Research Projects

The Clinical Outcomes Research Initiative (CORI) is a research group which studies gastrointestinal practices and diseases in practice-based settings. Physicians participating in the CORI consortium from across the USA produce endoscopy reports using a specialty electronic health record. Data from those reports are sent electronically to a central data warehouse where they are pooled with data from other consortium participants. Studies on the processes and effectiveness of current practrices are performed with this data; in addition, prospectifve studies are performed using this data to direct recruitment. PI: David Lieberman; Director of Development and Technology: Judy Logan
[CORI Website]

Improving Quality in Cancer Screening: The Excellence Report for Colonoscopy
This AHRQ-funded initiative is designed to evaluate and improve the quality of screening and diagnostic colonoscopies in ambulatory care settings. Using the Clinical Outcomes Research Initiative (CORI) consortium and the National Endoscopic Database (NED), researchers will develop and test the Excellence Report, a quality report card for gastrointestinal (GI) endoscopy, focusing on nationally-recognized quality process measures for colonoscopy. PI: Judy Logan

Clinical Decision Support (CDS) in Community Hospitals: Barriers & Facilitators
This is a National Library of Medicine continuation proposal from the computerized provider order entry (CPOE) research teams at OHSU and Brigham and Women's Hospital. The project examines CDS in community hospitals and to quickly disseminate knowledge about what does and does not work. The three aims of the project are to detail the barriers and facilitators to CDS use in community hospitals, plan and draft CDS guidance tools, and finalize and widely disseminate CDS tools that can be used by community hospital staff to facilitate future CDS implementations by reducing barriers and strengthening facilitators. PI: Joan Ash

Assisting Systematic Review
This new investigator-initiated project studies the hypothesis that machine learning-based text classification techniques can add significant efficiencies to the process of updating systematic reviews (SRs). SRs are essential components in the practice of Evidence-based Medicine (EBM), providing recommendations for medical treatment, diagnosis, prognosis, and etiology based on the best available biomedical evidence. Because new information constantly becomes available, medicine is continually changing and SRs must undergo periodic updates. Currently the identification and assembly of the appropriate literature necessary for creating and updating SRs is a manually driven process that requires a significant amount of time from highly trained human experts. By applying automated document classification techniques early in the process, at the stage of identifying and screening the possibly relevant literature, the workload of the systematic reviewers at these early stages can be reduced. This allows the reviewers to better spend their time conducting more detailed analyses, writing more complete reports, and producing a greater number of complete reviews more quickly. Recently, the Cochrane Collaboration estimated that at least 10,000 total SRs are needed to cover most health care problems. At present, the number of reviews is less than half this after 10 years of concerted effort by the EBM community. New trials are currently published at a rate of more than 15,000 per year - therefore the need for efficiency in preparing and updating reviews is urgent. PI: Aaron Cohen

Care Management Plus

This 4 year project funded by the John A. Hartford foundation has the hypothesis that care managers and information systems can be used for patients with complex needs so that care of patients will be more efficient and quality of care will improve. PI: David Dorr
[ CMP Website ]

Computerized Physician/Provider Order Entry

Implementation of computerized physician/provider order entry (CPOE) is being increasingly encouraged as an important solution to the challenge of medical error reduction. Use of CPOE is not yet widespread, in part because it has a reputation for being difficult to implement successfully. This project is funded by a grant from the National Library of Medicine to study success factors for implementing POE systems. PI: Joan Ash
[CPOE Website]

Image Retrieval in Biomedicine

Finding and using images is an important aspect of biomedical practice, research, and education. ImageCLEF is an international challenge evaluation that provides standardized test collections for assessing image retrieval systems and algorithms. ImageCLEFmed is a task within ImageCLEF that focuses on image retrieval in the medical domain. PI: Bill Hersh
[ ImageCLEFmed Website ]

RxSafe: Shared Medication Management and Decision Support for Rural Clinicians
This research will investigate the feasibility and impact of novel approaches to clinician decision support in multidisciplinary ambulatory care, emphasizing high-risk transitions of care. We expect improvements in medication management by providing the means to effectively share medication information, by sharing the benefit of corrections or improvements to the regimen made by any team member but visible to all the others, and by providing clinicians using the system with access to evidence-based information at the time and place it is needed. PI: Paul Gorman

Using IT To Improve Medication Safety for Rural Elders
This project in rural Lincoln County, funded by the Agency for  Healthcare Research and Quality, serves frail elderly residents of  care facilities who are served by one hospital, six pharmacies, and  four clinics in the area.  All but two of these use computer systems  to keep track of medication information, but none of these computers  systems exchanges information with the others.  As a result, local  clinicians depend on "phone, fax, or foot' to share information.  The  RxSafe system will help nurses, doctors, and pharmacists to be on the  same page when it comes to the medication management of these  residents.   To do this, RxSafe connects the disparate medication  information systems, employing the emerging national e-prescribing  standards, and working within the Common Framework that has been  developed by the Connecting For Health initiative. PI: Paul Gorman

Barriers and Drivers of Health IT Use for the Elderly, Chronically Ill and Underserved
As recommended by the AHIC, this project will "conduct a synthesis of current knowledge from existing studies of health information technology use by elderly, ill, and underserved populations including an analysis of barriers and drivers. The barrier and driver analysis should elucidate for which subpopulations barriers can be overcome and how." PI: Paul Gorman

Fogarty ICTIRT
The Neurotrauma Research Group (NTRG) is a collaboration of investigators and programs from OHSU, the University of Washington, and the University of Rosario (Rosario, Argentina). The long-term objective of the NTRG is to improve treatment and outcomes for people who sustain traumatic brain injury (TBI) through research, education, and the development and dissemination of evidence-based guidelines. Through multiple funding sources, the NTRG conducts brain trauma research and educational programs in the United States and Latin America. With the International Collaborative Trauma and Injury Research Training program (ICTIRT), funded by the National Institutes of Health, the NTRG makes use of the research and educational resources of OHSU and the University of Washington to train Latin American investigators in brain trauma research, and to establish a curriculum of trauma research at the University of Rosario in Argentina. PI: Nancy Carney

Supporting Rapid Transfer of Information and Responsibility
This National Science Foundation project is a collaboration among investigators in the Department of Medical Informatics and Clinical  Epidemiology, Kaiser Center for Health Research, and the Human  Factors Research Program at the University of Maryland.  The research  is concerned with the phenomenon of critical discourse, the rapid  transfer of information and responsibility in multidisciplinary teams working in settings with high stakes, high degrees of uncertainty,  and highly constrained time and resources.   The project includes field studies to describe this activity, as well as development and  examination of technologies to support multidisciplinary collaboration in this environment. PI: Paul Gorman
[ Project Website ]