High Velocity Medicine

Moving Toward High Velocity Medicine
From The Rock:
Where is the wisdom that we have lost in knowledge?
Where is the knowledge that we have lost in information?
-- T.S. Eliot, 1934
A while back I had a great conversation with Vi Shaffer and folks from Gartner. I've known Vi forever and find her to be a smart and insightful person who knows the field of health informatics quite well. We were talking about 'nanotechnology' for some background research I was doing for projects at Partners, and we got to talking about clinical decision support and knowledge management along the way. This conversation helped me crystallize one of the major issues I see confronting US healthcare delivery, and for that matter, healthcare delivery around the world: the need for High Velocity Medicine. Try this on for size:

 Many in the field have talk for years about the DIKW pyramid, which I think started simply as the DIK pyramid. In any case, the basic idea is that we live in a world of data about nearly everything -- especially as more and more of our lives are 'quantified', and the world is painted in a digital facsimile of itself (aka the Matrix!). Data are the objective facts about our existence, or for example: the height of Mt Everest in the graphic. Data may be compiled, however, to give us insights about things, or information. For example, the average height of mountains in Nepal, or a book about Mt Everest. When internalized, information becomes knowledge, or something we know first hand. But it is the last step, transforming knowledge into practice that is the true goal -- knowledge acted upon is wisdom. And with wisdom and experience, we need to know which data to look at again to complete the cycle, and create a virtuous learning system.



Some have suggested that it may take as long as 25 years or more to covert new data into wisdom, e.g. to take a new clinical trial result and see it in routine clinical practice. In a seminal paper, Lau described in the NEJM in 1992 how long it took for thrombolytic therapy in acute MI to diffuse into routine practice -- the first clinical trial was performed in 1960, and thrombolytics in AMI became routine only in 1985. 

Given the appalling rate at which we routinely exercise our wisdom in daily life, or the practice of medicine, it is no surprise that this graphic is shaped like a pyramid -- perhaps implying less wisdom among us mortals, than knowledge, information, or data.

Yet, it is a central idea of biomedical informatics that we can improve upon this process of translation from data to wisdom, and apply it in the routine practice of medicine as wise, seasoned clinicians. We need to dramatically accelerate this translation of data to knowledge, and discern when is new knowledge 'true' and ready for general use, vs. that knowledge which is observed, and possibly even replicated, but perhaps not yet generally applicable.

To me, this suggests a new theory of evidence given the volume, speed, and diversity of sources of new data (aka 'big data') -- moving from hypothesis-driven clinical trials toward PheWAS (phenome-wide) and GWAS (genome-wide) association studies… with biologic plausibility and clear certainty assessments. But that's another topic.

With the explosion of biomedical data -- data on genotypes, phenotypes, social interactions, behavior, geospatial, populations, community, and more to come -- it is incumbent upon us to focus on the acceleration of this translation from data to wisdom, it's dissemination, and implementation in the tools of the day: electronic health records. I suggest this will lead to 'high velocity medicine': when we are agile with new data, compile it into information with routine ease, quickly understand and interpret it into new knowledge, and act upon it as wisdom with alacrity. This is the fuel for Berwick's "escape fire" - avoiding the current conflagration through the transformation and optimization of healthcare delivery, and the primary pursuit of health and wellness.

Discussion with Vi Schaffer of Gartner

Lau J, Antman EM, Jimenez-Silva J, et al. Cumulative meta-analysis of therapeutic trials for myocardial infarction. N Engl J Med 1992;327:248–54.
Berwick, D. Escape Fire: Lessons from the Future of Healthcare. 2002, The Commonwealth Fund.

Folks, posting for a friend to promote what sounds like a great conference:

H@cking Medicine at MIT

We are proud to announce Hacking Medicine, the first event of its kind. Engineers, scientists, physicians, and entrepreneurs, in one location, creating disruptive healthcare solutions today. If you want to radically change healthcare, then apply now: http://bit.ly/nXIF3P. Bring your skills, your ideas, or both. We're selecting 100 people just like you. Be part of the inaugural conference. Leave with a team and a hack on its first steps towards becoming a company and disrupting healthcare. For more information or to apply to be one of the 100: http://hackingmedicine.mit.edu/ Hacking Medicine takes place October 22nd and 23rd at the Media Lab at MIT. If you are selected we will send you more detailed logistics. Healthcare needs you, come be part of the solution. The Hacking Medicine Team

The Clinical Decision Support Consortium

Background

Electronic health records (EHRs), when used effectively, can improve the safety and quality of medical care. For maximum benefit, however, EHRs must be paired with clinical decision support (CDS) systems to effectively influence physician behavior. CDS includes a variety of techniques designed to facilitate and guide doctors’ decision making toward evidence-based practice. Common examples of CDS include computerized checks for drug interactions and electronic reminders for screening tests like mammograms and Pap smears.

While the evidence that CDS can be effective is clear, current use and adoption of CDS is limited. In fact, most of what we know about CDS comes from only four academic medical centers and integrated delivery networks. Wider adoption of decision support has been held back by a variety of issues, including:

• Difficulty translating medical knowledge and guidelines into a form that can be used by EHRs.
• Technical challenges in developing a standard representation for CDS content that could be shared across sites.
• Absence of a central knowledge repository where human readable and executable guideline knowledge can be shared and stored.
• Challenges in integrating decision support into the clinical workflow and other barriers to IT adoption
• Limited capabilities for clinical decision support in commercially available electronic health record (EHR) systems.

