Thursday, October 13, 2022

Medical Cognition tools to resolve clinical complexity: System 1 (heuristics,eyeballing) and system 2 (evidence based reflective sharing) toward a dynamic ontology in healthcare

Abstract :


The entire write up below is a preliminary draft with extensive quotes borrowed from various resources to provide the audience here with a fair idea of Medical cognition, which is a broad area consisting of various system 1 and system 2 human cognitive tools to resolve clinical complexity (diagnostic and therapeutic uncertainty). These tools are often used through various medical cognitive platforms such as synchronous face to face interactions (often system 1) and asynchronous communication and learning between multiple stakeholders in connected web space (user driven healthcare UDHC, patient journey records PaJR) and blended to form "case based blended learning ecosystems CBBLE (often a blend of system 1 and 2). 

Introduction :

"Medical cognition research looks at the psychological processes underlying healthcare performance, focusing on in-depth analysis of the perceptual and cognitive processes that lead to observable behavior in all healthcare stakeholders. The focus is on understanding the knowledge structures and mental processes in healthcare brought to bear during cognitive activity (e.g., problem solving and decision making)."(Patel 2001). More about our past work around it here: http://medicinedepartment.blogspot.com/2021/06/evolution-of-model-forpatient-centered.html?m=1



Clinical complexity consists of a few defining characteristics such as uncertainty, non linearity, unpredictability and yet an overall pattern leading to resolution through attractor states over time. (Greenhalgh 2001) As physician attractors we are uniquely privileged to "be" with our patients regardless of the diagnosis and that is the only way we may know our patient's outcomes where our "being" with them is the most significant (and often overlooked) intervention. 


"Decision-making is complex. It is partly based on the dual-process theory of Epstein and Hammond, recently popularized in Daniel Kahneman’s book “Thinking Fast and Slow.” Two families of cognitive operations, called System 1 (intuitive) and System 2 (analytical) thinking, are used in decision-making. System 1 thinking is often described as a reflex system, which is “intuitive” and “experiential” or “pattern recognition”, which triggers an automated mode of thinking. 

System 2 is the more “analytical,” “deliberate” and “rational” side to the thinking process. It is pieced together by logical judgment and a mental search for additional information acquired through past learning and experience. The data are then processed carefully, through a conscious application of rules, making it a much slower and cognitively demanding process but more likely to lead to better decisions. The analytical system is engaged usually when there is uncertainty, complexity, or the outcomes give little room for error but there is time to think."




We regularly use "medical cognition" system 1 and system 2 tools to tackle clinical complexity and some of these are are often used through various medical cognitive platforms such as synchronous face to face interactions (often system 1) and asynchronous communication and learning between multiple stakeholders in connected web space (user driven healthcare UDHC, patient journey records PaJR) and blended offline and online to form "case based blended learning ecosystems CBBLE (often a blend of system 1 and 2).

Glossary of terms :

user driven healthcare UDHC : Subset of "Medical Cognition' globally where multiple users, all healthcare stakeholders including patients, interact online to understand and take decisions on meeting patient requirements. 



Here's about how it transformed into the current CBBLE since 2017 at Narketpally : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6163835/

CBBLE (pronounced cable) : Case based blended learning ecosystem that is available locally in many institutions and some are connected globally to each other. CBBLE is different from UDHC in that it is not purely online but blended offline and online. 

PaJR : The key concept lies in the use of regular patient reported outcomes to locate the phase of illness in 

System 1 thinking and human visual pattern recognition through approximations rather than accurate measures :

Humans may, to quote, "share a commitment that there is one true value for any experienced visual magnitude in the world – even if our analog magnitude systems are incredibly limited in their ability to discern what value this is. Visual analog magnitudes do not represent "fuzzy" or "noisy" estimates – rather they represent precise estimates that are subject to epistemic limitations."



Although a large part of human activity and system 1 decision making happens through heuristics and visual pattern recognition (often called eyeballing), this area of human endeavor is relatively less well studied.

Here's a link to some articles and review quotes on system 1  heuristics and  medical eyeballing that we have compiled : https://userdrivenhealthcare.blogspot.com/2022/10/eyeballing-as-system-1-intuitive.html?m=1

Our current projects in Medical cognition and decision making attempt to address this. 

Projects on visual approximation and shared pattern recognition to :

Document and validate medical outcomes of reduced muscle mass and increased visceral fat in various manifestations of the metabolic syndrome in the form of non communicable diseases NCDs such as 

Diabetes 

Hypertension 

CAD, CVD, CKD etc 

Alcoholic liver disease and chronic vascular complications of alcoholism overlapping with usual vasculopathy of metabolic syndrome 

Current project plans around the above are linked here below :


System 1 visual pattern recognition of muscle mass and trunkal fat needs a database of patients with varying degrees of muscle mass and fat correlated with metabolic syndrome outcomes. Our project objective is to create that database but meanwhile here's https://www.ruled.me/visually-estimate-body-fat-percentage/,
what others have shared around normal muscle mass and visceral fat and at what level that may be assessed visually where it may begin to look abnormal hinting at disease. According to the images shared by the authors who created the link above, all current sedentary humans appear likely to develop metabolic syndrome soon if they haven't already. The images that the authors have shared appear self explanatory (system 1?) although there is no scientific system 2 validation. 

System 2 validation : Rationale for various interventions to build muscle mass and prevent metabolic syndrome :

Scientific article : πŸ‘‡




Document and validate medical outcomes of certain types of fever patterns :

Current planπŸ‘‡


Past system 2 work πŸ‘‡


System 1 attempts in this area in the past :


Creating a dynamic ontology toward faster (system 1) and safer (system 2) medical cognition in healthcare :

Here is a link : https://medicinedepartment.blogspot.com/2022/02/?m=0, to 5000 case reports (and growing daily) from our case based reasoning database dashboard that represent single thematic analysis projects to identify the system 1 and system 2  issues in each and review/audit their challenges toward developing better solutions. 


Will be looking forward to your help in classifying these individual patient case reports into topical areas that represent 
knowledge, which can be used to integrate and analyze large amounts of heterogeneous data, allowing precise classification of any subsequent patient toward better diagnosis, care management, and translational research. This will help to create a dynamic medical ontology that may have a profound impact on how healthcare is practiced today. 

More here : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6503847/ on medical ontology and one of the challenges that you may notice in our dashboard link above also mentioned here in the medical ontology article and that is, "phenotypic information about individual patients is often insufficiently detailed or inaccessible, thus obstructing the detection of similarities and the classification of patients into clinically useful groups." This is a major area that we are currently grappling with that is primarily to do with how we capture and frame the patient data and is primarily because of social political factors governing medical education today. 


References:

1) Plsek PE, Greenhalgh T. Complexity science: the challenge of complexity in health care. BMJ. 2001 Sep 15;323(7313):625–8. doi: 10.1136/bmj.323.7313.625.

2) Patel VL, Kaufman DR. Medical informatics and the science of cognition. J Am Med Inform Assoc. 1998 Nov-Dec;5(6):493-502. doi: 10.1136/jamia.1998.0050493. PMID: 9824797; PMCID: PMC61330.



No comments:

Post a Comment