To improve the evidence base and quality of care, HHS is investing hundreds of millions of dollars in medication surveillance and comparative effectiveness research that relies on observational cohorts. Because the voice of the patient is largely missing in these data sources, key efficacy endpoints and quality of life metrics may not factor into clinical and policy recommendations. Also absent is key information that patients know, can accurately report, and are willing to share. Capturing data about, for example, over-the-counter and complementary medications, disease endpoints, and adverse events below threshold for a typical healthcare encounter, may greatly enhance knowledge of disease progression and best practices. We seek to develop targeted HIT and computational approaches to add the voice of the consented patient to cohort research. We build on two widely deployed open source technologies-the i2b2 analytic platform and the Indivo open source personally controlled health record. We create an infrastructure supporting a self-perpetuating cycle where shared information for research drives improved care, which motivates sustained engagement and further consented sharing, thus, further improving research and care. We call this model "The Learning Cohort" (TLC) because participants are given a toolkit providing them direct access to query registry data - a toolkit which safely constrains a patient's query capability to validated aggregate measures and offers a consumer-oriented presentation of preprocessed, summarized live registry data germane to personalized health choices. The setting is the Childhood Arthritis and Rheumatology Research Alliance (CARRA) and its scalable, 60 site patient registry which has access to half of all US children with rheumatic diseases. The project creates a translation-to-practice pathway for standardized measures including those from the NIH Patient Reported Outcomes Measurement Information System (PROMIS) to capture patient reports for the registry. Further, we provide patients with a PCHR "app" enabling them to consent to join a distributed cohort, which automatically and anonymously shares data and observations with public health authorities for postmarketing surveillance. All software code will be made open source through www.INDIVOhealth.org and www.i2b2.org. This project leverages technology produced under an Office of the National Coordinator of Health Information Technology- funded Strategic Health IT Advanced Research Projects (SHARP) project (www.SMArtPlatforms.org). SMART, which provides a common application programming interface enabling HIT platforms to run substitutable apps, allows us to create a pathway for wide diffusion of the software technology produced under this proposal.
The Institute of Medicine has called for a learning healthcare system that generates and applies the best evidence for the collaborative healthcare choices of each patient and provider;to drive the process of discovery as a natural outgrowth of patient care;and to ensure innovation, quality, safety, and value in health care. In response, we propose The Learning Cohort, which promotes a self-perpetuating cycle where individuals participate and consent to report and share information for research. The research data are used to improve and results are shared back with patients. This engagement, in turn, motivates sustained involvement and further consented sharing.
|Mandl, Kenneth D; Kohane, Isaac S; McFadden, Douglas et al. (2014) Scalable Collaborative Infrastructure for a Learning Healthcare System (SCILHS): architecture. J Am Med Inform Assoc 21:615-20|
|Natter, Marc D; Quan, Justin; Ortiz, David M et al. (2013) An i2b2-based, generalizable, open source, self-scaling chronic disease registry. J Am Med Inform Assoc 20:172-9|