Showing 17 ideas for tag "data"

Goal 4: Develop Workforce and Resources

Quantitative training in the era of big data

Should training in biostatistics, computer science, bioinformatics become broader for the entire biomedical community in this era of very large data sets?

 

How can we grow the number of specialists in biostatistics (and these related fields) without causing an oversupply?

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Compelling Question (CQ)

Details on the impact of addressing this CQ or CC

Broad impact over all of biomedical science.

Feasibility and challenges of addressing this CQ or CC

Training could be enhanced within the NHLBI, or externally through increased funding for appropriate programs in quantitative methods for biomedical scientists and increased funding for graduate training in biostatistics (and related quantitative areas).

Challenges: Poor training and disinterest in the mathematical sciences among many medical scientists. Fewer biostatisticians (etc.) than needed to fill current positions.

Name of idea submitter and other team members who worked on this idea NHLBI Staff

Voting

40 net votes
55 up votes
15 down votes
Active

Goal 2: Reduce Human Disease

Moonshot: Turning the BMT EMR into a Research Record

The critical challenge is to develop a standards-based BMT electronic medical record (EMR) and integrate research capacity into the architecture of EMR systems. The ultimate goal would be to build de-identified complete data-sets which can be used to support observational studies and clinical trials, improve transplant outcomes and inform public policy.

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

Clinical research is constrained by a clumsy method of acquiring biomedical data, generally relying on manual capture of information from EMR back-to-paper which is then transcribed into registry or specific clinical trial databases. This method is labor intensive, fraught with opportunities for error, and increasingly difficult to defend in light of the high costs associated with clinical trials. Adoption of standards-based clinical documentation and creating access to source clinical data would reduce or do-away with resource-intensive, very expensive and time-consuming data abstraction, enhance data quality and depth, and accelerate translational research.

Feasibility and challenges of addressing this CQ or CC

As BMT centers increasingly adopt EMR systems in the United States, a vast and potentially very useful data resource is being created. However, most EMR systems offer very generic formats for clinical documentation and the medical information is inconsistently expressed in vocabulary, structure, and format. One challenge is the development of common standards-based clinical documentation format and its adoption by EMR system vendors and BMT institutions to support structured data sharing.

Large transplant centers can build their own integration engines to link EMR with stem cell lab, HLA, donor care, workflow etc. However, a broad implementation of integration IT solutions would be needed amongst centers conducting BMT clinical trials.

While CIBMTR's FormsNet application and Clinical Trials Network allow electronic data submission, data professionals still need to manually enter the data. Another challenge in creating a centralized data resource would be to build interoperability between transplant centers and research entities. An alternative to a large centralized database could be a distributed research network which allows data holders to maintain logical and physical control over their data and mitigate security, proprietary, legal, and privacy concerns.

Name of idea submitter and other team members who worked on this idea Rakesh Goyal

Voting

57 net votes
71 up votes
14 down votes
Active

Goal 3: Advance Translational Research

Maximizing Previous Investment in Existing Cohorts

Everyone would like to see integration of genomic, metabolomic, epigenomic, proteomic, transcriptomic, etc. data analyzed in the context of clinical disease, environmental influences, and even end-organ effects (lung versus heart or blood as an example). Rarely can this occur on small cohorts, but rarely are funds available to take maximum use of existing large cohorts and the samples and information collected within... more »

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

The impact would be huge as it would leverage already extremely expensive cohorts to maximum potential, allowing for exploration into clinical subphenotyping, disease mechanisms, personalized medicine, surrogate endpoints, biomarker exploration, etc. Maximizine output on previous investment is the clearest impact, since even simple analysis in a large number of samples adds up to a very hefty sum. Additionally, data from samples becomes more valuable with longitudinal follow-up of available subjects.

Feasibility and challenges of addressing this CQ or CC

The challenges include the expense of analysis in large cohorts and the ability to attract and fund high level biostatistical faculty at top-notch institutions and get them engaged fully in the problem. Biostatisticians of high caliber will not engage without funding and without an ability to “train” students using the data and explore their own research interests within the context of the overall clinical problem. Funding mechanisms that are large (to allow for deep phenotyping of cohort samples on multiple platforms in multiple sample types) and that seek to generate solid and ongoing collaborations between the data generators and the data analyzers must emerge.

Name of idea submitter and other team members who worked on this idea Wanda K. O’Neal, PhD

Voting

22 net votes
34 up votes
12 down votes
Active

Goal 2: Reduce Human Disease

Data integration in congenital heart disease research

Lack of data integration in the congenital heart disease research community is inhibiting low cost, research opportunities.

