Notice Number: NOT-HL-19-685
Key Dates
Release Date: March 6, 2019
Response Date: May 15, 2019
NOT-HL-19-701
Issued by
National Heart, Lung, and Blood Institute (NHLBI)
Purpose
This Notice is a time-sensitive Request for Information (RFI) inviting comments and suggestions to be considered during the potential development of a challenge prize for heart failure (HF) classification systems.
Background
Adult heart failure (HF) is commonly classified for diagnosis and subsequent treatment by a single metric left ventricular ejection fraction (LVEF). However, HF is a multi-organ syndrome with multiple related but seemingly disparate phenotypes, and LVEF fails to capture this complexity. The many complex interrelated phenotypes, comorbidities, as well as lifestyle and environmental factors associated with HF outcomes make it a disorder well-suited to benefit from a big data approach. Adult HF, particularly HF with preserved ejection fraction (HFpEF), is a current priority area for NHLBI and was the focus of a 2017 NHLBI workshop Research Priorities in Heart Failure with Preserved Ejection Fraction (HFpEF) . As such, the NHLBI is exploring the feasibility of designing a potential challenge prize that would focus on utilizing computational methods with currently available data to create new phenotypic classifications for adult HF.
Section 2002 "Eureka Prize Competitions" of the 21st Century Cures Act, enacted on December 13, 2016 (P.L. 114-255), requires NIH to support and report on prize competitions in areas of biomedical science that could: 1) realize significant advancements and 2) improve health outcomes in human diseases and conditions that have a disproportionately small research investment relative to expenses for prevention and treatment, represent a serious and significant disease burden, or for which there is potential for significant return on investment.
Section 2002 prize competitions, like other NIH prize competitions, must be carried out pursuant to NIH’s existing prize authority, i.e., the America COMPETES Act (P.L. 111-358), as revised by the American Innovation and Competitiveness Act (P.L. 114-326).
Information Requested
This RFI seeks input from stakeholders throughout the scientific research community and the general public. Individuals wishing to provide feedback may respond to any number of the following four topics:
- Scientific gaps and needs to be addressed by HF classifications or subphenotyping,
- Important existing data needed to develop novel HF classification systems,
- The feasibility and structure of a challenge prize competition on HF classification systems, and
- Potential judging criteria.
TOPIC 1: HF classification or subphenotyping
HF is commonly classified by a single metric based on LVEF. Although this classification is useful in identifying patients likely to benefit from certain drug or device therapies, it fails to capture the substantial heterogeneity within a given LVEF range, especially in patients with HFpEF. Other classification systems based on severity of symptomatic limitations, peak aerobic capacity, or structural cardiac changes have been developed to address this heterogeneity but do not fully capture it. The NHLBI is interested in stimulating research to develop novel phenotyping classification schemes that would better capture the inherent complexity of the HF phenotype to more accurately target optimal treatment and predict outcomes in HF patients.
To achieve these goals, the NHLBI is seeking input from domain experts to better understand what a novel HF phenotypic classification system might look like and how different classification approaches may be best compared. We hope to receive clear recommendations that will help us better understand the current landscape of HF patient classification, the problems or concerns with current HF classification systems, what research scientists would seek from a novel HF classification scheme, and features of a new classification scheme that would be useful for clinical practice.
For individuals wishing to comment, feedback on the following issues would be particularly useful:
- Whether a potential classification scheme should focus on specific HF subphenotypes (e.g., HFpEF vs. all adult HF phenotypes)
- Characteristics of a new phenotyping classification scheme in adult HF that would be useful for basic science (e.g., biomarkers, post-translational modification of proteins) or for clinical application (e.g., easily measured lab values)
- Ways in which a new adult HF classification scheme could be applied to precision medicine approaches to HF research
- Other areas or considerations related to the development of novel classification schemes of HF
Individuals interested in providing commentary may wish to consider the recommendations that emerged from the NHLBI’s 2017 workshop Research Priorities in Heart Failure with Preserved Ejection Fraction (HFpEF) . In addition, the NHLBI Strategic Vision offers insights into potential critical challenges and compelling questions that are guiding the Institute’s future research priorities and investments.
TOPIC 2: Types of data needed to develop novel adult HF classification systems
The NHLBI has made considerable investments in the creation of deep data resources, from longitudinal, deeply phenotyped epidemiology cohorts to recent large-scale omics data generation by the Trans Omics for Precision Medicine (TOPMed) program. The Institute is also currently embarking on a project to create a cloud-based platform intended to democratize data access and computational tools and support data interoperability via Data Storage, Toolspace, Access, and analytics for biG-data Empowerment (Data STAGE). With these data and tools in hand, NHLBI is now seeking opportunities to utilize them in applying computational analysis and machine learning approaches toward its mission to prevent, evaluate, and treat heart, lung, blood, and sleep diseases.
The NHLBI is seeking input from domain experts to better understand the types of existing data and the types of currently available computational or storage resources that would be needed to develop novel HF classification schemes. We hope to receive clear recommendations that will help us ensure that potential challenge prize entrants have access to currently existing datasets and resources that would be needed to successfully develop novel HF classification systems.
