Assistant Professor of Medicine Tiago Colicchio, Ph.D., whose research focuses on increasing the situational awareness of clinicians through the design of electronic clinical documentation systems, joined the Department of Biomedical Informatics and Data Science (DBIDS) in 2017 after receiving his Ph.D. from the University of Utah in biomedical informatics. Recently, he spoke about his journey to being part of UAB, his passion for electronic health records (EHR), and his overall visions as it relates to the world of informatics in the future.
Q: For those who may not be aware of who you are, tell us a little bit about yourself.
A: From a professional standpoint, I describe myself as a clinical informatician working to improve care outcomes with Electronic Health Records. I have a diverse technical and cultural background, with undergraduate training and MBA in my home country Brazil, Masters’ in IT in France, and doctoral training in Biomedical Informatics from the University of Utah. In addition to being a clinical informatician, I’m a philosophy enthusiast, I enjoy reading classical philosophy, as well as Western and Russian literature, and spending time with my family, my wife and our four children who make my journey so much fun.
Q: What (or who) were some of your inspirations for wanting to get into biomedical informatics research?
A: My first job was an internship in informatics at a large health system in Brazil, when I was still in college bank in 2005. In 2011, I still worked at the same place when I visited some friends who lived in Salt Lake City at the time, and I fell in love with the beautiful scenery of the Salt Lake Valley. When I returned to Brazil, I searched for graduate programs in Utah, and by coincidence, I found Guilherme Del Fiol, a Brazilian physician who was assistant professor in the Department of Biomedical Informatics (DBMI) at the University of Utah. I contacted Guilherme via LinkedIn and we set a conference call, when he told me all the history of informatics in Utah (which I completely ignored at the time). I discovered that Utah has the honor of having the first informatics academic department in the world, founded by Homer Warner, one of the founding fathers of the field of informatics. When I became aware of all the history and the treasury of knowledge accessible to informatics students in Utah, I applied for their Ph.D. program, starting my transition from industry to academia from 2012 to 2013.
Q: How important was it for you to create a homegrown electronic health record (EHR) system in Brazil?
A: While working in Brazil I had to wear many hats. For a couple years, I was an IT analyst responsible for maintaining all medication systems internally developed at a large tertiary care hospital. At that time, the organization was expanding its business operations and transitioning from a single site hospital to a health system with multiple outpatient clinics and other hospitals to be constructed in different states across the country. I was then promoted to IT project manager and led the implementation of a commercial EHR (Tasy, Philips Clinical Informatics) in the main hospital and three multispecialty clinics. These projects allowed me to interact on a daily basis with physicians, nurses, and pharmacists, which was an extraordinary experience that allowed me to have a deep understanding of healthcare processes and clinical decision-making, despite the fact that I’m not a clinician by training. I then became IT systems manager and led a team of software engineers who developed a computerized order entry system with advanced decision support, replacing the same functionality in the commercial EHR with our own solution. At the same time, I also served as Project Management Officer (PMO) for the IT department and Software Quality Assurance (SQA) manager. All these responsibilities contributed to a tacit knowledge that helps me identify gaps in current research, in areas that are often understudied, but are extremely meaningful from a clinical standpoint.
Q: With overseeing large EHR implementations in both U.S. and non-U.S. organizations, what are some of the challenges that you have seen (from your perspective)?
A: In Brazil, the main challenges related to EHR adoption and use are the lack of affordable and high-quality EHRs, and most importantly, the lack of qualified informatics personnel, due to the lack of training programs and migration of many qualified professionals to the U.S (I tried to compensate that by writing an introductory book on informatics for the Brazilian audience, which I did in 2020). However, my experience with EHR customization was smoother in Brazil than in the U.S., in part because I was a health system employee and had more autonomy and resources to modify and build applications locally. In the U.S., national adoption of commercial EHRs created an innovation vacuum, with many hurdles to make modifications to applications used at the point of care.
Q: With your EHR research, how do you feel it will help increase clinicians’ situational awareness? Are there any particular areas that you have focused on in your research to help improve the medical decision-making process?
A: I’m currently working on two large projects focused on improving decision support alerts. One is focused on identifying patients with inaccurate allergy records to drugs that they likely tolerate, and use this data to suppress false-positive alerts that are often fired to clinicians, disrupting their workflow and causing them to overlook relevant alerts. The second project focuses on creating standardized methods to evaluate the burden and effectiveness of decision support alerts. I expect that combined, these projects will improve clinicians’ ability to dedicate more attention to alerts with real potential to prevent patient harm, and researches will have a better understanding of the full impact of CDS alerts on clinician behavior and care outcomes.
Q: As Informatics is now a department, what new opportunities do you believe will arise as it relates to your particular area of research within DBIDS?
A: I’m excited with the opportunity to join the new department and participate in the recruitment process that will bring several qualified informatics researchers to Birmingham. I think UAB has enormous potential to become a national and international hub for informatics research and education.
Q: What are some of the ways in which you have seen EHR improve from when you first started up to now?
A: Now, national EHR adoption is not something to pursuit like when I started, it’s a reality. EHRs have been adopted in virtually every single U.S. hospital and clinic, and there is no coming back from this. There are obvious improvements with data accessibility and availability through many EHR functions that facilitate seamless capture and retrieval of a plethora of data that were partially (or totally) inaccessible to clinicians using paper records. However, we are just getting started and have a long way to position EHRs not only as a tool that facilitate data acquisition and presentation, but that is actively used as a valuable complement to complex medical decisions, improving clinician’s ability to keep up with ever-increasing medical knowledge and health care complexity.
Q: What is your vision as it relates to EHR within informatics in the next four to five years?
A: I think we are passing through one of those big transitions that happen to every industry. As commercial EHRs become consolidated as the main platform of a digitized health system, and researchers in academic departments find new ways to collaborate with their academic medical centers that often function as separate units of the same organization, new modalities of informatics innovation will emerge. There is a lot of potential for interoperable applications such as SMART on FHIR and CDS Hooks apps that replace one-size-fits-all functions of commercial EHRs with more efficient, custom-built applications. Further, we now have a wave of methods collectively known as artificial intelligence, that include several modalities of machine learning, including large language models, which are promising tools to improve data processing and present several opportunities for future research. There is a lot of opportunities and challenges ahead of us, a whole blue ocean to be navigated.