Our Continuously Learning Health System
In this forum, I have often reflected on the following general guiding principles underlying the Strategic Plan for UF&Shands:
- The patient comes first.
- The faculty-hospital mindset of “we” and “they” is replaced by “us.”
- We advance Forward Together.
Our promise is to provide the best possible care for our patients, now and in the future. As the state’s most comprehensive integrated academic health center, a fundamental part of our promise is to learn from the care we provide and to contribute new knowledge that can help our community and our country meet the health challenges of today and tomorrow.
In this issue of On the Same Page, I will share how the concept of “us” is expanding to include our patients as participants and partners in our promise to improve health, contributing to what has become known as a “continuously learning health system.” In this spirit, we will soon launch an exciting and transformational research initiative that will eventually touch all areas of our health care enterprise.
Called UF Consent2Share, this initiative provides patients at UF&Shands an opportunity to help advance health research in one of two ways: 1) by allowing UF to store blood or tissue leftover from their health care visits for future use in research; and/or 2) by allowing UF researchers to contact them about future health studies for which they might be eligible. UF Consent2Share now allows researchers to link the consent process with available health care information and biospecimens.
With appropriate IRB approvals, researchers can study and learn from patients who consent to share their biospecimens and health information for research through the UF Clinical and Translational Science Institute (CTSI) Biorepository and the UF&Shands Integrated Data Repository (IDR), which have extensive safeguards in place to protect patient privacy.
These three efforts – UF Consent2Share, the CTSI Biorepository and the IDR – are collaborations that reach across our clinical, education and research enterprises. They serve as three major pillars of an exciting and evolving initiative championed by the CTSI: UF Health Research Connection, which aims to expand access to and linkages between institutional resources that bridge health research and patient care.
Nationally, one of the long-entrenched challenges of the research community has been to expand participation of diverse populations in research studies, which in turn speeds the progress, significance and relevance of research in improving health and health care. Only about 2% of the U.S. population participates in clinical research studies, and, until recently, many studies have shown that participation rates in the United States were quite low among under-represented populations, especially African Americans. With the efforts of Clinical and Translational Science Award institutions across the U.S., however, the community’s trust in the research enterprise has been growing. Led by UF's Linda Cottler, Ph.D., professor and chair of epidemiology, the Sentinel Network study by five CTSAs has shown that among 6,000 community members, African Americans report more willingness to participate in research than respondents of other racial/ethnic groups, even when blood samples or genetic materials would be collected.
In 2011, faculty at UF&Shands saw nearly 1.5 million patients in our hospitals and practices. On average, UF investigators launch approximately 1,000 clinical research studies each year, and approximately 400 of those studies are seeking participants at any given time. UF Consent2Share will help to bridge the patient-research divide in our community, and has the potential to serve as a national model for how academic research and health-care systems can work together to expand access to and strengthen the impact of health research.
What follows is a brief report on the successful pilot launch of UF Consent2Share, as well as an overview of how it will enable research through the IDR and CTSI Biorepository. Two individuals have been a unifying force across all three of these efforts: Peter Iafrate, Pharm.D., chair of the UF Institutional Review Boards, and Gigi Lipori, senior director, planning and analysis, Shands at UF. Their tireless dedication to developing ethically sound processes that protect patients and enable the secure flow of data between these resources is to be commended. An important nod also goes to Kari Cassel, UF&Shands Chief Information Officer, and Mike Conlon, Ph.D., CTSI associate director and chief operating officer, whose leadership and expertise have been instrumental.
Through this initiative, all eligible patients will have the opportunity to allow UF researchers to 1) store blood or tissue, collected during their health care visits, for future use in research, and/or 2) contact them in the future about research studies for which they might be eligible.
Here’s how the process works: At check-in, trained admissions staff briefly discuss UF Consent2Share with patients, who then receive the research consent form to review. Saying yes to one or both options is completely voluntary and is not required to receive care. Patients are offered a copy of the consent form to keep, and their responses are recorded and managed as part of their Epic electronic medical records. For patients who have questions about or wish to change their consent, a helpline and CTSI research subject advocate are available as resources.
As in all of our clinical and research efforts, the protection of patient privacy and personal health information is of paramount importance. For patients who choose to participate, their related medical information will be stored on a password-protected and encrypted computer server and will only be available to researchers conducting IRB-approved studies. Access to data will be expertly managed by the IDR and CTSI Biorepository staffs.
