failure pediatric activity
  • Historical data
  • Sample size/ pediatric size
  • Endpoint/testing number of pediatric /adults
  • Support and incentivizes for cross-cutting pediatric technology
  • Congress to encourage data collection
  • Lifetime of device
  • Lenient process to use collection data for approval devices in pediatric
  • Difference in implants regulatory process
  • Study lifetime of devices over time
  • Mechanism of disease adults/ pediatric
  • Mechanical stress up for pediatric
  • Evolving timeline of technology
  • Acceptance of evolving data for current times
  • Utilize current clinical practice as reference points
  • Evaluation of label use
  • Impact of behavior/psychosocial factors
  • Unique anatomy development
4) Financial Issues
  • Cost of data approval
  • Cost of sharing through partnership
  • Create financial Incentives partnership NIH/FDA
  • Change law clinical trial cost hospital
  • Reimbursement issues
  • Lack of financial incentives
  • Incentives to get companies to invest
  • Funding to report the research data already known about
5) Level playing field  
  • Landscape fairness
6) Growth and Development
  • Prevalence of disease/condition
  • Impact on different body systems
  • Immunogenicity
  • Hormonal Influence
  • Disease process no analogous
  • Growth and develop
7) Effectiveness
  • Foreign Development
  • Reliability of data
  • CE Market reliability
  • Publication acceptability
  • How much data is enough?
  • Post market follow-up
  • Compliance and adherence
  • Liability issues for FDA and firm
8) Need to engage pediatricians
  • Exercise mobility of children
  • Difficult in gathering information from users
  • Patient reported outcomes
9) Study Design Concerns
  • Data- how to share with all the stakeholders
  • Foreign data
  • IRB interoperation of HDEs “research”
  • Statistical validation
  • Need to talk to clinicians re: end points, use
  • Link clinical and device data
  • Definition of “effective”
  • Appropriate endpoints for effectiveness
  • Applicability of some disease processes
  • Biological Plausibility
  • Incorporate study data for extrapolation
  • Look at HDE as available data
  • Encourage industry to obtain HDE
  • Age vs maturity e.g.: skeletal maturity and effect of disease
  • Species difference in physiology and anatomy e.g. : “juvenile” animal models
  • Lack of registry
  • No guidance on adverse events
  • Case by case data only
  • Scientific differences (path physiology)
10) Labeling
  • No established regulatory framework specific to extrapolation
  • Labeling pediatric specific
  • How does the activity level of children impact the functioning of the device compared to the older populations?
  • Long term issues/effects
11) Need for guidance
  • Regulatory definition of adverse events
  • Should there be a more novel pathway?
  • Labeling e.g.; Dialysis not to be used in patients under 20 kg but no other option
  • No focused framework
12) Human Factors
  • User reliability (if needed to interface with device)
  • How to measure human factor effect

Topic 3: Overcoming the challenges and limitations- Identifying methods to address the pitfalls and data gaps, including statistical approaches and modeling

Focus Question: How might these and other approaches be used to address data gaps and support extrapolation and establishment of effectiveness claims for pediatric patients?

1) Trials need to be less burdensome – options and incentives.
  • Randomize sites vs. randomize patients when the disease course is similar (same criteria Randomized Clinical Trails (RCTs)
  • Unethical to RCTs
  • Transparency justification device making process
  • Address hospital cost for clinical trials
  • Pre market flexibility
  • To get on the label do not ask for randomized control
2) Challenges and benefits of Bayesian approach
  • In support of statistical approach
  • Use data submitted for adult premarket approval (PMA) to demonstrate effectiveness for peds as historical control
  • Preferable to use Bayesian than small sample size
  • Prior study agreement for Bayesian approach
3) Other approaches as sources data
  • Simulated clinical computational model
4) Data Appropriateness
  • Is it appropriate to borrow? Data availability?
  • Ensuring historical controls are appropriate
  • Equivalence claims for pediatric based on historical data
  • Validation of data
  • Data format
  • Source of data and degree of bias
  • Assess data
  • Define types – implant vs. external
  • Device types are they current?
  • Same devices may be easier than others
  • Similar diseases and similar endpoints
  • Long-term growth vs. device size
  • Age vs. size of patient
  • Spread across subpopulations
  • To organize available data by age to get subpopulation information
  • To identify ease of use of data for adolescents
  • Ethical issues addressed up front
  • Global consent form
  • Discussion with parents who experience multiple device failures
  • Less stringent institutional requirements
  • How do you know if you have “enough” data?
  • U.S vs. non U.S. data
  • Data validation
  • Number of patients need for each sub population
  • Larger amount of data available that might be appropriate for adolescents vs. younger patients
  • Need for specific expertise
  • Exchangeability of data – if disease is similar
  • Combination products look at CDER and CBER for data
  • Accept the statistical models that have been preferred