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

Improve our ability to leverage big data. We could accumulate larger sample sizes to ask/answer important epidemiology or comparative effectiveness questions in the congenital heart disease population.

Feasibility and challenges of addressing this CQ or CC

Bioinformatics, data science, and information technology capabilities are reaching a point to support such large scale data integration efforts.

Name of idea submitter and other team members who worked on this idea NHLBI Staff

Voting

13 net votes
19 up votes
6 down votes
Active

Goal 4: Develop Workforce and Resources

Credible Data and Analysis of the Biomedical Research Workforce

There is a need for sensible policies that require collection and scientific analysis of credible data relating to the biomedical workforce. The data currently available are weak – for example no one knows, to a factor of 2X, the actual number of postdocs in the United States. The absence of credible human resource and labor market data on the biomedical research workforce is very surprising. NIH could contribute greatly... more »

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Feasibility and challenges of addressing this CQ or CC

NIH has begun to develop its own capacity for such data collection and analysis, a very positive step. In addition, NIH may wish to consider modest research grant funding for research on the biomedical workforce by academic labor economists.

Name of idea submitter and other team members who worked on this idea Michael S. Teitelbaum

Voting

6 net votes
15 up votes
9 down votes
Active

Goal 3: Advance Translational Research

Build a National Surveillance of Chronic CV and Lung Diseases

There is a need to build a robust coordinated surveillance system on the incidence and prevalence of chronic diseases. Surveillance data are needed to:

•Describe and monitor the burden, trends, and patterns of these diseases

•Set parameters and metrics of research priorities

•Identify where to target resources for prevention, treatment, and delivery of care

•Track and monitor progress toward public health disease... more »

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

The high prevalence of chronic cardiovascular and lung diseases has created burden in increasing healthcare costs and high mortality rates in the US compared to other developed countries. Even so, they remain among the most preventable health problems. A national surveillance system for chronic cardiovascular and lung diseases would enable data-driven decision-making about public health strategies for prevention, management, and cost containment.

Feasibility and challenges of addressing this CQ or CC

A 2011 Institute of Medicine (IOM) report concluded that a coordinated surveillance system is needed. It proposed a framework for such a system that would integrate existing information through collective efforts of multiple stakeholders. The time is right to gain from and build upon numerous ongoing broad initiatives in biomedical Big Data, including growing health IT adoption mandated by the HITECH Act, ONCHIT efforts to achieve health IT interoperability, the NIH BD2K initiative, and the multiorganizational network participating in FDA Mini-Sentinel, HCS Collaboratory, and PCORnet, among others. The NHLBI is well-positioned to lead, develop and implement the IOM’s recommended framework and system. (IOM report - http://www.iom.edu/Reports/2011/A-Nationwide-Framework-for-Surveillance-of-Cardiovascular-and-Chronic-Lung-Diseases.aspx))
Existing data sources (i.e., population surveys, registries, cohort studies, administrative data, and vital statistics) do not individually provide nationally representative data, cannot be linked, and are not currently readily accessible to all levels of users. One potential way to build such a system is to integrate and expand existing data sources.

Name of idea submitter and other team members who worked on this idea NHLBI Staff

Voting

5 net votes
13 up votes
8 down votes
Active

Goal 3: Advance Translational Research

Embedding Clinical Trials in Learning Health Systems

What are the best methods for using genotype information and other EMR data to randomize heart, lung, blood, sleep patients to different treatment strategies? One big challenge is how to consent patients for this sort of trial. Must patients be consented separately for every such trial or could there be blanket consent for participating in the learning health care model? This would also require a paradigm shift in how... more »

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Compelling Question (CQ)

Details on the impact of addressing this CQ or CC

If successful this approach should enable the conduct of cheap pragmatic trials that are fueled by data from clinical care. The integration into clinical care helps assure efficiency and generalizability of results.

Feasibility and challenges of addressing this CQ or CC

The advent of electronic medical records and the explosion of big data technology has made it possible to gain access to and analyze data in a manner that would have been unthinkable 10 years ago. This is already going on in other fields.
Health care systems are increasingly using "big data" approaches to track outcomes in the patients treated with different strategies and drugs, and apply the knowledge gained from outcomes in previous patients to inform decision making in subsequent patients ("learning"). This approach could be used to personalize treatment. A recent example from cancer is to genotype lung tumors, and tailor the treatment of a new patients to drugs producing good results in patients with similar tumor genotypes. When two or more treatments produce similar results, one could randomize. Cardiovascular disease presents a challenge in using genotyping information to personalize treatment, because the manifestations are the results of complex genetic and environmental risk factors.