But can a process like this really work? Internal Medicine & Medical Specialties Clinic Manager Pat Meek and her team stepped up and agreed to help us find out by piloting this process in the liver clinic at Shands Medical Plaza. With support from Peter Iafrate, Gigi Lipori and UF&Shands Information Technology, the IMMS team launched the pilot on Sept. 11. After just over a month, the team was ready to expand the pilot to all 12 divisions practicing at the IMMS clinic, which sees approximately 225 patients per day.
As of November 27, 3,247 patients have completed the consent forms, with approximately 79% (2,547) consenting to be contacted about future research studies and approximately 85% (2,731) consenting to allow UF to store their biospecimens for future research. This level of participation is extraordinary and suggests enormous potential for this initiative.
Katie Adams, supervisor and financial support analyst, and Megan O’Connor, office representative, took the lead for IMMS and served as the point persons for the pilot. Contrary to the team’s initial expectations, most patients are returning the consent forms at the check-in desk, as opposed to waiting to talk to their physicians. In reflecting on the pilot process, the clinic team reported that they did not find the process overly burdensome from an administrative standpoint and were surprised at the positive reactions to it among patients.
“No one thought we’d be able to make such an impact at the front desk,” Pat Meek observed. “Patients are so excited about it and the staff is very engaged.”
A phased roll-out will begin in early 2013; department chairs will soon receive a communication seeking clinical services that would like to volunteer to roll this out to their patients. Meanwhile, as part of the pilot, the consent team is monitoring and studying the process to ensure that the consent is being administered in an appropriate manner.
Integrated Data Repository (IDR)
The IDR is a large-scale database that collects and organizes information from across the academic health center to support improved care and research. It is built on a platform called “i2b2” (Informatics for Integrating Biology and the Bedside), which is an NIH-funded National Center for Biomedical Computing based at Partners HealthCare System. The IDR first became operational in 2011, but is now expanding its capabilities.
The IDR consists of a secure, clinical data warehouse (CDW) that aggregates data from the various clinical and administrative information systems, including the Epic electronic medical record system. The CDW contains demographics, diagnoses, procedures, lab results, personalized medicine genotyping data, health care visit details, biospecimen availability – and now, thanks to UF Consent2Share, information about whether patients have agreed to be contacted about research studies and/or to have their leftover samples collected and stored for future research.
While the IDR’s CDW cannot be directly accessed for research, a HIPAA-compliant “Limited Data Set” that replicates data from the CDW (with identifiers removed) has been approved for faculty researchers to query for the purposes of identifying cohort counts as they prepare grant proposals, plan clinical trials, write IRB protocols and conduct other preparatory research activities. Researchers can also apply for IRB approval to obtain more extensive research data sets from the CDW, which may be used for various research purposes ranging from answering additional study feasibility questions to conducting complete analyses for retrospective studies.
Led by Gigi Lipori, the IDR team develops and implements tools that increase access to institutional data — in a secure and HIPAA-compliant environment — for research, clinical and educational activities. The IDR team is a multidisciplinary group that includes Shands Decision and Support Services, UF&Shands IT, UF faculty, the UF IRB and the UF CTSI.
A simple application of an IDR is cohort identification — the ability to ask and answer straightforward questions regarding patient populations and over time. Cohort identification can be used to improve quality of care. It can also be used for clinical trial planning, as was demonstrated by UF College of Medicine investigators Azra Bihorac, M.D., and Mark Segal, M.D., Ph.D., with the help of a CTSI Pilot Project grant awarded in March 2011 to support use of the newly launched IDR for cohort identification.
Drs. Bihorac and Segal sought to identify a unique cohort of patients undergoing in-hospital surgical procedures for their pilot project titled “Database Communication Enables Machine Learning Classifiers To Predict Postoperative Acute Kidney Injury In Intensive Care Unit (Declare).” The incidence of acute kidney injury is as high as 5% to 7% of all hospitalized patients and leads to increased morbidity and mortality. Drs. Bihorac and Segal hypothesized that by using a clinical decision support system, they could develop predictive models for perioperative acute kidney injury using the IDR for cohort identification. Through the IDR, their team was able to bring together comprehensive data related to preoperative, intraoperative and postoperative factors that were scattered in multiple non-communicating databases. The cohort allowed them to apply, for the first time, a novel type of clinical decision support system — a machine-assisted classifier — to identify potentially modifiable preoperative and intraoperative factors associated with the occurrence of perioperative acute kidney injury. Co-investigators on the pilot project included Patrick Tighe, M.D., Kent Berg, M.D., M.B.A., and data analyst Sharmila Mohan.
