Clinical informatics interventions have been at the heart of UNC initiated programs that improve care through research and practical application. The primary care practices on campus have implemented diabetes registries utilizing point of care reminders that optimize on site usual care and population management techniques utilizing enhanced care teams that achieve chronic care outcomes that reach and exceed many national benchmarks. Similar programs in depression management and anti-coagulation management have also achieved these high standards. The clinical informatics push has not been limited to campus. Through the NC Regional Extension center for Health Information Technology, an arm of the NC Area Health Education Centers Program, health information technology has been spread to nearly 4000 providers across the state of North Carolina. After reaching CMS defined “meaningful use” many of these practices have enrolled in a program that incorporates algorithmic care, quality improvement techniques, and organizational redesign to optimize chronic disease management and preventive care. These outreach efforts have left UNC well positioned to partner with these practices for future research efforts including patient-centered outcomes research (led by the Sheps Center for Health Outcomes Research), dissemination and translational research (supported by our CTSA organization, NC TRACS).
Ongoing Clinical Informatics Projects
UNC is home to one of the 12 NIH-funded Patient Reported Outcomes Measurement Information Systems (PROMIS) research sites. UNC Site focuses on validation and linking of pediatric and adult PROMIS item banks for four pediatric chronic illnesses: asthma, cancer, nephrotic syndrome and sickle cell disease
Interactive Cancer Communication Systems (ICCS) Directed Physical Activity Enhancement for Colon Cancer Survivors. The project tests a Survivor Comprehensive Health Enhancement Support System (SurvivorSurvivorCHESS), an ICCS using a smart phone to increase physical activity, decrease distress, and improve quality of life in colon cancer survivors during their transition from initial treatment to extended survival. More
Effect of Glucose Monitoring on Patient and Provider Outcomes in Non-Insulin Treated Diabetes. This study is supported by the Patient Centered Outcomes Research Institute (PCORI) award. Electronic health record data, glucose meter readings and messaging, and other study-related data will be captured via an internet-based system designed by the Sheps Center for Health Services Research IT team in collaboration with the Carolina Data Warehouse, the practice network, the glucometer manufacturer, and the UNC project team. More
Readmission Risk Prediction Model Using Natural Language Processing. The overall goal of this research project funded by IBM and Allscripts is to use both unstructured (extracted through natural language processing) and structured data from our institution’s clinical data warehouse to derive and validate an accurate model that predicts 30-day hospital readmission risk in hospitalized patients. This goal will be accomplished by using the Carolina Data Warehouse for Health (CDW-H), IBM Content Analytics, and SPSS to extract and analyze both structured and unstructured data stored in the CDW-H on all eligible patients discharged from UNC Hospitals over the past 5-years. We will compare the overall accuracy (c-statistic) of prediction models that use structured data only, with models using both structured and unstructured data elements. We hypothesize that the accuracy of the prediction models will improve with the inclusion of psycho-social information from free-text narrative clinical notes and reports extracted using NLP.
A Patient-Centered Hybrid Portal: Transforming the Medical Record into Meaningful Information for Patients. This project is funded by a Clinical and Translational Science Awards (CTSA) program at the University of North Carolina at Chapel Hill (UNC-CH). The overall scope of this project is to utilize both unstructured (extracted through natural language processing) and structured data from clinical, billing, and administrative data repositories from the CDW-H to create a proof-of-concept prototype for a hybrid patient portal that combines the features of traditional patient portals and patient health record tools and facilitates informed decision making
Program for Assisted Transition from Hospital to Home (PATHH). This study, funded by Merck Sharp and Dohme, is enrolling patients hospitalized with congestive heart failure and providing each patient (and/or caregivers) with a web-based personal health record (Blue Medics) that will help them keep track of their medications, doctor appointments, daily weights, and vital signs after discharge. The goal of the project is to reduce 30-day hospital readmission rates for heart failure patients.
Developing a Decision Support System to Facilitate Accurate Radiographic Interpretation by Emergency Room Physicians. This is a joint project between UNC-CH and the Renaissance Computing Institute. The project will develop and validate a decision support system that uses image analysis technology to help emergency room physicians identify abnormalities on chest x-rays (example; lung nodules) requiring further evaluation after patients are discharged from the emergency room.
Treating the Whole Patient: Evaluating the Use of Depression Treatments during the Year Following a New Colorectal or Breast Cancer Diagnosis. This project utilizes secondary administrative claims data contained within UNC’s Lineberger Integrated Cancer Information and Surveillance System (ICISS). A part of this project focuses on obtaining and linking pharmacy claims data from Medicare Part D plans to allow the research team to evaluate mental health pharmacotherapy in the Medicare-eligible population along with other insured NC residents. The overall aims of the project are to: 1) evaluate patterns of antidepressant and mental health counseling services use among patients during the year following a new cancer diagnosis, 2) determine whether patient or cancer-related characteristics are associated with treatment use patterns, and 3) determine whether provider or systems-level characteristics are associated with treatment use patterns.
ICCS Directed Physical Activity Enhancement for Colon Cancer Survivors. The project team is developing an Interactive Cancer Communication System (ICCS) for smart phones called Mobile Comprehensive Health Enhancement Support System (mCHESS). This innovative smart phone intervention draws on our previous successful research to increase physical activity and to improve quality of life of cancer survivors; mCHESS will provide health information, tools, and a support system adapted from these evidence-based programs (FRESH START and CHESS). The primary aim of the study is to determine if mCHESS offers a more effective behavioral change intervention compared to usual care practice.
Self-Generated Health Information (SGHI) Consortium. University of North Carolina at Chapel Hill School of Information and Library Science and the RTI International Center for the Advancement of Health IT have joined forces to lead the creation of a Self-Generated Health Information (SGHI) Consortium to develop an SGHI Exchange Market. SGHI is information created, recorded, gathered, or inferred by or from individuals by a variety of applications and devices, such as Mobile Health Applications, Activity Tracking Sensors, Running Applications, Calorie Counting Applications, Genome Analysis Applications and many others. The standards-based SGHI Exchange Market will use open architecture in combination with rewards-based strategy to promote sharing of information between providers, technology companies, payers, researchers and individuals.