Keywords
patient knowledge - human factors - health informatics - patient-centered care - knowledge
- cognition - patient care
Background and Significance
Background and Significance
There has recently been more research on characterizing patient behavior within and
outside of clinical settings, mainly in the context of chronic or long-term illness,
to inform the development of human-centered technologies. Patient ergonomics, a subdiscipline
of human factors, studies the health-related work of patients, caregivers, and community
members.[1] Limited research has been concerned with “getting inside patients' heads” or patient
cognition. Cognition refers to the processes (associated with knowledge, thinking,
memory, attention) underlying outward behaviors. Patients possess unique and privileged
knowledge particularly about historical and daily experiences with their illness.
Although clinicians need this information from patients to provide patient-centered
care, patient knowledge may not be systematically elicited or documented by clinicians.
In this article, we propose a theoretical characterization of patient knowledge and
explore its relevance to care provision by applying traditional cognitive science
and human factors (a core competency of health informatics[2]) perspectives to define patients as knowledge workers across the continuum of care.
To demonstrate that patients possess and share knowledge relevant to the clinical
picture, we present vignettes about Mr. Jones.
-
Background—A patient has experience and comprehension of illness, medication side
effects, and self-care management strategies: Mr. Jones, a 50-year-old male with a master's degree, was diagnosed with multiple
myeloma 4 years ago. Mr. Jones also suffers from hypertension, for which he takes
blood pressure medication and exercises regularly. Mr. Jones' cancer has relapsed
multiple times but it has been controlled with chemotherapy and stem cell replacement.
He regularly sees his oncologist and his primary care provider. Since his diagnosis,
Mr. Jones has had several emergency room visits and hospitalizations at both his local
hospital and the academic medical center where he gets his cancer care. With the help
of his wife, Mr. Jones manages the logistics of his doctors' visits and chemotherapy
infusions, as well as medications and self-care activities. He keeps his test results
in a binder and adjusts lifestyle and activity levels according to his oncologist's
guidance and personal strategies. Recently, Mr. Jones has begun to suffer from debilitating
neuropathy caused by one of his cancer medications. He also had a stairlift installed
in his home to prevent falls due to fatigue and lightheadedness. Mr. Jones has an
informed understanding of his condition and the mechanisms of action of his medications.
He has also developed strategies for managing his life, given medication side effects.
-
Patient and caregiver communication of patient status may impact care safety and quality: While neutropenic due to chemotherapy, Mr. Jones had an acute event that lands him
in the local hospital's emergency room, rather than the academic medical center where
he receives his cancer care. The electronic health record (EHR) systems are not integrated.
The emergency room clinicians do not follow guidelines for neutropenic precautions
and prepare to conduct a series of unnecessary tests. Mr. Jones highlights his immunocompromised
status to the clinicians and asks his wife to provide blood test results from the
previous day. Clinicians put in place neutropenic precautions.
-
Strategically withholding patient knowledge may impact decision making regarding treatment
options: Mr. Jones is eligible for a clinical trial for a promising new medication. Mr. Jones
sees this clinical trial as his last hope and strategically withholds from his care
team the poor quality of life that he is experiencing at home for fear of not being
included.
-
Experiences of medication side effects may impact medication options: Mr. Jones is interested in exploring a new hypertension medication. When asked by
his primary care physician about bothersome side effects of his cancer medications,
Mr. Jones forgets to mention neuropathy. Mr. Jones is thus prescribed a medication
that exacerbates his neuropathy, thus impacting his quality of life as well as his
functionality.
Despite the impetus toward patient–clinician teaming in health care and policy for
systematic sharing of patient-generated health data,[3] research to characterize patient cognition as underlying their contributions to
the clinical information space is limited. This is a challenge for both technology
development and synchronous and asynchronous communication. The content of patient
knowledge informs thinking and underlies behaviors of managing wellness, health, and
interfacing with the health care system. Qualitative researchers like nurse and sociologist,
Corbin and Strauss have a history of capturing patient experiences with chronic illness
and characterizing activities of self-care management as “work.”[4] However, rather than answering research and applied questions of how to support
the work, their research objective was the discovery of patient experiences of work.
