MDEpinet Tools and Services
CRN Maturity Model
MDEpiNet has developed a CRN assessment tool based on the seven domains of maturity. An expert group of MDEpiNet collaborators from academic, clinical, industry and regulatory settings participated to develop the framework for CRN maturation. The seven domains of the CRN maturity model are: Device Identification; Patient reported outcomes and Patient engagement; Data Quality; Efficiency; Governance & Sustainability; Health Care Quality Improvement; and Total Product Life Cycle. Each CRN will be able to use this tool to self-assess their maturity.
Active Surveillance Tools
To support long-term device surveillance MDEpiNet is advancing active surveillance methodologies for Coordinated Registry Network. The Center is developing a flexible tool to provide users with timely and comprehensive evaluations of medical device safety signals. There are various projects underway that use an integrated tool to implement DELTA and other outlier detection in CRNs.
Evidence Review and Synthesis Methodologies
Systematic Literature Review is an important tool used by MDEpiNet Coordinating Center. We have a systematic process of study identification and data abstractions. The studies are reviewed by experts using Medline, Embase, Cochrane Controlled Trials Register, reference lists of articles, annual reports of major registries, summaries of safety and effectiveness for pre-market application and mandated post-market studies at the FDA. Data synthesis are routinely conducted using advanced methodologies.
Validation of Claims Data and Library of ICD-9 and ICD-10 Codes
To facilitate device research in contemporary era, adapting to the transition from ICD 9 to ICD 10 is critical. Various conditions and events have been defined and translated to ICD 10 algorithms to support research based on CRNs. The clinical accuracy of ICD 9, ICD 10, and CPT codes is at times unknown, although critical to device and comparative effectiveness research. Validation studies of these coding definitions have been implemented to verify their clinical accuracy.
IDEAL-D: a rational framework for evaluating and regulating the use of medical devices. This approach enables innovators and manufacturers to initiate clinical studies of devices at every stage of developmental total product life cycle (TPLC). Starting from first in human studies followed by early stage case series, comparative investigations, and surveillance studies, IDEAL-D highlights study designs and robust methodological approaches to evidence generation.
Distributed Analytical Methodologies
To develop methodologies that enable the distributed analysis of international data. In this approach, a standardized data extraction is implemented and distributed to participating registries. Each registry then completes the analyses of their own registry and completely de-identified data summaries are sent back to the coordinating center. Data are then combined using multivariable hierarchical models to evaluate comparative outcomes of devices regarding the main patient-centered outcomes.
Objective Performance Criteria Development
OPC is a numerical target value of safety or effectiveness endpoints derived from historical data from various data sources, such as clinical studies and/or registries. OPCs may be used in single-arm device evaluation that has regulatory implications. The Center is developing advanced methods to construct OPCs, combining estimates obtained from different approaches, ranging from direct analysis of registry or claims data to those reported in literature.
Advanced Study Design and Analytics
Utilizing expertise in epidemiology, health services research, biostatistics, and regulatory science to help investigators construct studies to answer important clinical and regulatory questions pertinent to medical devices. The Center offers advanced study design and analytical guidance and support (e.g. propensity score, inverse probability weighting, classification trees, and other methods) to help investigators navigate around complex device questions.
The use of CRNs and RWD aid in overcoming many limitations associated with post-market studies and may also reduce the costs and save time for evidence generation. To demonstrate the Return on Investment (ROI) of CRNs, methodologies calculating the ROI and Days Saved among necessary regulatory studies conducted in CRNs compared to those conducted in the absence of CRNs were developed. These methodologies aid in the evaluation of CRNs, demonstraion of favorable ROI, and emphasize the value of RWD sources.
Patient Partnership Development
Patients are an important partner in the MDEpiNet public-private partnership and a critical voice of many of MDEpiNet’s projects. MDEpiNet patient partners work alongside clinicians, researchers, device manufacturers, FDA and other federal agency staff to develop and improve real-world data collection and analysis in a variety of clinical areas. Patient partners are identified for CRN or a clinical area through a working group and included in decision making process for project related tasks like 1) roundtable meetings 2) core minimum dataset development 3) stakeholder engagement meetings.
