The Next Frontier in Health Care Reform

May 7, 2010

By Cara Bonnett
Duke Office of News & Communications

Durham, NC — Ed Hammond, who helped design Duke’s first electronic health record 40 years ago, gets a firsthand reminder of the need for personalized medical care every time he visits the doctor. “My son also works at Duke and has the same name as I do,” said Hammond, a professor of community and family medicine and biomedical engineering. “There’s almost a 1 in 10 probability that when I’m seen, it will be my son’s record. It’s understandable: Names are not the best identifier.” Inspired by new efforts to put information technology to work in transforming health care, Hammond, a national expert in data standards and health informatics, came out of retirement last year to head the new Duke Center for Health Informatics.

The center, which brings together Duke’s Schools of Medicine and Nursing, the Pratt School of Engineering and the Fuqua School of Business, will train the next generation of doctors, nurses and health-care administrators in implementing and managing electronic medical record systems to improve patient care. At least 50,000 health-care informatics professionals are needed in the next few years to help meet a federal mandate for digitization of all health records by 2014.

Stimulus Funding from the American Recovery and Reinvestment Act

Ed Hammond, “Curriculum Development Centers,” National Institutes of Health, $1,820,000 for 2 years. Ed Hammond, “Consortia for University-based Training of Health IT professionals in Health Care,” National Institutes of Health, $2,167,121 for 3 years. Hammond has won two stimulus grants under the American Recovery and Reinvestment Act totaling just under $4 million to get the training program started.

Duke’s programs will include a new Master of Management in Clinical Informatics degree, being offered for the first time this fall through Fuqua, as well as degrees from the certificate to the doctoral level. The center builds on a rich tradition of health informatics at Duke, which started one of the nation’s first programs under Hammond’s guidance in the 1970s, said Meredith Nahm, the center’s associate director for academic programs. Bringing together more than 50 Duke faculty, the center will emphasize an approach that includes medicine, nursing, business, economics, biomedical engineering, computer science and public policy. The goal: designing interoperable systems that present customized data for different uses. “Because nurses and doctors use information differently, for example, they’re interested in different aspects of the same information,” Nahm said. “The systems need to pivot to present information in a way that’s appropriate to the situation and a particular user.” System integration issues make it difficult to share the data that clinicians need to see the entire picture of a patient. And incomplete information can result in medical errors, which claim as many as 98,000 lives each year, according to a report by the Institute of Medicine. “As we develop an interconnected health system, medical records can’t exist in silos,” said Constance Johnson, director of Duke’s nursing informatics program. Hospitals have lagged in adopting analytical tools that would allow them to capitalize upon existing data stores to increase efficiency, reduce costs and improve health outcomes, said Jeff Ferranti, the health system’s associate CIO for health analytics and patient safety. The federal government plans to distribute $19 billion to implement electronic medical record systems, now in use by only about 8 percent of the nation’s 5,000 hospitals and 17 percent of its 800,000 physicians. “We’re at a critical crossroad: Our ability to gather data has outpaced our capacity to aggregate information to advance care,” Ferranti said. “That’s the primary motivation for the center. We’re creating a learning health care system, where the data collected isn’t just a byproduct of care but is reintegrated into the process, where it is aggregated and synthesized to identify trends and improve patient care.”

Duke, for example, has collected data since 2004 on patient safety incidents – an estimated 1,800 reports each month – and aggregated the data “to identify problems before they become problems,” Ferranti said. “Before electronic records, we had 23 disparate, paper-based reporting systems. Now we have five years of codified safety information that we can use to study issues from falls to transfusion deviations to adverse drug events.”

The new center will leverage Duke’s strong integration between health system operations and research by offering students access to Duke’s data warehouse, which contains more than 3.8 million patient records from the system’s three hospitals and 100 clinics in the Triangle area. “At the end of the day, what we’re really doing is speeding the translational process from bench to bedside to public health initiative,” Ferranti said. For Hammond, the center offers new hope for putting health data to work to improve the human condition. The past 40 years have brought technological advances that enable new possibilities, he said, from tools that can help doctors tailor treatments to individual patients’ needs to software that researchers can use to identify the best candidates for clinical trials. “We live in a new world, and we need new, innovative thinking to create a seamless system, without boundaries, where we can exchange data across different sites of care, to create a single complete picture of what happens to the patient,” Hammond said. “This is an exciting time.” The creation of the center, which is housed within the Duke Translational Medicine Institute, was supported by a grant from National Center for Research Resources (NCRR) through the American Recovery and Reinvestment Act (ARRA).

Shaw receives F31 National Research Service Award

September 19, 2010

Ryan Shaw, one of the doctoral students in the Duke University School of Nursing, has received a F31 NRSA award for his fellowship proposal entitled “A mixed methods RCT to sustain weight loss among persons with obesity using theoretically tailored mobile telephone short message service (SMS).” The award is for a three-year period, from December 1, 2010, to March 31, 2013, in the amount of $124,140. His fellowship sponsor is Constance Johnson, PhD, RN.