Designing human-centered health information technologies (HITs) to support capturing,
sharing, and integrating of patient-held knowledge requires the understanding of not
only what patients do, but when, why, and how they do it.
Human factors approaches[5]
[6]
[7] have been used to characterize decision making of professional experts in real-world
domains of aviation, nuclear power, military, and health care. Findings have informed
models of knowledge acquisition and decision making.[8]
[9]
[10] Only recently have patients become recognized as active participants and decision
makers in and outside of the health care system.[11] Despite the relevance of cognitive science and human factors literature to this
applied problem space, most theory and research on patient cognition have stopped
short of applying it.
Why Is Patient Knowledge Relevant?
Patient knowledge is relevant because of its relationship with clinical outcomes.
For instance, research by Khan et al in pediatric hospital settings highlights that
families play a role in patient safety through vigilance and reporting of potential
errors.[12]
[13] Cognitive processes associated with decision making underlie vigilance and reporting.
Work by Weiner and Schwartz[14] spanning the last decade stresses the importance of avoiding contextual mistakes
in medical decision making by physicians asking the “right” questions of patients.
By gathering contextual information from patients, particularly about life and social
constraints, physicians can better inform patient-centered treatment plans, highlighting
the value of patient contributions. This work is an example of the importance of clinician-driven
knowledge elicitation from patients.
Further, lower health literacy is associated with lower rates of preventative care,
medication management challenges, more hospitalizations, more emergency care use,
poorer health, and higher mortality rates.[15]
[16]
[17] On the other hand, patients and caregivers with higher health literacy may actively
influence their care plans through engagement and shared decision making.[18] In a BMJ Comment, Kennedy highlights the idea that patients are experts in their “experience, feelings,
fears, hopes, and desires.”[19] Thus, even in the absence of illness-specific knowledge, patients may be experts
in their own right. The term “expert patient” refers to patients who are not just
health care consumers but producers of health through deep knowledge and comprehension.[20]
[21]
[22]
[23]
[24] The use of this terminology highlights the applicability of traditional approaches
of studying expertise to patients.
Cognitive Science Perspectives on Knowledge Acquisition
Traditionally, cognitive scientists have examined the acquisition of knowledge in
service of education. In 1980, Dreyfus and Dreyfus proposed a stage model of knowledge
acquisition where a learner progresses from a rule-bound novice to an intuitive reasoner
making decisions based on tacit knowledge.[9] Generalized to nursing practice,[25] the Dreyfus model continues to be heavily cited in medicine today. In addressing
designing medical education to support knowledge acquisition, Ericsson's deliberate
practice theory is widely accepted. This theory's idea is that expert performance
is a function of intentional practice designed by educators. This practice incorporates
opportunities for problem solving, immediate feedback, evaluation, and repeat performance.[26] Ericsson's theory has been generalized to other professional domains such as transportation,
entrepreneurial skills, writing, teaching, and even music, chess, drawing, math, and
software design.[27]
In complement, human factors scientists have been seeking to understand knowledge
through theoretical perspectives such as naturalistic decision making (NDM) that accounts
for expert performance in complex real-world domains.[6] NDM provides both a theoretical perspective and a set of research tools to examine
how professionals (primarily subject matter experts or SMEs) solve problems, make
decisions, and perform cognitively and perceptually complex work. This work has traditionally
been conducted in domains marked by uncertainty, time limitations, and high consequences
for erroneous actions. Because it is not feasible to bring real-world complexity into
the laboratory, methods include interviews to elicit challenging lived incidents from
SMEs and identify and unpack decision points. Again, much of this research has been
focused on professional domains to capture expertise and deliver it to novices as
part of training, education, and decision support solutions. In health care, substantial
literature exists using the NDM perspective to characterize clinicians' knowledge
and decision making.[28]
[29]
[30]
[31] However, again, the topic of patient cognition has received little attention even
in human factors.
Of particular relevance to informatics is the understanding that many types of patient
knowledge, especially in the cases of chronic and long-term illness, are not acquired
in a deliberate manner through formal education and training, but are emergent as
a function of experience and continuity across time and space of the care continuum.
Some types of patient knowledge may be acquired through implicit learning or learning
that occurs without conscious effort,[32] and may potentially yield deep understanding. Although authors suggest that self-management
of chronic illness like diabetes “draws on the same cognitive skills found in experts
from diverse professional domains,”[33] few studies have focused on examining the acquisition and content of such knowledge.