To facilitate consensus on important aspects of registry advancement, MDEpiNet supports the CRNs to conduct the Delphi process by convening key stakeholders. Undergoing a Delphi process is a preferred method for reaching concordance about a core minimum dataset as traditional consensus panel approaches has challenges such as bias and lack of anonymity. As a result, the Delphi process was developed to achieve consensus while minimizing bias inherent in group dynamics and face-to-face responses.To facilitate consensus on important aspects of registry advancement, MDEpiNet supports the CRNs to conduct the Delphi process by convening key stakeholders. Undergoing a Delphi process is a preferred method for reaching concordance about a core minimum dataset as traditional consensus panel approaches has challenges such as bias and lack of anonymity. As a result, the Delphi process was developed to achieve consensus while minimizing bias inherent in group dynamics and face-to-face responses.
Data Standardization/Minimum Core Data Elements/Harmonization Tools
The minimum core data sets are drafted by the subject matter experts in each CRN. Once the concepts are drafted, they are modeled in Enterprise Architect (a Unified Modeling Language (UML) tool) as class diagrams. Each data element is assigned to one or more data class(es) within the clinical workflow. If a data element is included in the common clinical data set (i.e., common across CRNs) it will be referenced in the CRN specific model. All CRN specific data elements are modeled in the CRN specific model as a data class, class attributes or included as a value set. Use of a standardized process to define and map to existing exchange and terminology standards enables the development of clinical domain technical specifications (e.g., FHIR® Implementation Guides). The work will result in a catalogue of mimimum core data sets for the various clinical domains represented by the CRN community.
Clinically Meaningful Device Attributes
Device libraries can be used for identification of devices and their attributes for device research and surveillance within CRNs. The libraries enable focus on individual device, device categories, or device characteristics. The libraries can also enhance the capacity of the FDA’s GUDID for research and surveillance. For example, the ICOR device library is a global, standardized classification system of hip and knee implantable devices, to advance the implementation of unique device identifiers and FDA postmarket surveillance.
Data Linkage Tools
To advance linkages between registries and routinely available data sources (e.g. claims and administrative data), MDEpiNet is developing and refining anonymous linkage algorithms to augmenting research capacities of CRNs by bringing together registries, claims data, and electronic health records. Data linkage with indirect identifiers is reliable with high sensitivity and accuracy. It is a cost-effective way to obtain long-term outcomes and has positive implications for long-term device surveillance.
To incorporate natural language processing (NLP) into the building blocks of CRN. NLP is a valuable method that can be used to parse unstructured text data and extract information from unstructured text data, including medical notes, radiology reports, and device adverse event reports. For processing large amount of text data, NLP is efficient and labor saving. The Center is currently also advancing NLP methods that will help with data collections to enhance sustainability of CRNs.
Assessing adverse events related to hysteroscopic sterilization device removal using NLP: Extracting information from prostate cancer biopsy reports, magnetic resonance imaging and partial gland ablation operative reports
AI and Machine Learning
To facilitate the evaluation of patient outcomes and predictors in the context of medical devices. Machine learning methods have the advantage of taking into account complex interactions between predictors and help identify patient populations among whom the treatment works best. These information may help clinicians understand population-specific treatment effect better and assist with clinical decision making.
Blockchain is a novel technology that advances data safety, security, and reliability through the (1) assurance of immutable data provenance, (2) implementation of smart contracting, and (3) implementation of the electronic consent form and helps build trust with diverse group of stakeholders. These aforementioned aspects of blockchain enable various stakeholders to share their data within an ecosystem, which in turn can increase the velocity of research.
Mobile AppsMobile Apps
The Mobile Apps engine integrated with HIVE technology provides novel and robust means of capturing data through patient-facing and physician-facing portals. MDEpiNet Coordinating Center currently supports various HIVE projects in women’s health technologies and cancer settings. Patient and physician registry platforms are being developed to support national and international collaborations.