Ultimately, the team’s vision is to create a unique tool for developing real-time models for predicting acute kidney injury that can be used both in practice and research to enable prospective implementation of new evidence-based interventions for patients carrying the highest risk for acute kidney injury. Drs. Bihorac and Segal envision these predictive models as a stepping stone in the development of widespread predictive models for other perioperative complications that align with the strategic goals of our anesthesiology and surgery departments, which seek to excel in prevention of perioperative complications.
In October, Dr. Bihorac received notification that she will receive the Society of Critical Care Medicine’s 2013 Vision Grant to continue the research initiated with the CTSI pilot grant. The grant will support the ongoing interdisciplinary research partnership between the UF College of Engineering’s industrial engineering department (Drs. Panos Pardalos and Petar Momcilovic), the UF College of Medicine’s anesthesiology and critical care departments, and the IDR.
“The CTSI funds and collaboration with UF&Shands IT were critical for the success of this grant, as they provided preliminary data and allowed us to assemble the Declare cohort for analysis,” Dr. Bihorac said, who acknowledged the support provided by IDR team members Gigi Lipori, Colleen Ebel, Maggie Downey, Felix Liu, Dick Deason and Christopher Harle in assembling the cohort. “This is just the beginning.”
Led by Michael Clare-Salzler, M.D., professor and chair of the Department of Pathology, Immunology and Laboratory Medicine, the CTSI Biorepository provides a centralized, large-scale, cost-efficient alternative to individual tissue and data banks. With the informed consent of patients, it uses state-of-the-art repository protocols for obtaining and storing tissues, and for ensuring that accurate information about the tissues can be provided to researchers in accordance with approved protocols.
It collects biospecimens for an institutional library that can be used for research purposes by any investigator with IRB-approved protocols. It also provides storage services for biospecimens collected by investigators, including Dr. Marco Pahor’s national clinical trial on lifestyle modification in elderly populations. UF Consent2Share will enable the CTSI Biorepository to expand the institutional library in a strategic manner, so as to provide samples of high value to the UF research community.
The CTSI Biorepository follows strict standard operating procedures that are based on best practice guidelines set forth by the International Society for Biological and Environmental Repositories, the Office of Biorepositories and Biospecimen Research and the College of American Pathologists (CAP). As a testament to the quality of its operations, this fall the CTSI Biorepository became one of the first 12 biorepositories in the country to receive CAP accreditation.
Engaging Our Communities
In addition to these efforts, UF is at the forefront of new efforts to engage underrepresented populations in the research enterprise through HealthStreet — one of our community engagement programs. Operating from the Department of Epidemiology under the CTSI umbrella, HealthStreet aims to produce outcomes that matter to the community by: 1) linking people to medical and social services based on their needs and concerns, 2) giving people an opportunity to learn about and participate in health research, and 3) building needed trust between scientists and the community. Already, 2,000 people in the Gainesville area alone have been assessed and linked to medical and social services, of whom 75% have been given opportunities to participate in health research studies. Over 85% of those served are from minority or underserved populations — African American, Asian, Hispanic, elderly, disabled or persons from rural areas in need of services.
A Continuously Learning Health System
Efforts such as these make UF more competitive for research awards nationally, which allows it to be one of the best venues for new discoveries of treatments, interventions and policies that will improve the health of Floridians. It has been argued that achieving higher quality care at lower cost will require health-system feedback loops that ensure continuous "learning,” as the lessons from research and care experiences are systematically captured, assessed and translated back into improved health care. In this context, the Institute of Medicine convened a Committee on the Learning Health Care System in America. The product of the committee’s work, Best Care at Lower Cost: The Path to Continuously Learning Health Care in America, makes the case that emerging tools like medical informatics, connectivity, team-based care and systems engineering techniques make the envisioned transition to a “continuously learning health system” possible, and are already being put to successful use in pioneering health care organizations.
With the CTSI initiatives described here, under the superb leadership of David Nelson, M.D., we can be proud to be among the pioneering health care systems that are well on their way to learn from our patients in a manner that will that improve their health care and lower cost.
David S. Guzick, M.D., Ph.D.
Senior Vice President, Health Affairs
President, UF&Shands Health System