For instance, Lippa et al[33] suggest that patient's problem detection skills and strategies acquired through
practice in diabetes self-management are associated with better adherence and greater
glycemic control, which may indicate proficient performance. Recently, Holden et al[34] used the NDM approach of incident-based interviews to characterize decision making
personas to inform the design of solutions to support patients with heart failure.
Their study highlights the need to understand cognitive processes, which are associated
with decision making to inform human-centered design.
Traditionally, there are two types of knowledge, explicit and tacit. Explicit (e.g.,
declarative, expressive) knowledge can be easily articulated, codified, and transferred
to others. For example, explicit knowledge may be acquired through book learning.
In contrast, tacit (e.g., implicit), proposed by Polanyi,[35] is acquired through practical, lived experience, and often cannot be readily articulated.
Knowledge associated with carrying out skills or tasks (e.g., riding a bike, tying
shoelaces) is termed procedural knowledge and is considered implicit. Although there
are some slight variations in the literature regarding terminology and distinctions,
we emphasize that a substantial literature exists in cognitive science and propose
the need to extend and adapt accordingly to examine and characterize patient knowledge.
Further, patients possess and manage knowledge that falls under the theoretically
accepted categories of knowledge of explicit, tacit, and procedural and additional
categories of privileged knowledge or “knowledge in the head,” such as preferences
and history. Privileged patient knowledge that is unelicited due to barriers (e.g.,
time constraints, individual differences, bias) will remain so.
Patient Is a Knowledge Worker
Patient Is a Knowledge Worker
Holden et al[11] positioned the patient, alongside the clinicians, at the center of the health care
complex sociotechnical system in SEIPS 2.0. The authors highlight that the patient
conducts “work” ranging from logistical planning and execution to medication to self-care
management, even in cases where there is no disease present. The SEIPS 2.0 framework
highlights an active patient role across patient-only activities as well as patient–clinician
teaming. A body of literature now exists examining patient work[36]
[37]
[38]
[39] focused on understanding daily contexts and activities, culminating in the idea
that investigating health care workflow is incomplete without capturing patients'
health-related activities in clinical and daily-living settings.[40] As a function of these activities, patients possess, manage, apply, create, and
share knowledge. A recent review of qualitative literature on heart failure proposes
a characterization of patient knowledge based on content, development, application,
communication, and experience, highlighting that patient knowledge is implicit, explicit,
and dynamic.[41] Through such work, patients are knowledge workers. A knowledge worker is a term that has been traditionally applied to professionals
who conduct thinking and reasoning as part of their work.[42]
[43]
What Knowledge Do Patients Possess?
Bodenheimer et al[44] summarized one role of patient knowledge during the self-management of chronic diseases
in primary care. They highlight the importance of patients knowing how to identify
their problems from their perspective, take actions to address these problems, and
adapt as circumstances change. However, only a few[45] studies investigate how a patient acquires such knowledge in the context of a complex
illness marked by a particular trajectory, and how that unfolds across time and settings.
Recently introduced SEIPS 3.0 highlights the concept of the patient journey across
space and time, with patient knowledge playing a role in processes and outcomes.[46]
Given that no taxonomy of patient knowledge exists, we propose first to apply theoretically
accepted categories of knowledge of explicit, tacit, and procedural to patient knowledge,
as shown in [Table 1]. [Table 1] contains cognitive science knowledge types and their descriptions, along with examples.
Two observations emerge: (1) multiple knowledge examples belong in the tacit category,
suggesting that patients may have trouble articulating them, and (2) there are additional
knowledge examples that potentially do not fit in this framework.