Data Sources

  • Data from Adult and General Hospitals vs. Pediatric Facilities (different providers for children)
  • International data may suggest effectiveness of other uses/methods of implantation of devices
  • Look at data sources for drugs whose devices are used already, e.g. infusion pumps
  • Published literature may allow pooling of data (for metaanalysis)
  • Data from drug studies where devices are used too e.g. infusion of drugs
  • Kids treated at “adult” vs. “pediatric facility, e.g. “ARC”
  • Diseases and endpoints same for adults and kids
  • Data with similar diseases similar endpoints
  • Big hospital databases
  • Clinical trials
  • Registries
  • Medical Device Report (MDR) data
  • Health Hazard Evaluations (HHEs) include information from FDA
  • Post market safety data concerns may also include problems with effectiveness
  • Recall data
  • Exchangeability of data – if disease is similar
  • Accept the statistical models that have been proffered
  • Pull from different data sets and evaluate pool ability of data
  • Combination products – look into CDER and CBER data sets
  • How to use off-label data to meet pediatric needs?
  • Combine data from similar devices
  • Different stage of disease time point of intervention
  • More limited Aims with expanded post-market analysis
  • Reconsider age barriers and definitions
  • Validate imputed control
  • Consider - value of off-label use
  • Closer frequent monitoring
  • Use anecdotal data (to start)
  • Non-FDA use of outside of the US data
  • Appropriate historical control
  • Cluster design control
  • Labeling should not be the only risk mitigation
  • Smaller Subset Studies
  • Use of meta analysis to combine different sources of data together
  • New/different outside RCT’s
  • Define missing information in current data pool
  • Publish data: of off-label use, or post market data
  • Smaller clinical trails and shorter studies based upon the device
  • Use of historical controls and treatment arms
  • Use more modeling

Unmet needs/ways to get missing data

  • Look for patterns
  • Early identification of challenges
  • More FDA input on study designs
  • Think Tanks
  • Consider innovation pathway
  • Develop registries
  • To look at unmet pediatric needs
  • To identify patterns in the data for certain devices
  • Use data from off-label use
  • Public access data banks.
  • Use more modeling
  • Use modeling for effectiveness
  • In-silico modeling used for artificial pancreas – Doing simulated studies better
  • Virtual clinical studies
  • Develop more tissue banks – take sample when implanted
  • “Virtual Clinical Trial”
  • Look at different more creative models such as computer modeling
5) Benefits of using historical controls
  • Historical data surgical controls to compare to device arm
  • Use adult data as historical data for control arm
  • Historical Controls may be good for “pure” pediatric devices if there is no adult data
  • Historical controls may be more desirable when there is and ethical or surgical risk
6) More collaboration/Sharing of resources
  • Collaboration with hospitals and users
  • Industry collaboration
  • Partnering with National Institute of Child Health & Human Development (NICHD)
  • More collaboration between agencies
  • Lack of research
  • Lack of funding
  • More intra agency collaboration
  • More collaboration with trade organizations
  • Collaboration among industry, academia, and clinicians
  • Multi disciplinary team of clinicians for input
  • A separate pediatric section – increased clinical expertise, increase funding, internalize experts
7) Reducing the burden on companies
  • consider alternatives to multi-center trials
  • Smaller studies
  • Use patient as own control
8) Design alternatives
  • Adaptive design
  • Establish standardize risk/benefit scales
  • Combine small trial with ps registry approaches
  • Better publicize FDA’s willingness to use the Bayesian approach
  • Look for alternatives from other fields
  • Development of new validated accurate precise endpoints or surrogates
  • Small trial size design similar to National Aeronautical and Space Administration (NASA) – Safety and effectiveness of single arm
  • Acceptance of “standard of care” as control arm