Nov 03 2015
“You are over pronating!” bleeps Bluetooth enabled, sensor-infused smart socks just 10 minutes into your daily running routine.
Or, “check your blood pressure, temperature and heart rate using Scanadu, with no cuffs, in just 15 seconds!” quips Kevin Maloy of MedStar Institute for Innovation during our fourth annual Georgetown Biomedical Informatics Symposium on October 16, organized and hosted by the Innovation Center for Biomedical Informatics. Emerging trends in informatics and health IT were demonstrated and discussed with over 350 attendees from academia, industry and government. The event benefited from strong support of institutional and industry sponsors. Find out more about the symposium’2015 here. I present Cliff Notes version of four major themes here for your quick browsing.
- Quantified Self
We are increasingly wearing bands to track how much we move, strapping on watches to listen to our heartbeat, and logging what we eat and drink. The underlying proposal is that describing ourselves with these numbers will put healthcare back in the hands of people. Will the “quantified self” become a major driver in how diseases are prevented or treated? This is one of the intriguing questions that our symposium explored.
Informatics opportunity: Design in healthcare is an opportunity to improve signal and reduce noise in a system that is over stretched, under utilized and very expensive.
- EHRs and other emerging health technologies
Digitized health is a dream come true for many. But are electronic health records (EHRs) actually getting in the way of physician productivity? At our symposium, Mike Hogarth of University of California, Davis presented results from a survey of 410 internists that estimated that 42 minutes are lost each day by physicians due to EHRs. About 80% of key clinical data are in the form of unstructured narratives – a mess he referred to “dirta,” instead of data. This information requires enormous quality control, structuring, and integration – a reality that raises the question: can practice-based evidence be generated through retrospective studies of EHR datasets?
Informatics opportunity: Nigam Shah of Stanford University suggested that enterprise wide data governance at hospital systems, or a green button function within EHRs, could help clinicians use aggregate patient data to support decisions at the point of care. Ben Schneiderman of University of Maryland demo’ed EventFlow, a tool for visual analysis of temporal events within patient records to enable advanced healthcare discovery. Zak Kohane of Harvard University, in his keynote lecture, cited clinical research data integration software such as i2b2, tranSMART, and SMART Health IT apps as solutions to the “dirta” problem in healthcare innovation.
- Trends in Precision Medicine
A lot of the excitement at the symposium – amplified by the talks on targeted therapies in pancreatic cancers and a panel discussion on Next Generation Sequencing (NGS) in the clinic – was focused on Precision Medicine.
Mike Pishvaian of Georgetown University and Dr. Jon Brody of TJU discussed PANCan’s “Know your tumor” program. This program has found that 43% of patients had actionable findings from molecular profiling, resulting in modified treatment recommendations and better responses.
Regeneron’s Rick Dewey asked a provocative question: what if everybody’s genome was available in his or her medical record? Rick and Marc Williams of Geisinger described a collaboration between Regeneron and Geisinger to use EHRs and exome sequencing data from over 200,000 individuals for faster drug discovery. It was a treat to hearabout Geisinger’s GenomeFirst initiative, which is implementing genome inference engines – clinical decision support and predictive models to enable Precision Medicine in a unique way with teams of clinicians, genetic counselors, nurse practitioners and informaticians.
No scientific symposium is complete without an award! The (iPAD winning) best poster award went to Ao Yuan, graduate student in Biostatistics at GU for his work on a semi parametric model for the analysis of patient sub-groups in precision medicine trials.
The Precision Medicine journey is underway, and is already improving medicine. Informaticians are vital to this journey. More work is needed to collect the right patient cohorts for research, to identify the right markers to test, and to develop the appropriate targeted therapies.
The Symposium explored what’s next for all of us in this important journey?
Informatics opportunity: Curating evidence of biomarker association with drug response, novel data wrangling approaches to extract and analyze useful clinical and genomic data to drive new hypothesis generation and clinical decision support, and data science approaches to connect genotypes to phenotypes are a few of many opportunities for informaticians to meaningfully participate in the precision medicine revolution.
- Security, Privacy and Trust principles for patient-centered studies
The symposium was a perfect lead-in to a great roundtable discussion on a much-needed security framework for President Obama’s Precision Medicine Initiative at the White House OSTP. I was humbled by the discussion with experts in cyber security, patient privacy, trust principles, and data breach. Will “white hat hacking” help? How can we use it in the context of protecting healthcare data and participants from willful misuse?
Informatics opportunity: DJ Patil, US Chief Data Science Officer emphasized the need for IT teams to focus on data infrastructure, auditing and monitoring of patient genomic data, data transmission and access infrastructure, including tiered data access.
It is so gratifying to see informaticians providing thought leadership across the full spectrum of clinical research and care. Let’s continue the conversation – find me on e-mail at email@example.com or on twitter at @subhamadhavan.