The AHRQ Clinical Decision Support Consortium

While these issues have been barriers for adoption of clinical decision support systems they are surmountable, as evidenced by a small number of sites where decision support is pervasive. We believe that the biggest challenge to fostering widespread adoption of clinical decision support is in documenting, generalizing, and finally translating the experience from these advanced sites to a broader community of care sites. To address this challenge, investigators from Brigham and Women's Hospital, Harvard Medical School, and Partners HealthCare (PHS), have formed the AHRQ Clinical Decision Support Consortium (CDSC) in collaboration with the Regenstrief Institute, Kaiser Permanente Northwest Research Group, the Veterans Health Administration, Masspro, GE Healthcare, NextGen and Siemens Medical Solutions.

The goal of the CDSC is to assess, define, demonstrate, and evaluate best practices for knowledge management and clinical decision support in healthcare information technology (IT) at scale – across multiple ambulatory care settings and EHR technology platforms.

Our work is guided by a series of high-value research questions:

• How do we improve the translation of knowledge in clinical practice guidelines into actionable clinical decision support in healthcare information technology?
• How do we optimally represent knowledge and data required to make actionable clinical decision support content in human readable and machine readable and executable forms?
• How do we collate, aggregate, and curate knowledge content for clinical decision support in a knowledge portal used by members of the CDS Consortium? How may we use such a tool to support knowledge management and collaborative knowledge engineering for clinical decision support at scale, across multiple healthcare delivery organizations, and multiple domains of medicine?
• How do we demonstrate broad adoption of clinical decision support at scale in different healthcare IT products that are used in disparate ambulatory care delivery settings? Such demonstrations may show the utility of a simplified clinical decision support knowledge specification in human readable form, as well as the utility of publicly available CDS web services, and their incorporation in CCHIT-certified electronic health records (EHR).
• How do we define and evaluate best practices in response to the above assessments and demonstrations? Evaluation must include an assessment of how to incorporate clinical decision support services at scale in a variety of vendor healthcare information technologies, as well as products developed in academic settings. Further, how do we deploy clinical decision support services in healthcare information technology in a manner that improves CDS impact?
• How do we broadly disseminate the lessons learned over the course of these investigations to key audiences, such as the academic informatics community, patient safety and quality groups, medical specialty societies, small office practice settings, and others?


CDS Consortium Activities

The CDS Consortium will carry out a series of activities over the next two to five years:

• Carry out an assessment of knowledge management practices: CDSC researchers will travel to a representative Consortium member site to gather information on their decision support systems and knowledge management practices. This team will use a mix of on site assessment methods, including in-depth interviews, and observations, as well as conduct follow-up visits to the sites after demonstrations have been completed.
• Translate guidelines into actionable decision support tools: The CDSC Knowledge Translation and Specification team (KTS) will select medical guidelines and develop standards and mechanisms for these guidelines into unambiguous knowledge specifications. Those specifications will be used in the subsequent service and demonstration projects.
• Build a knowledge portal and repository (KPR): The KPR team will develop knowledge management tools for use in the development, review, publication, cataloging and archival of clinical guidelines in human and machine-readable forms.
• Develop CDS Services: The CDS services team will take the knowledge specifications developed by the KTS team and develop publicly available web services that will implement the KTS team’s logic. These services will be made available to clinical information systems.
• Carry out demonstrations of CDS across sites: The CDS Demonstrations Team will implement decision support interventions using the content and services developed in the consortium. The initial demonstrations will occur in the Partners Longitudinal Medical Record (LMR) and will incorporate other sites in future years.
• Build “dashboards” to measure the success of our CDS: The CDS Dashboards Team will develop performance reporting tools and clinical dashboards so that providers and site clinical quality staff can review adherence to CDS Consortium guidelines.
• Evaluate our work: The evaluation team will coordinate CDSC evaluation activities across all teams and act as expert consultants to each team as they develop and carry out their evaluation plans.
• Make recommendations: Based on what we learn, the CDSC will make recommendations to electronic health record vendors, clinical content vendors and regulatory and certification authorities about best practices and capabilities for decision support.
• Disseminate our results: Along with Masspro, the Massachusetts quality improvement organization, and the AHRQ National Resource Center for Healthcare IT, we will share and publish our findings and clinical decision support content and best practices developed by the CDS Consortium. We will incorporate lessons learned about CDS into online learning modules for physicians as part of Masspro’s DOQ-IT University initiative, and in the AHRQ National Resource Center.

Conclusion


The CDS Consortium is confident that working together, with Agency for Healthcare Research and Quality (AHRQ) support, significant progress towards widespread adoption of clinical decision support can be made in a short period of time. We expect to deliver first results by the end of the first year of the study. More detailed information is available upon request. Please visit our public website for CDSC study: http://www.partners.org/cird/cdsc

HL7: Clinical Decision Support

This is the presentation made to the CIC (Clinical Interoperability Council) at HL7 in April, 2009, on Clinical Decision Support.

HL7: Clinical Decision Support

View more presentations from bfm.