Biomedical Informatics

By Peter Elkin, MD

Biomedical informatics is the field of medicine that deals with the representation, use, analysis and prediction of health information. Subfields include bioinformatics and health informatics, which includes public health informatics, clinical informatics (a new board-certified medical subspecialty), biomedical ontology, imaging informatics and human-factors engineering for health.

To understand where biomedical informatics is moving health care, one needs to look first at the wonderful progress that we have brought to health care to date. Informatics created MedLine and Pubmed, which expose the biomedical literature to clinicians and biomedical researchers all over the planet and are the basis for systematic reviews and meta-analyses that drive clinical guidelines and evidence-based medicine. Informaticians helped map the human genome, and it is estimated that this may have saved five to ten years of laboratory work. This has led to precision medicine, which is changing the practice of medicine through the use of genomic and proteomic data to direct therapy.

Informaticians developed standardized electronic prescribing methods, which increased the safety of prescriptions nationally. They developed the technologies that drive electronic health record systems (EHRs). Along with EHRs, informaticians developed order sets and clinical-decision support that help clinicians to practice accurate and evidence-based medicine. These are now bolstered by cellphone applications that bring medical knowledge to the palm of your hand. Unanswered questions in the practice of health care will less often delay or lead to non-evidence-based practice.

We have begun to change health care from a cottage industry where clinicians practice as they were trained, to a systematized practice of medicine. The Affordable Care Act was made possible in part due to the development of systems that use interoperable semantic data that monitors the quality of care a patient receives. Natural-language processing and standardized observational databases have strengthened our ability to algorithmically understand patients’ longitudinal clinical history of care and their clinical outcomes. Today, there are systems to determine the risk of individuals based on their health conditions. In the future, we will monitor all providers’ risk-adjusted quality performance measures. This will allow us to incentivize clinicians to provide high-quality care to patient populations that have high-risk conditions.

Looking to the future, the goal is to produce a learning-health system where we learn from our practice of medicine every day in order to provide better care and to keep our populations healthier in the long term. In order to accomplish this goal, we need to work together. We need to use our data to derive value. This can be accomplished in part by using artificial intelligence and machine learning to create predictive analytics to determine who is at risk for bad outcomes, and to give those individuals access to intensive preventive therapy. An example might be to provide more frequent screening to individuals at high risk for cancer, thus leading to fewer cancers.

Other examples are new models of collaborations between specialties to ensure that we communicate and share best practices. This could include electronic tumor-board applications, which have the capability of bringing expertise to bear on a patient’s case across distance and time.

Biomedical informatics with its new specialty of clinical informatics is leading health care toward a set of human-computer partnerships that will keep our populations healthy, and will expediently and accurately treat the sick. Informaticians at UB and around the world will need to partner with all specialties to advance and improve health and health care through faster and more systematized clinical practice, medical education and biomedical research.