Name of idea submitter and other team members who worked on this idea NHLBI Staff

Voting

4 net votes
15 up votes
11 down votes
Active

Goal 1: Promote Human Health

Predictive analytics to engage healthy behaviors and maintain health while reducing cost

Predictive Health employs the principle that using modern health testing and predictive analytics will better define true health (not just absence of disease) and, in combination with large-scale data analytics, will facilitate predicting deviations from the healthy trajectory earlier than traditional disease diagnosis, thus allowing more effective and less costly interventions to maintain health. Predictive Health educates... more »

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

By addressing this CC the health of the nation will be improved: better national and individual understanding of health, greater longevity of sustained health and productivity, reduced costs of healthcare by focusing on health than on disease diagnosis and management.

Feasibility and challenges of addressing this CQ or CC

The Emory/Georgia Tech Predictive Health Institute (http://predictivehealth.emory.edu) was founded ~10 years ago by launching educational (http://predictivehealth.emory.edu/education/index.html) and scientific (http://predictivehealth.emory.edu/chd/index.html) programs to achieve the Predictive Health goals. The scientific approach created a prospective longitudinal cohort of individuals who have been richly phenotyped according to traditional medical testing (clinical laboratory, stress testing, etc) and research testing (genomics, metabolomics, regenerative cell potential, oxidative stress) to create the deepest understanding of current and future health. The success of the Predictive Health Institute demonstrates the feasibility and potential success of such a critical challenge to both human health and healthcare expenditures.

Name of idea submitter and other team members who worked on this idea Greg Martin, for the Emory/Georgia Tech Predictive Health Institute

Voting

5 net votes
9 up votes
4 down votes
Active

Goal 3: Advance Translational Research

Create a National Action Plan for COPD

Lead a coordinated effort of government, patient advocacy organizations, professional organizations, payers and others to plan and implement a coordinated plan to improve COPD awareness, education for patients and healthcare professionals, treatment strategies, research and data collection, policies and public health infrastructure and programs.

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

Unlike most leading causes of death and disability, there is no coordinated effort to lower disease burden associated with COPD. Coordinated plans provide a forum for identifying the most pressing issues that must be tackled, for setting goals and convening partners from different disciplines and are a framework upon which policy change can be achieved. In order to make meaningful progress in the impact that COPD is having on patients, health systems and payers, coordinated planning and action is needed and NHLBI can lead the way but time is of the essence.

Feasibility and challenges of addressing this CQ or CC

There are proven models of multi-stakeholder, public and private partnerships to tackle disease burden and create national plans. There are also multiple national and regional organizations standing ready to assist in these efforts.

Name of idea submitter and other team members who worked on this idea COPD Foundation Board of Directors, COPDF MASAC, COPDF State Advocacy Captains

Voting

6 net votes
6 up votes
0 down votes
Active

Goal 3: Advance Translational Research

Leveraging big data for T4 translation research

What approaches can help leverage the emerging big data in health and health care for observational and interventional implementation research in heart, lung, blood, sleep diseases?

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Compelling Question (CQ)

Details on the impact of addressing this CQ or CC

• Integration of big data analytics into T4 research study design and interventions development
• Innovative linkages across multiple health and non-health sector data
• Innovative methods to analyze big data linked across sectors
• Various communities are using big data analytics to understand population health data (e.g. electronic medical records s) and opportunities exist for consolidation of these efforts and standardization of methodologies

Feasibility and challenges of addressing this CQ or CC

• NIH now has focus on big data in its formative stages
• Significant amount of NIH’s budget is/will be dedicated to big data research
• NHLBI can leverage NIH’s investment by foster research in D&I big data analytics and systems science
• Future investment in big data should yield opportunities and focus efforts

Name of idea submitter and other team members who worked on this idea NHLBI Staff

Voting

0 net votes
16 up votes
16 down votes
Active

Goal 1: Promote Human Health

leveraging EHR through improved partnerships

Large medical delivery systems have an abundance of information stored in their EHR systems. What steps are necessary to develop partnerships between NHLBI and medical delivery systems or other agencies such as AHRQ to access these EHR systems, develop a common terminology (if icd 9 and other common codes are ineffective for merging data) and add that data to dbGap for NIH/community use?

