There is a lot of talk about data. But as Philip Scott said in the opening session of the MCBK conference, “Data does not save lives, knowledge can.” Matt Samore in our seminar Friday showed a data-knowledge-wisdom pyramid where data is in the “know nothing” category and described how COVID illustrated the hierarchy..
Jeremy Wyatt in the opening keynote said that clinicians have internal knowledge and external knowledge, and the best way to provide that external knowledge is through decision support systems. But there is a long path in an LHS to generating knowledge, with bias and confounding effects weakening the knowledge we generate (see figure below). We experienced many of the confounders in the D2K portion of the LHS short course.
MCBK Conference Opening Remarks by Chuck Friedman and Philip Scott
We have a lot of examples of biomedical knowledge that come from analyzing data or from experts deliberating, and they vary in how much work it takes to get them into practice:
Policies
Decision trees
Best practices (guidelines)
Causal/propositional networks
Computable phenotypes
Predictive/explanatory models
There are two ways to represent the knowledge: human readable in words, pictures, and equations or computable (machine-implementable).
Knowledge is the keystone that holds the LHS cycle together, and because it takes 17 years to translate knowledge to practice, knowledge is the bottleneck for achieving the LHS. Closing the gap requires knowledge that can be directly executed in a computer and infrastructure to curate, manage, and disseminate it.
This is a clarion call for our Centre. In addition to partnering with health services to translate their data to knowledge, how do we help mobilize that knowledge into more computable forms that can be shared across organizations to benefit patients more quickly? Send me your ideas!
Regional Linguistic Quirks (RLQ): Interacting with service providers on the phone or at the counter is often a less stressful experience for me here than in the US because of the much more frequent use of “no drama”, “no worries” or phrases like “beautiful” when I respond to a simple request like giving them my order.
Comments