linda crane lecture 2018 - wordpress.com · 2018-02-28 · basic epistemological axioms ......
TRANSCRIPT
Text to Traci….
“I’ve decided to write the Linda Crane lecture as a speech - so I can practice it ahead of time and know that I say all that I want to say... It's certainly a different approach for me and I am not altogether comfortable with it yet; but this is certainly a different kind of lecture for me and I know that I am not altogether comfortable with it yet.”
Dr. Linda CraneHumble - Decisive - Passionate
“Humility is not synonymous with passivity or indecisiveness. One can hold beliefs passionately yet with humility. An awareness of the slipperiness of truth, the subtlety of error, and the human appetite for illusion does not mean one cannot believe and act with intensity. Acting with less than perfect knowledge is part of the risk of being human. But we should therefore reject self congratulating narrowness, always seeking a deepening and broadening of our understanding rather than a hardening of it.”
Daniel Taylor, The Myth of Certainty: The Reflective Christian and the Risk of Commitment
Knowledge based practiceSynthesis: Causal Models, Causal Knowledge
Acting with less than perfect knowledge is part of the risk of being human.
The Clinician’s Dilemma
Basic epistemological axioms
• Common Sense RealismLaw of Non contradictionLaw of CausationBasic reliability in sense perception’Analogical use of language
Knowledge and CompetenceCompetence requires knowledgeTherefore practice competence requires practice knowledge
Equivalent to:
Knowledge Based Practice
Uncontroversial:
Practice knowledge is a necessary but not a sufficient condition for practice competence
A knowledge based practice can help set the stage for practice competence
Knowledge Based Practice
• Synthesis• Building causal models• Generating inference rules based on the models
to inform practice based on that knowledge• Testing the established causal models
• Doing this all explicitly; not simply implicitly
Evidence Based Practice Won’t Work
• It will not resolve the clinician’s dilemma• It will not create perfect knowledge
Evidence Based Practice Won’t Work
• It will not resolve the clinician’s dilemma• It will not create perfect knowledge
Context –
Yes, I am caricaturizing EBP, yes I realize no one is making the above claims
Early experience
• September 1992, MS in PT student• Told we had to pursue the MS degree to help
improve the profession’s ability to generate and use evidence for practice
• But also told that an answer to my question could not be answered and that I would eventually know the answer as I gained practice experience
Logic & Epistemology• Formal logic and inference
DeductionInductionAbduction
• Common Sense RealismNon contradictionAnalogical use of languageBasic reliability of sensory perceptionCausation
• Syntopical reading – synthesis
PTJ, 2003, Invited Commentary
“What can be done to stimulate more research in physical therapy that has direct clinical relevance?”Alan Jette
2005 - letter to editor, PTJ:“The simple study of parts in isolation -reductionism - is the modus operandi of the scientific method, attempting to isolate sources of variation. Clinicians, however, are faced with all sources of variation at the same time and must deal constantly with the full burden of the complex system.”
Collins, 2005
Critical Realism
Philosophical foundation for knowledge based practice
Explicitly identifies the human mind as part of the process of knowledge
Ontology (being: the way things are) determines epistemology (knowing: the way things are known)
The way practice is, should determine the way we know about practice
Critical Realism Definition (McGrath)
“Reality is apprehended by the human mind, which attempts to express and accommodate that reality as best it can with the tools at its disposal -such as mathematical formula or mental models.”
Alistair McGrath, Scientific Theology, Volume 2: Reality
Critical Realism
Critical realism explicitly identifies the human mind as part of the process of knowledge. This makes explicit use of models to represent the reality encountered. In other words - the mental models that are constructed are knowledge and their goal is to fit with reality even when we cannot empirically verify each and every component of such models.
Critical Realism
Ontology (being: the way things are) determines epistemology (knowing: the way things are known)
The way things are – there is a differentiation and stratification of reality
Critical Realism
McGrath: “The nature of reality is such that certain things can only be known to a certain extent, and in a certain way – and that this is the reality of the situation. We are not in a position to determine whether and how things may be known: that is decided by the things themselves.”
