when i say When I say . . . distributed cognition Brian Hazlehurst
Two heads are better than one. . . I’m lost without my patient notecards. . . Residents these days don’t even know how to differentiate heart murmurs. . . These are everyday expressions commonly heard in clinical practice that indicate features of distributed cognition. Firstly, the ‘two heads’ expression conveys the idea that teams of individuals are capable of creating activity that is more robust than that which can be accomplished by an individual acting alone. Secondly, tools used by practitioners, such as patient notecards, change the nature of the cognitive work performed, often making tasks faster and easier. And thirdly, culture, which includes the institutions that train residents, changes what an individual needs to know in order to be a successful clinician. Distributed cognition is a theory that attempts to account for these everyday features of human behaviour, whether that behaviour is taking place in a hospital, on the bridge of a fishing vessel, in the cockpit of an airliner, in an office, or anywhere else a concentration of human energy occurs in the service of achieving complex activity and outcomes. Distributed cognition is a theory of human cognition that describes how information processing is distributed across people and their workplace (situated teams), their technologies (tools), and their social organisation and its influences through time (culture). Traditional theories of higher-level cognition, by which I mean reasoning, decision making and intelligent action, have focused on the individual mind as a context-free engine for intelligence. These theories have been heavily influenced
Center for Health Research, Kasier Permanente Northwest, Portland, OR, USA Medical Education 2015: 49: 755–756
by analogy with how a digital computer functions as an abstract symbol processor to achieve complex outcomes. Indeed, the image of the clinician as an isolated diagnostician has represented an archetype of higher cognition modelled by a generation of computer scientists in search of the key elements of intelligence. Of course, there is a lot that can be learned about the basic infrastructure of cognition (e.g. perception, motor skills, limits of memory, skills and strategies used) by paying careful attention to the demands placed on isolated individuals performing tasks. However, it has become increasingly clear that cognitive tasks are always highly coupled to the social and material contexts in which they are learned and performed.1 Distributed cognition provides a theoretical framework for the empirical investigation of how this works. By contrast with traditional theory, the theory of distributed cognition holds that intelligent action emerges from processes in which the brain and body are deeply embedded: successful surgery results from well-orchestrated team interactions; the clinician’s thoughts are closely coupled to her notecards, which provide reminders and support her clinical reasoning, and the resident’s skill set is organised by ever-evolving technology and values institutionalised in programmes for training new clinicians. Teams, tools and culture are not just ‘add-ons’ to symbol-processing brains – and two heads are not better than one just because they represent two abstract symbol processors instead of one – rather, the theory of distributed cognition claims that these contexts are an essential part of cognitive processing. The processing of information
Correspondence: Brian Hazlehurst, Center for Health Research, Kasier Permanente Northwest, 3800 North Interstate Avenue, Portland Oregon 97227, USA. Tel: 00 1 503 335 6349; E-mail: [email protected]
doi: 10.1111/medu.12672
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 755–756
755
B Hazlehurst is truly distributed across the brains of team-mates, from tools to brain, and so on.2 It is these loops of information processing that create intelligent action, and these loops cannot be broken down into discrete units of individual brains. The brain, of course, has an important role to play in how distributed information processing happens, but so too do the structured environment, the tools, the social organisation and the institutionalised cultural values. The theory of distributed cognition has practical implications for health care delivery and medical education. Quality improvement can apply this theory to put into place systems that make it easier to achieve desirable outcomes while reducing the ‘cognitive overload’ that results from the imposition of too many new responsibilities on the individual practitioner. Patient safety also applies this theory, designing and implementing processes that coordinate individuals with one another and their tools, and are thereby more resistant to mishap. A third important area for the application of the theory is technology design. The increasing use of electronic health records (EHRs) means that new tasks are created for clinicians, and new opportunities for sharing work (and catching mistakes) arise with ease of access to key patient information. At the same time, new opportunities for errors have emerged. For example, medication orders placed through the hospital EHR may pass drug interaction tests built into the software, but when this software is not fully integrated with that of other systems, such as the pharmacy inventory or surgery schedule, which create the information loops of distributed cognition, new errors arise.3 Finally, the theory of distributed cognition has practical implications for how we teach the next
756
generation of clinicians; it guides us toward curricula that investigate and teach about the nature of distributed action (including collaboration and teamwork) and the cognitive functions of tools and structured environments for achieving optimal patient care. Clinicians’ responsibilities increasingly require them to have skills to effectively manage information through communication, situation awareness and reasoning about complex sequences of events involving the diverse array of actors who touch the patient. Distributed cognition provides an important framework for changing how we understand intelligence and how it is achieved in clinical practice. Indeed, we are already seeing this theory at work in several aspects of health care delivery. If we can also change our medical education to emphasise the skills needed for collaboration, teamwork, information management and the practical use of helpful tools, we will better equip tomorrow’s caregivers.
