 |
Examples of ongoing research in Public health informatics are organized below by faculty member. This is not an exhaustive list. Additional information about current research projects can be obtained at the Department of Biomedical Informatics Website and the Mailman School of Public Health’s Website. Faculty Rita Kukafka, DrPH, MA Public Health Informatics Track Coordinator Dr. Kukafka's research interests focus broadly on use of informatics methods and applications to advance public health practice and research. Training activities include establishing a public health informatics track in the Department of Medical Informatics, and coordinating the health promotion disease prevention track in the School of Public Health with a new focus on interactive health communication. She is currently principal investigator/co-principal investigator on the following funded research projects: 1. An Internet-Based Information Technology to Reduce Prescription Errors with HIV/AIDS Anti Retroviral Medications The goal of this project is to implement and evaluate an information technology intervention to augment professional education and technical assistance efforts intended to reduce prescription errors, improve quality of care and resource utilization in ambulatory care settings. It employs a web-based, interactive decision-support system used by physicians at point of care when they are formulating a patient’s medication regimen, and patient tailored education designed to designed to enhance patient adherence to antiretroviral therapy. This component of the intervention is designed to support the physicians’ efforts to implement the patient counseling components of the guidelines. In addition to the web-based, interactive software, an agency and provider IT needs assessment assesses institutional barriers and physicians’ readiness to use such IT. Results of the needs assessment is used to tailor for the staff of each agency push technology designed to motivate use of the software. The intervention is being implemented in 42 agencies in three states. Agencies are recruited from a larger pool that is receiving a program of continuing professional education and technical assistance offered by the NY/NJ AIDS Education and Training Center (AETC) and the Midwest AETC. Many of the agencies receive Ryan White funding. They all care for substantial numbers of patients from vulnerable populations, and they serve medically underserved communities. The evaluation design randomly assigns agencies into three arms: (1) an enhanced intervention that couples access to the web-site and the push technology intended motivate use of the system, (2) access to the web-site system without the motivational component, and (3) non-intervention control group. Outcome data on medical errors, resource utilization, and patient health status will be collected through chart abstraction. Use of the system will be documented through short provider interviews and usage data collected from the website. Major products include the web-site system, a training manual, and forms to conduct an Information Technology needs-assessment. 2. Automated ICF Coding Using Medical Language Processing The World Health Organization's newly-revised International Classification of Functioning, Disability and Health (ICF) has been cited as a potential code set for reporting issues of measurement and interpretation of functional status. Functional status information is more optimal than past classification systems in carrying out public health practice because it views health broadly considering developmental, behavioral, emotional, social and environmental conditions. Public health has long viewed health within this ecological paradigm. In order for the ICF to be widely adopted, it is critical that we evaluate the ICF as a possible mechanism for that purpose and explore methods to facilitate the inclusion of ICF information into standardized records. The proposed research addresses this task. This research is to evaluate the function of the ICF for capturing and encoding clinical data from patient records. The goal is to explore the feasibility of extracting and encoding functional status information from patient records. The study will evaluate 1) the ICF classification completeness and comprehensiveness in coding clinical concepts and, 2) the MedLEE NLP system in parsing clinically relevant concepts, and coding in the ICF classification. Eneida Mendonça, MD, PhD 1. CAP: Community Access Process The goal of the SASA-CAP project is to develop an integrated delivery system for the uninsured in 5 communities in New York City, including Morris Heights, Washington Heights, Inwood, Harlem, and South Bronx. Main objectives are: a) Create an integrated and efficient health care delivery system through the use of case management using a Health Priority Specialist to link frequent users of the emergency room to primary care practices, b) coordinate and track specialty referrals between hospitals, off-site practices, community practitioners and groups and c) Conduct eligibility assessments to facilitate enrollment in public health insurance programs and re-enroll individuals into Child Health Plus, Medicaid, and Family Health Plus. 2. Immunization Registry Through this collaboration, and with support from the National Immunization Program of CDC, we have developed EzVAC, a provider-based immunization registry and a state-of-the-art information system that was developed to serve clinical needs as well as health outreach and education purposes. EzVAC is an interactive, multi-institution information system that aggregates data from a variety of sources, including community hospitals, clinics and private practices. Built primarily to collect immunization data, EzVAC is a flexible platform that includes an immunization database, hospital registration systems, a Web-based registry server, Web user interface, and reminder, recall and forecast applications. The system has been in use since April 1999 at all the general pediatric practices affiliated with the Columbia Presbyterian Medical Center (CPMC) campus and the New York Weill Cornell campus of the New York-Presbyterian Hospital. EzVAC currently contains immunization records of over 100,000 children delivered by 600 providers, located at the hospital network of pediatric practices, school-based clinics, in-patient wards, the pediatric emergency department and the offices of private physicians in the community David Kaufman, PhD 1. Usability Evaluation of an Innovative Smoking Cessation Program: The Smoke Clinic is an innovative