“Intervention strategies that generate successful sustainability and maintenance of health behaviors have proven elusive. The integration of emerging social and behavioral theories with informatics has the potential to advance the science of health maintenance. Technological mediums can be used as tools to deliver tailored information to persuade and transform health behaviors and promote maintenance. This study uses theoretically tailored messages sent via mobile telephones to promote sustainability of weight loss among people with obesity.”

HL7 and IHTSDO Expand Collaborative Efforts



Ann Arbor, MI- Health Level Seven (HL7) and the International Health Terminology Standards Development Organisation (IHTSDO) announced Thursday, July 7th that they have expanded collaborative efforts. According to leaders from both groups, the closer collaboration will foster healthcare information interoperability and lead to improvements in patient safety by futher facilitating the use of IHTSDO standards with HL7 standards.

IHTSDO and HL7 are jointly working on a number of key global healthcare information technology standards initiatives, including:

  • IHTSDO is making SNOMED CT codes and descriptions freely available for release in HL7 products, under  a public good license, in order to improve semantic healthcare interoperability across countries and to improve patient safety. HL7 will use the codesand descriptions to produce a number of SNOMED-enabled products, using universal realm bindings to SNOMED CT where appropriate.
  • HL7 is investigating use of the IHTSDO Workbench to maintain its vocabulary. Use of common tooling will improve harmonization, leading to benefits across the two organizations as well as for users of both standards.
  • HL7 and IHTSDO are jointly reviewing and streamlining the request submissions process, to enable requests for additions and amendments to SNOMED CT codes to be made more effectively. Improvements are being made in HL7’s internal processes, and also in IHTSDO’s request submission process.
  • IHTSDO has announced its support for the development of Release 2 of the Common Terminology Services standard (CTS2) by HL7 and OMG. This standard will provide a consistent interface into terminologies in EHR environments.

According to HL7 Board Chair Bob Dolin, MD, “Expanding on the coordination and collaborative efforts between HL7 and IHTSDO, these joint initiatives signify a major step toward the construction of globally applicable healthcare interoperability standards that will require far less realm-localization. HL7 is committed to working with IHTSDO and other standards bodies to advance the delivery of safe and effective patient care.”

IHTSDO Chief Executive Officer Jan-Eric Slot said, “Making it easier to use HL7 standards with SNOMED CT will facilitate global standards-based healthcare implementations, where it will add the semantic dimension to interoperability.” He continued, “The closer working relationship between these two leading organizations will improve the uptake and effectiveness of these jointly developed standards by members of both HL7 and IHTSDO.”

Health care informatics: the new kid on campus

October 21, 2010

Reconsidering your career? Focusing on your future? Picking a profession? The health information technology field has much to offer.

The Office of the National Coordinator for Health Information Technology has estimated that over the next five years the health care industry will face a workforce shortage of 50,000 qualified HIT professionals. To close the gap, the ONC, through the American Recovery and Reinvestment Act of 2009, is funding informatics training programs at universities and community colleges nationwide.

Ed Hammond, PhD, DCHI director, and Meredith Nahm, PhD, DCHI associate director of academic programs, are quoted in this article.

Read the full article in the CAP e-Newsletter: Health care informatics: the new kid on campus

Duke University Medical Center creates solution for shortages in health informatics

August 27, 2010

By Duke Medicine News and Communications

A national shortage of more than 50,000 information technology professionals is a serious shortfall as hospitals strive to create robust and comprehensive electronic medical records.

One helpful solution is to broaden existing informatics programs to create new professionals of many types. On Aug. 2, 2010, a new certificate in Health Informatics was launched at Duke University Medical Center by broadening an existing informatics post-masters certificate to open access to individuals with a clinical undergraduate or graduate degree.

Read the full press release.

Certificate Student Earns Paid Internship at Allscripts



Amanda Truijillo, a student in the Graduate Certificate in Health Informatics program, was offered a full-time paid internship for the summer at Allscripts with the Care Management team in Professional Services. DCHI hosted a senior management visit to Duke in May following the HIMSS meeting in New Orleans where Allscript’s voiced an  interest in recruiting Duke students.

Informatics Research Seminar: VisualDecisionLinc– A Comparative Effectiveness Approach To Advance Decision Support in Psychiatry

October 13 @ 4:00 – 5:00 pm


Speaker: Ketan Mane and Charles Schmidt
Presented from Duke University


In this seminar, we describe VisualDecisionLinc (VDL), a novel visual analytics based decision support tool. VDL is based on comparative effectiveness research that allows clinicians to quickly assess the treatment outcomes of similar patients to guide their decision making process. Presented in a dashboard style, VDL is designed to process and analyze large number of patient visits, and distill the therapeutic knowledge related information from comparative population to a summarized visualization based data views. Inbuilt user interactions are added to refine the derived therapeutic knowledge to best meet the clinician’s needs for an informed patient centric decision.