REFERENCES 1 Hutchins E. Cognition in the Wild. Cambridge, MA: MIT Press 1995;1–381. 2 Hazlehurst B, Gorman PN, McMullen CK. Distributed cognition: an alternative model of cognition for medical informatics. Int J Med Inform 2008;77:226–34. 3 Koppel R, Metlay JP, Cohen A, Abaluck B, Localio AR, Kimmel SE, Strom BL. Role of computerised physician order entry systems in facilitating medication errors. JAMA 2005;293:1197–203. Received 16 September 2014; editorial comments to author 16 October 2014, accepted for publication 1 December 2014
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 755–756
Two heads are better than one. . . I’m lost without my patient notecards. . . Residents these days don’t even know how to differentiate heart murmurs. . . These are everyday expressions commonly heard in clinical practice that indicate features of distributed cognition. Firstly, the ‘two heads’ expression conveys the idea that teams of individuals are capable of creating activity that is more robust than that which can be accomplished by an individual acting alone. Secondly, tools used by practitioners, such as patient notecards, change the nature of the cognitive work performed, often making tasks faster and easier. And thirdly, culture, which includes the institutions that train residents, changes what an individual needs to know in order to be a successful clinician. Distributed cognition is a theory that attempts to account for these everyday features of human behaviour, whether that behaviour is taking place in a hospital, on the bridge of a fishing vessel, in the cockpit of an airliner, in an office, or anywhere else a concentration of human energy occurs in the service of achieving complex activity and outcomes. Distributed cognition is a theory of human cognition that describes how information processing is distributed across people and their workplace (situated teams), their technologies (tools), and their social organisation and its influences through time (culture). Traditional theories of higher-level cognition, by which I mean reasoning, decision making and intelligent action, have focused on the individual mind as a context-free engine for intelligence. These theories have been heavily influenced
Center for Health Research, Kasier Permanente Northwest, Portland, OR, USA Medical Education 2015: 49: 755–756
by analogy with how a digital computer functions as an abstract symbol processor to achieve complex outcomes. Indeed, the image of the clinician as an isolated diagnostician has represented an archetype of higher cognition modelled by a generation of computer scientists in search of the key elements of intelligence. Of course, there is a lot that can be learned about the basic infrastructure of cognition (e.g. perception, motor skills, limits of memory, skills and strategies used) by paying careful attention to the demands placed on isolated individuals performing tasks. However, it has become increasingly clear that cognitive tasks are always highly coupled to the social and material contexts in which they are learned and performed.1 Distributed cognition provides a theoretical framework for the empirical investigation of how this works. By contrast with traditional theory, the theory of distributed cognition holds that intelligent action emerges from processes in which the brain and body are deeply embedded: successful surgery results from well-orchestrated team interactions; the clinician’s thoughts are closely coupled to her notecards, which provide reminders and support her clinical reasoning, and the resident’s skill set is organised by ever-evolving technology and values institutionalised in programmes for training new clinicians. Teams, tools and culture are not just ‘add-ons’ to symbol-processing brains – and two heads are not better than one just because they represent two abstract symbol processors instead of one – rather, the theory of distributed cognition claims that these contexts are an essential part of cognitive processing. The processing of information
Correspondence: Brian Hazlehurst, Center for Health Research, Kasier Permanente Northwest, 3800 North Interstate Avenue, Portland Oregon 97227, USA. Tel: 00 1 503 335 6349; E-mail: [email protected]
doi: 10.1111/medu.12672
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 755–756
755
B Hazlehurst is truly distributed across the brains of team-mates, from tools to brain, and so on.2 It is these loops of information processing that create intelligent action, and these loops cannot be broken down into discrete units of individual brains. The brain, of course, has an important role to play in how distributed information processing happens, but so too do the structured environment, the tools, the social organisation and the institutionalised cultural values. The theory of distributed cognition has practical implications for health care delivery and medical education. Quality improvement can apply this theory to put into place systems that make it easier to achieve desirable outcomes while reducing the ‘cognitive overload’ that results from the imposition of too many new responsibilities on the individual practitioner. Patient safety also applies this theory, designing and implementing processes that coordinate individuals with one another and their tools, and are thereby more resistant to mishap. A third important area for the application of the theory is technology design. The increasing use of electronic health records (EHRs) means that new tasks are created for clinicians, and new opportunities for sharing work (and catching mistakes) arise with ease of access to key patient information. At the same time, new opportunities for errors have emerged. For example, medication orders placed through the hospital EHR may pass drug interaction tests built into the software, but when this software is not fully integrated with that of other systems, such as the pharmacy inventory or surgery schedule, which create the information loops of distributed cognition, new errors arise.3 Finally, the theory of distributed cognition has practical implications for how we teach the next
756
generation of clinicians; it guides us toward curricula that investigate and teach about the nature of distributed action (including collaboration and teamwork) and the cognitive functions of tools and structured environments for achieving optimal patient care. Clinicians’ responsibilities increasingly require them to have skills to effectively manage information through communication, situation awareness and reasoning about complex sequences of events involving the diverse array of actors who touch the patient. Distributed cognition provides an important framework for changing how we understand intelligence and how it is achieved in clinical practice. Indeed, we are already seeing this theory at work in several aspects of health care delivery. If we can also change our medical education to emphasise the skills needed for collaboration, teamwork, information management and the practical use of helpful tools, we will better equip tomorrow’s caregivers.
REFERENCES 1 Hutchins E. Cognition in the Wild. Cambridge, MA: MIT Press 1995;1–381. 2 Hazlehurst B, Gorman PN, McMullen CK. Distributed cognition: an alternative model of cognition for medical informatics. Int J Med Inform 2008;77:226–34. 3 Koppel R, Metlay JP, Cohen A, Abaluck B, Localio AR, Kimmel SE, Strom BL. Role of computerised physician order entry systems in facilitating medication errors. JAMA 2005;293:1197–203. Received 16 September 2014; editorial comments to author 16 October 2014, accepted for publication 1 December 2014
ª 2015 John Wiley & Sons Ltd. MEDICAL EDUCATION 2015; 49: 755–756