The way practice is, should determine the way we know about practice
What about transitivity?• Evidence leads to knowledge• Knowledge informs practice• Therefore, evidence informs practice
Critical realist response: evidence does now translates to knowledge directly; evidence contributes to the transformation of knowledge as something that is ontologically distinct from what evidence can translate to.
2005 - letter to editor, PTJ:“The simple study of parts in isolation -reductionism - is the modus operandi of the scientific method, attempting to isolate sources of variation. Clinicians, however, are faced with all sources of variation at the same time and must deal constantly with the full burden of the complex system.”
Collins, 2005
Complexity – as a number of alternatives
AT ES0 00 11 01 1
22 = 4 23 = 8 24 = 16AT ES RT0 0 00 0 10 1 00 1 11 0 01 0 11 1 01 1 1
AT ES RT IMT0 0 0 00 0 0 10 0 1 00 0 1 10 1 0 00 1 0 10 1 1 00 1 1 11 0 0 01 0 0 11 0 1 01 0 1 11 1 0 01 1 0 11 1 1 01 1 1 1
25 = 3226 = 6427 = 12837 = 218747 = 16,384
SYNTHESIS: CAUSAL MODELS, CAUSAL KNOWLEDGE
THREE CLAIMSTWO PREMISESTHREE EXAMPLESONE SUMMARYONE CONCLUSION
Claim 1
Causal inferences are the most pervasive in all clinical reasoning
P e | i( ) > P(e)P e1 | i1( ) > P(e1 | i0 )
Claim 2Clinicians reason from cause to most likely effect when prescribing interventions and from effects to most likely cause during their evaluation
P i | e( ) = P(e | i) ⋅P i( )P e( )
P e | i( ) > P(e)P e1 | i1( ) > P(e1 | i0 )
Claim 3Causal inferences are central, either explicitly or implicitly, in all research as we attempt to generate understanding about cause and effect relations from our structured observations
Premise 1
Explicitly outlined causal models provide us with a framework to iteratively build and share knowledge for practice and connects practice to research as well as research to practice
Premise 2
Causal models provide frameworks upon which inference patterns can be taught, learned, practiced, and improved upon
Features (common to all examples)• Causal models are a representation of clinical
knowledge, and thus can be used to develop clinical reasoning
• Causal models represent both causal structure and degrees of belief for research and practice
• Tools such as DAGitty (open source) and SAMIAM (open source) for the development, analysis and use of physical therapy causal models in practice and research
Examples
1. Social determinants of health2. Differential diagnosis3. From Effect Sizes to Conditional probabilities
Differential Diagnosis: the abductive challenge
Deduction:Premise 1: If X, then YPremise 2: XConclusion: Therefore Y
Abduction:Premise 1: If X, then YPremise 2: YConclusion: Therefore X
Differential Diagnosis: abductive adjustment sets
• The process of differential diagnosis is essentially the process of identifying an adjustment set and testing (in some way) the members of that set
• First select the appropriate causal model for the given situation, then reason through it (implicitly or explicitly); reasoning explicitly may help reduce the risk of flawed reasoning and traps
Differential Diagnosis: abductive adjustment sets
• The abduction adjustment set must include all possible causes of the effects under consideration.
• Next step – more detail into each cause
!The adjustment set for the cardiac cause is now: {Ischemia, Pericarditis, Valve Dysfunction, Heart Failure}
Data collection and use is driven by tools
With these tools in mind –
Data collection and use can be considered
With these data products available –
We can enter dialogue that attempts to bridge the crevasse between research evidence and the complexity of practice
Effect Sizes to Conditional probabilities
Effect sizes are extremely useful
But a clinician might also value a conditional probability when an intervention (or set of interventions) is proposed
NMES for heart failure
Gomes Neto et al. Journal of cardiopulmonary rehabilitation and prevention, 100(August):1, 2016.
Suggestion
Let’s use causal models to provide a synthesisof what we know to develop causal knowledge from which to further develop with empirical evidence, and from which knowledge we can practice
Causal models represent a synthesis of knowledge for practice with a critical realist epistemology that explicitly identifies the human mind as part of the process of knowledge and where the models that are constructed fits with reality even when we cannot empirically verify each and every component of such models; but where knowledge assumptions that are encoded in the models are clear, combine knowledge with reasoning, and are subject to empirical verification when possible.
Conclusion