Ketan Mane is a Senior Research Informatics Developer in the ‘Health Informatics and Biosciences Group’ at Renaissance Computing Institute (RENCI). His work focuses on applying visual analytics approaches in decision support role to help clinicians identify viable treatment options at the point of care. Mane has a background in biomedical engineering, and holds a Ph.D. in Information Science from Indiana University, Bloomington. Prior to joining RENCI, Mane was part of the InfoViz. lab at Indiana University directed by Dr. Katy Borner. He has also worked as a Postdoctoral Research Fellow at Los Alamos National Lab (LANL). His research interest include: information visualization, visual analytics, comparative effectiveness research, decision support tools, health informatics, knowledge domain visualization.

Charles Schmidt provides technical leadership and management for RENCI biological and medical science related projects in the areas of patient monitoring, systems biology, and genomics. Prior to joining RENCI, Schmitt was the senior computer scientist at BD Technologies, Research Triangle Park, where he assisted in software development and bioinformatics support for programs in stem cell research, immune function, medical diagnostics, genomics, and proteomics. He also served as the primary architect and developer of the MPM software informatics platform. Schmitt is a member of the Association for Computing Machinery, Institute of Electrical Engineers, and the International Society for Computational Biology. He holds a B.S. degree in physics and a Ph.D. degree in computer science from UNC-Chapel Hill.

Informatics Research Seminar: What is Biomedical Informatics and Why is it Hard?

October 20 @ 4:00 – 5:00 pm


Speaker: Todd R. Johnson, PhD
Presented from Duke University



Over the past 30 years, computer power has drastically increased and many activities now benefit from information technology. Despite some early successes in biomedicine, most notably in genomics and proteomics, many areas of biomedicine have not benefited. . This is particularly prevalent at the clinical level, where the failure rate of information technology is over 50%, and studies have shown that even successful IT has had little effect on either cost or quality of care. Although many factors influence the success of IT, in this talk we argue that the nature of biomedical information and information processes makes them inherently difficult to automate with computers. We lay out this argument by proposing a framework for understanding the nature of information, information processing, and the role of computers and computation in information systems. This framework pinpoints several sources of difficulty with the use of computers in biomedicine and suggests that to effectively use computers in biomedicine, we must fundamentally change the way we design, view, and interact with our information systems. We conclude by briefly describing a two-pronged approach to designing information systems for biomedicine. The first, which we call Collaborative Semantic Information Processing, involves designing information systems that explicitly support the complementary abilities of humans and machines for processing meaning vs. symbols. The second approach involves a shift from primarily symbolic, logical modeling of biomedical concepts to a probabilistic approach that explicitly represents the uncertainty and vagueness implicit in most biomedical concepts and terms.


Professor and Director, Division of Biomedical Informatics, Dept. of Biostatistics, College of Public Health
The University of Kentucky

Dr. Johnson’s research uses cognitive science, computer science, and human factors engineering to solve biomedical informatics problems. He views informatics as the science of information, where information is defined as data + meaning. In other words, informatics is the science of meaningful data.

Informatics Research Seminar: Using the Open Metadata Registry (OpenMDR) to create data sharing interfaces

October 27 @ 4:00 – 5:00 pm

Speakers: David Ervin & Rakesh Dhaval, MS
Presented from UNC-CH



The domains of clinical and translational research are collaborative in nature and team science necessitates using information systems that are both locally relevant and globally interoperable. The Translational Research Informatics and Data Management (TRIAD) grid is an initiative to provide innovative methods for integrating heterogeneous information across institutional boundaries to increase the speed, efficiency, and impact of clinical and translational research. As an extension of the caGrid middleware, TRIAD is a service-oriented infrastructure designed to support translational research by enabling the creation of scalable, secure and knowledge-anchored data-sharing environments.

OpenMDR is a suite of tools that provides grid-compatible semantic metadata management capabilities, including the creation of locally relevant ontology-anchored data elements and conduct of federated queries and retrieval of semantic metadata from repositories across grid-enabled networks, including TRIAD and caGrid. The suite comprises of four different components: 1) MDR Core, 2) MDR Query, 3) MDR Plug-in, and 4) MDR Domain Model Generator.

MDR Core is an ISO11179 semantic repository capable of storing, versioning, and maintaining semantic and representational metadata. MDR Query is a grid service used to search multiple semantic repositories, giving developers the option of using other semantic metadata management tools in addition to those provided by the NCI such as caDSR and EVS. UML modelers use MDR Plug-in within Enterprise Architect to search for semantic metadata from multiple registries. Service developers can use the MDR Domain Model Generator to create semantic metadata for caGrid and TRIAD grid data services. Each of these projects provides functionality that enables federated semantic metadata annotations to be created and used in Grid Service Registration and Discovery. OpenMDR is designed to be locally deployed, populated, and curated. This allows service developers and institutions to maintain locus of control for their data and terminologies while facilitating rich semantic interoperability with other institutions, and maintaining a fast and agile process for annotating and delivering a strongly typed and semantically anchored grid services into production. Such services provide out-of-the-box data sharing functionality and through the use of existing grid tooling and shared services relating to discovery and federated query capabilities to address issues.


David Ervin, Center for IT Innovations in Healthcare, The Ohio State University
Rakesh Dhaval, MS, Center for IT in Healthcare, The Ohio State University