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Compelling Question (CQ)

Details on the impact of addressing this CQ or CC

Accessing the EHR systems of the large care systems would provide a wealth of knowledge/data for the research community.

Feasibility and challenges of addressing this CQ or CC

The stored data will only expand in the EHR systems and it would certainly be of value and interest to any number of research projects.

Name of idea submitter and other team members who worked on this idea NHLBI Staff

Voting

1 net vote
13 up votes
12 down votes
Active

Goal 3: Advance Translational Research

Advancing translational research requires timely in-depth analysis of large datasets

1. NHLBI investments over the last decade in terms of genomic approaches have yielded many research findings.
2. Rapid analyses of early data identified the "low hanging" fruit (and perhaps some/many important results were missed); this limited the scope of translation (partly because of relatively limited discovery?)
3. Important data are being generated at much greater rate than the data are processed/analyzed thoughtfully.... more »

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

If we invest in a few multi-disciplinary data analysis centers (integrating biology with statistical genetics) involving active NHLBI participation (cooperative agreements), the massive amounts of data generated at huge cost could be analyzed more thoughtfully if time and resources are made available. This way, we may be able to identify many more research findings of much greater potential for translation.

Feasibility and challenges of addressing this CQ or CC

It is feasible to pursue this challenge through cooperative agreements whereby NHLBI scientists can actively participate and ensure that strategic investing is following strategic paths for deep discovery.

Name of idea submitter and other team members who worked on this idea DC Rao

Voting

1 net vote
11 up votes
10 down votes
Active

Goal 4: Develop Workforce and Resources

Leveraging big data for T1 translational research

How best to train T1 investigators in using big health care data to test their basic science hypotheses related to heart, lung, blood, and sleep disorders and thus generate sufficient confirmation to justify clinical interventions.

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Details on the impact of addressing this CQ or CC

• Hypotheses that emerge from small studies of patients focused on mechanistic questions often are not able to be tested at the population level and thus remain unconfirmed. The increasing availability of big data from EHRs permits corroboration at the population level, but requires skills in framing queries and minimizing bias and confounders.
• The emergence of large clinical data sets such as PCORNet and the NIH Collaboratory make this particularly timely.
• By sampling existing data sets rather than having to conduct new randomized studies, this type of research can be performed at relatively low cost and in a much more timely way.

T1 investigators are not usually trained in the techniques of using large clinical data sets and so require targeted training.

Feasibility and challenges of addressing this CQ or CC

Developing educational programs for T1 investigators should be straightforward. Resources will be required to support the queries by T1 investigators required as part of the training, but this should be modest compared to other forms of research.

Name of idea submitter and other team members who worked on this idea Barry Coller

Voting

2 net votes
3 up votes
1 down votes
Active

Goal 2: Reduce Human Disease

Data from regulatory studies a barrier to evidence-based medicine

Alignment of regulatory, healthcare, and research arms of the government is poor. There is a need to improve the design, quality and usefulness of data from regulatory studies to address major clinical questions and also to facilitate scientific inquiry. This is a barrier to evidence based medicine and improved treatments.

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Critical Challenge (CC)

Name of idea submitter and other team members who worked on this idea Society for Vascular Surgery

Voting

2 net votes
3 up votes
1 down votes
Active

Goal 2: Reduce Human Disease

Harnessing the Tsunami of Patient-Generated Health Data

How can the NHLBI foster the development of effective tools and methodologies to harness the tsunami of patient-generated health data into a valuable resource for conducting patient-centered research? Some challenges to overcome might include a) how to enable the collaboration of the more traditional clinical scientists with scientists from other disciplines such as informatics or computational and data-enabled science... more »

Is this idea a Compelling Question (CQ) or Critical Challenge (CC)? Compelling Question (CQ)

Details on the impact of addressing this CQ or CC

Methodological research results in this area could be translated into clinical settings – specifically, they could help physicians interpret such data without being overwhelmed by their sheer volume, and even further use them as helpful adjunct tools to more traditional ways of diagnosing and/or treating diseases.

Feasibility and challenges of addressing this CQ or CC

The increasing number of smartphones, mobile apps, and remote monitoring devices are producing a vast amount of patient-generated health-related data. However, there are no widely established tools, methodologies, or strategies to ensure optimal use and management of these data. The time is right to move forward quickly in furthering research in this field over the next 5 to 10 years.

Name of idea submitter and other team members who worked on this idea NHLBI Staff

Voting

-1 net votes
10 up votes
11 down votes
Active