web-based comprehensive smoking cessation clinic. The objective of this evaluation is to characterize the dimensions of the Smoke Clinic web site that promote productive use and meaningful engagement, as well as, to document features or aspects of the interface that may impede such use. Our approach employs two classes of usability evaluation including a) usability inspection and b) usability testing. Usability inspection is designed to characterize dimensions of the interface that adhere to usability principles and to note violations of these principles. We employ 2 classes of inspection methods including a cognitive walkthrough which is scenario or task based and a heuristic evaluation which emphasizes dimensions of the screen. The system is evaluated on the basis of a small set of well-tested design principles such as visibility of system status, user control and freedom, consistency and standards, flexibility and efficiency of use. Jacqueline Merrill, RN, MPH, DNSc Refining Organizational Network Analysis Methods for use in Public Health The purpose of this research is to refine organizational network analysis, an empirical, descriptive technique for modeling organizational structure, for use in public health agencies. This extends pilot research that examined communication and information flows in a local health department using the Organizational Risk Analyzer (ORA), a computational tool for the study of complex systems developed at Carnegie Mellon University. ORA enhances standard network analysis by capturing not only communication links between people, but links to the tasks they perform, and the resources and knowledge they use. The pilot showed network findings have utility for public health, but these techniques have been developed primarily to examine structure in the private sector. Our goals are to customize data collection to accurately capture public health tasks, knowledge and resources; to improve ORA reports for use by public health managers; and to field test these modifications in local health departments. The results will allow us to explore structural determinants of public health agency performance and, in turn, to support management, planning and resource allocation in public health systems.Suzanne Bakken, RN, DNSc, FAAN Suzanne Bakken conducts research at the intersection of clinical informatics and public health informatics. She direct s the Center for Evidence-based Practice in the Underserved which provides administrative support and funding for pilo t/feasibility studies that use informatics strategies in underserved populations. Areas of particular relevance to pub lic health informatics include health literacy and interactive health communication. Two clinical informatics projects led by Bakken, Mobile Decision Support for Advanced Practice Nurses (1 R01 NR008903) and Wireless Informatics for Saf e and Evidence-based (WISE) APN Care (HRSA-funded) include decision support and provision of access to low literacy m aterials that address three major public health issues: obesity, smoking, and depression. George Hripcsak, MD, MS George Hripcsak is professor and vice chair of the Department of Biomedical Informatics of Columbia University and senior informatics advisor at the New York City Department of Health and Mental Hygiene. Dr. Hripcsak's research includes clinical data exchange networks, with an emphasis on regional public health reporting; syndromic surveillance using natural language processing of clinical notes; and the dissemination of public health recommendations to community health centers via electronic health records. He is active in several clinical data exchange networks, including the New York Clinical Information Exchange (NYCLIX) and NewYork-Presbyterian's Systems Technology Interfacing Teaching and Community Hospitals (STITCH), and he is developing methods for evaluating their success. He has applied natural language processing techniques to detect adverse events for patient safety research. Students Michael Bales Michael Bales is a Ph.D. student in Biomedical Informatics at Columbia University. An epidemiologist by training, Michael has spent three years as a fellow in the Public Health Informatics at the Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia. Before coming to Columbia he served in West and Central Africa as a short-term consultant with the World Health Organization (WHO), where he helped the Togo Ministry of Health and the WHO offices in Kinshasa, Democratic Republic of Congo, to build national capacity in data management and analysis. While at Columbia Michael has worked with an interdisciplinary research team to conduct a formal evaluation of human and automated coding of functional status information (FSI) using the International Classification of Functioning, Disability, and Health (ICF) framework. He has also been involved in research pertaining to semantic networks, biomedical terminology structure, and medical language syntax. Delano McFarlane Delano McFarlane (advisor - Dr. Rita Kukafka) is conducting research and development on MyDOK, a personal health record system that uses a mobile data storage device. This system specifically designed to improve continuity of care and ease of disease management for children diagnosed with Type I Diabetes. I am also actively doing research on how to measure and improve the effectiveness of technology based health interventions in low income communities. This research is being done in conjunction with the Harlem Health Promotion Center. Jessica Ancker Jessica Ancker is a Ph.D. student. Her research is focused on using experiential learning to improve risk communication with the public Abstract: Clear provider-patient communication about risks is an essential part of shared decision-making, informed consent, and health education. Unfortunately, many patients have trouble understanding numerical expressions of risk such as rates, odds, percentages, and proportions. They also frequently misinterpret graphics. However, new research in cognitive psychology shows that people make different decisions when they experience probabilities (e.g., by drawing cards from a deck) than when numerical probabilities are described to them (in words, numbers, or illustrations). We are attempting to exploit these findings by developing an interactive electronic tool that provides a virtual experience of the probability of a health outcome. Qualitative and quantitative evaluation is planned. Such a tool might be particularly useful in demonstrating probabilities to patients with poor numeracy skills. |