For individuals wishing to comment, feedback on the following topics would be especially welcomed:
- Types of existing data that would be needed to develop a new HF phenotypic classification (e.g., genome sequence, imaging, biomarkers, omics, clinical trials, longitudinal phenotyping)
- Types of computational workspaces or analytical routines needed when developing a new phenotype classification scheme
- Whether NHLBI’s existing datasets are sufficiently comprehensive to support the development of novel HF classification schemes or whether additional extant datasets (e.g., electronic health records) would be needed
- For individuals having experience working with NHLBI data, hurdles to data access and potential solutions to those problems
- Other areas or considerations related to the available datasets and resources that potential participants might need to successfully develop novel HF classification schemes
Ideas may be submitted in any area of big data analytics, bioinformatics, biostatistics, data access, computational biology, and cloud computing.
TOPIC 3: The feasibility and structure of a challenge prize competition
The NHLBI is contemplating whether the development of novel classification schemes for HF is appropriate for a challenge prize competition. We hope to receive clear recommendations that would help us create a feasible, equitable, and successful challenge prize. For more information on the challenge prize format and currently active challenge prize competitions, commenters may wish to visit challenge.gov.
For individuals wishing to comment on the feasibility of a challenge prize, feedback on the following issues would be particularly useful:
- Clarity of the goal (see Topic 1) and its suitability for a prize model (as opposed to a grant, contract, or other mechanism)
- Pros and cons of using a pre-determined dataset (vs. open data) for developing a potential new phenotyping scheme for HF
- Currently available methods to facilitate equal access to NHLBI datasets for all potential participants
- Attractiveness of the question to a broad audience of entrants and ways in which the NHLBI can advertise the challenge prize to solicit entries from the general public, especially to bioinformaticians, programmers, computer scientists, and others with an interest in big data analytics
- Methodology for assessing overall success of the challenge
- Other ideas or considerations concerning the feasibility of using a challenge competition to foster equitable competition to develop novel systems for HF classification and the requirements of such an approach. Ideas that are amenable to small business participation are also welcome.
The NHLBI is considering ways to best structure a potential challenge prize, and hopes to receive clear recommendations that would help structure the competition phase of a challenge prize appropriately so that participants have adequate time to develop and test novel HF classification systems (see Topic 1).
For individuals wishing to comment, feedback on the following issues would be especially welcomed:
- Whether a competition would be better suited for a multi-phased approach (e.g., proposal and research phases, see challenge.gov for examples) or for a single phase, and the appropriate length of time for the phase(s)
- Appropriate milestones within the challenge (e.g., obtaining requisite IRB/data access approvals, submitting a statistical analysis plan, preliminary/pilot results), reasonable timeframes for completing these milestones, and whether additional points should be given to entrants who complete milestones on time (see Topic 4 below)
- Other suggestions for how a challenge could be structured to best facilitate fair and open competition and to best allow for entrants to be successful
TOPIC 4: Potential judging criteria
The NHLBI is seeking input on how different HF classification approaches may be best compared. We hope to receive clear recommendations that will help us create consistent, realistic, fair, and easily understood and applied judging criteria for a potential challenge prize.
For individuals wishing to comment, feedback on the following issues would be particularly useful:
- Ways in which the clinical utility and/or basic research worth of a new HF classification scheme could be evaluated and measured
- Examples of metrics that judges might use to identify a winner, and whether certain metrics should be weighted more than others during judging
- Ways to ensure that models, algorithms, machine learning/artificial intelligence, and statistical approaches used to generate novel HF classifications are rigorous and reproducible
- Ways that a challenge prize competition could use the success or accuracy of a novel classification scheme as a potential judging criterion
- Whether the accuracy or performance of all submitted classification schemes should be compared by applying them to a single dataset (e.g., a gold standard ), and if so, whether that dataset should be made publicly available to entrants during the competition phase of the challenge
- Ways to resolve potential scoring ties between multiple solutions so that a single winner could be identified
- If milestones are used during the competition phase of the challenge prize, whether timely completion/achievement of the milestones should be considered during judging (see Topic 3)
- Other areas or considerations related to the development of judging criteria or other ways to identify a winner
How to Submit a Response
All responses to this RFI must be submitted electronically to the following webpage at https://grants.nih.gov/grants/rfi/rfi.cfm?ID=88 by May 15, 2019. The suggested response length is up to 1000 words.
Responses to this RFI are voluntary. Do not include any proprietary, classified, confidential, trade secret, or sensitive information in your response. The responses will be reviewed by NIH staff, and individual feedback will not be provided to any responder. The Government will use the information submitted in response to this RFI at its discretion. The Government reserves the right to use any submitted information on public NIH websites, in reports, in summaries of the state of the science, in any possible resultant solicitation(s), grant(s), or cooperative agreement(s), or in the development of future funding opportunity announcements.
This RFI is for information and planning purposes only and shall not be construed as a solicitation, grant, or cooperative agreement, or as an obligation on the part of the Federal Government, the NIH, or individual NIH Institutes and Centers to provide support for any ideas identified in response to it. The Government will not pay for the preparation of any information submitted or for the Government’s use of such information. No basis for claims against the U.S. Government shall arise as a result of a response to this request for information or from the Government’s use of such information.
We look forward to your input and hope that you will share this RFI document with your colleagues.
Inquiries
Please direct all inquiries to:
Rebecca Beer, PhD
Program Director
Division of Cardiovascular Sciences
National Heart, Lung, and Blood Institute (NHLBI)
Email: NHLBIChallengeInput@nhlbi.nih.gov
Laura Hsu, DrPH
Challenge Manager
Division of Extramural Research Activities
National Heart, Lung, and Blood Institute (NHLBI)
Email: NHLBIChallengeInput@nhlbi.nih.gov
and Human Services (HHS)
