The National Library of Medicine seeks applications for novel data science/informatics approaches that can help individuals gather, manage and use data and information about their personal health.
Increasingly, consumers and patients have access to a broad and complex array of personal health information that is relevant to the state of their health. Health-related information can come from diverse sources, such as mass media and social networks, health care organizations, government agencies, clinicians, family members and friends. Health-related information also comes in many different formats, such as data from an individual’s electronic medical record, family histories and genealogies, data streams from activity trackers, personal genome sequences, articles, videos about diseases and treatments, and public research data sets. It is well known that, while patients discuss personal health decisions and health information with the clinicians from whom they receive care, they also seek health information from other sources that are increasingly digital, and are constantly changing, enriched with new streams of data and new types of data. National biomedical research initiatives are emerging, such as the Precision Medicine Initiative Cohort Program (“All of Us SM” https://www.nih.gov/research-training/allofus-research-program) and the Million Veteran Program (http://www.research.va.gov/MVP/ ), that invite citizens to share personal digital health data and biospecimens with researchers. There are also collaborative initiatives such as Patients Like Me © (https://www.patientslikeme.com/) wherein patients contribute personal health data for citizen science projects.
Whether a person wants to participate in one of these research initiatives, or just play an active role in staying healthy, she/he faces a challenge in deciding what personal health information is important enough to collect, store, manage and share. A patient with a chronic condition might want an easy way to monitor his/her progress between visits to a clinic. A parent caring for a family member with a debilitating condition might want to stay current about treatment options and clinical trials in that area. Either might be faced with test results or treatment options that they don’t understand. Someone who participated in a clinical trial might want to annotate records of his/her own results, to look for patterns. These examples are meant to highlight the scope and scale of the health information-related activities that an individual might wish to undertake. The individual must decide what kinds of data and information should be kept, and bring them together such that the combination serves as a trustworthy, usable, useful library of personal health information.
Vast stores of digital data about health and disease, genetics, environment and behavior are the basis of keen scientific interest about what new insights into health and disease can be gained by combining and 'mining' these datasets together. The phrase 'big data' is often used to characterize this rich, multi-dimensional, heterogenous research resource. Much of the early research in big data methods and approaches for health has centered on needs of researchers in biomedical, behavioral and social science fields. For example, research supported by the National Institutes of Health through its Big Data to Knowledge Initiative (BD2K) is developing novel methods to find and visualize trends in very large data sets, using computational intelligence to integrate data from disparate sources, matching genetic defects to health manifestations, designing tools that protect privacy while using person-level data, and developing techniques to bring together all of the health information about a single patient, to name just a few areas (https://datascience.nih.gov/bd2k/funded-programs). Many of the use cases for these approaches mirror information problems an individual faces when trying to gather all of her health information from all possible sources and use it for improving and maintaining her health.
To bring the benefits of big data research to consumers and patients, new biomedical informatics and data science approaches are needed, shaped to meet the needs of consumers and patients, whose health literacy, language skills, technical sophistication, education and cultural traditions affect how they find, understand and use personal health information. Novel data science approaches are needed to help individuals at every step, from harvesting to storing to using data and information in a personal health library. Areas of development suggested below are not meant to limit the scope or creativity of proposed projects.
- Constructing a personal health library: informatics approaches that help a person gather together different types of health data/information/knowledge into a single, searchable resource for personal use, including intelligent mapping tools for vocabulary used to describe elements of the library.
- Managing a personal health information library: novel informatics approaches that make it easy for an average user to expand or remove entries, make notes or corrections, including intelligent tools that alert the user to new information about topics covered in a personal health information library.
- Using a personal health library: data science and informatics approaches that make it easy to find and use the information stored there, including visual tagging, text summarization, graphics translation, knowledge mapping, suggestions for tutorials, analytic and visualization techniques that make the information understandable based on characteristics of the individual user or group.
- Digital librarian/assistant for personal health library: data science and informatics approaches that bring machine intelligence to the management and use of a personal health information library through personalized alerts and suggestions, literacy aids, translators or other approaches, taking into account characteristics of the individual user or group.
Applicants must base their proposed work on an informed profile of the intended users, and, the work should be developed through interaction with the user. The strongest projects will provide approaches that incorporate health data and information from more than one source, such as diagnostic images and links to full-text articles or genome sequence data linked to a family health history. An application should be centered on the problem area being addressed and the intended audience, propose a possible solution that employs novel data science or informatics, and undertake a pilot that will result in evidence of the degree of success and/or needed next steps. Applicants should expect to involve the intended users in their work.
Applicants may propose new tools or extensions to the capabilities of existing open source tools such as personal health record systems, by adding new features or extending capabilities of the tool. In either case, scientific innovation is key. Applicants are encouraged to take advantage of freely available public information resources available from NLM and others, such as MedLINEPlus, Genetics Home Reference, PUBMED Central, Online courses and tutorials.
Applicants should plan to undertake one or more pilots to test their ideas with the intended user group. If pilots focus on a single disease or health condition, applicants should provide assurance that their approach is generalizable to others. Awardees are expected to share the results of their work through publication, and through open source mechanisms for data or resource sharing. The plan for data/resource sharing will be discussed during the initial scientific merit review.
Projects that propose the following outcomes would not be appropriate for this FOA:
- A tool that requires the user to purchase a commercial off-the-shelf product.
- An information resource that requires payment for access to information.
- A tool that supports management of only a single kind of health data or information.
- An approach that does not allow the user to expand or update the contents.
- An approach that doesn't allow the user to update, annotate or add/delete data.
- An approach that limits the user's ability to share information from her/his personal health library with another person or organization.
Potential applicants are urged to discuss their proposed project with the Research Contact listed in Section VII Agency Contacts, for advice about the suitability of their idea for this funding initiative.
CU Internal Expression of Interest Deadline: N/A; email firstname.lastname@example.org if interested
NIH Letter of Intent Deadline (Optional): February 16, 2018
NIH Application Deadline: 5:00pm MST March 19, 2018
Any individual(s) with the skills, knowledge, and resources necessary to carry out the proposed research as the Program Director(s)/Principal Investigator(s) (PD(s)/PI(s)) is invited to work with his/her organization to develop an application for support. Individuals from underrepresented racial and ethnic groups as well as individuals with disabilities are always encouraged to apply for NIH support.
Limited Submission Guidelines
Only one application per institution (normally identified by having a unique DUNS number or NIH IPF number) is allowed.
Award Information and Duration
Up to $250,000 per year for up to 4 years.