Table 1
Overview of patient knowledge based on cognitive science knowledge types
|
Cognitive science knowledge types
|
Description
|
Patient examples of knowledge application
|
|
Explicit[71]
|
Readily codified, accessed, and verbalized
|
Articulation of medication and treatment protocols, and logistics of interactions
with health care system
|
|
Tacit, implicit, intuitive[71]
|
Not codified, not easily expressed or transferred
|
Judgment and decision making regarding situation awareness, physical and psychological
baseline, change perceptions, phenomenology of symptoms, etc
|
|
Procedural[72]
|
Task performance
|
Processes and strategies for self-care management, wound care, medication administration,
etc
|
Thus, we propose additional knowledge examples. We refer to these examples as patient factors (both static and dynamic), an all-encompassing term that refers to patients' privileged
knowledge. The examples of patient factors represented in [Fig. 1] may be privileged to the patient—some (e.g., attributes) may be more systematically
captured and potentially documented in the EHR. Still, most do not even have designated
fields in the EHR. These patient factors both inform and are a function of the patient
journey across the care continuum and may also evolve (e.g., preferences) as a function
of time and experience. [Fig. 1] represents the problem space of patient knowledge that is in particular need of
research attention.
Fig. 1 Overview of patient factors.
Acquisition of Knowledge
A crucial mechanism that accounts for privileged patient knowledge is experience and
continuity. Patients (and caregivers) are the only ones with continuity across the
care continuum—time, space (home, clinical settings, contexts of daily living), and
multiple clinicians. [Fig. 2] represents the patient care continuum as a spaghetti junction—a term used to describe a complex traffic interchange.[47] Not unlike the spaghetti junction, the patient journey (along with caregivers and
primarily in the home) across the care continuum is characterized by a winding path
with a variety of barriers (construction), highs and lows in physical and mental health
(under- and overpasses), the potential for guidance from lay and clinical caregivers
such as nurse navigators (a Global Positioning System), and interactions with the
health care system requiring planning, logistics, and information exchange (exits
off the highway). We note that health care interactions comprise a fraction of time
and space within the full patient journey. We currently have no measures to account
for patient continuity, and indeed, it is not explicitly represented in the EHR.
Fig. 2 The patient is the only one that has continuity across the entire care continuum
marked by time, settings (clinical, as well as contexts of daily living), and clinicians
(2018, Alexandria Cook).
Patients may also acquire knowledge across the patient journey via formal or deliberate
mechanisms. Through patient education, the health care system focuses on providing
patients with tools to support empowerment (understanding the health care system and
engaging in behaviors that influence situations and outcomes)[48] and engagement.[49] Numerous examples exist, including: American Cancer Society[50] provides patient educational resources specific to each cancer type, as does American
Heart Association,[51] on hypertension. Individual organizations, offices, clinics, and clinicians may
provide additional patient education. A growing body of literature exists on the acceptance
of consumer HITs.[52] Human-centered consumer HIT can provide opportunities for managing, storing, and
sharing health information (e.g., heart rate, blood glucose monitoring, etc.) and
learning about self-care management. Yet, nondeliberately or implicitly acquired knowledge
continues to remain poorly understood. Brown and Duguid's organizational knowledge
acquisition[53] theory, focusing on informal and fluid knowledge acquisition, calls for a focus
on practice or application. It also highlights the value of communities in learning.
Online patient communities that serve not just as spaces for support, but spaces for
information exchange (assuming moderators filter unreliable information), come to
mind.
Implications: Applying Patient Knowledge to the Clinical Information Space and Research
Patients are responsible for navigating an increasingly complex health care system
with information distributed across numerous clinicians and staff, technologies, paper
artifacts, and physical settings. Further, through privileged knowledge (phenomenology
of symptoms, case history), patients may be the drivers of their care, as suggested
by multiple sclerosis research.[18] Some examples of ways in which patients drive their care include making decisions
within and outside of their relationships with clinicians. Traditionally, when we
talk about patient decision making, it is often focused on selecting treatment options
as a function of patient–clinician interactions and the disease context. However,
patients make other decisions that receive less research attention, which may or may
not play a role in their outcomes. The following are just some examples of decisions
that patients routinely engage in independently (or in concert with their nonclinical
caregivers): self-detecting difference in health status,[54] seeking medical care (when, how, from whom, etc.),[18] managing medications,[55] releasing privileged information strategically,[18] self-medicating with over-the-counter medicines and self-treating,[44]
[56] and requesting and receiving preventative care (seeking, etc.).[57]
[58] No matter the level of knowledge, patients are decision makers.
In this article, we highlighted the clinical value of patient knowledge. We recognize
the unprecedented complexity of studying and characterizing patient cognition and
behavior, given they are “situated” within complex sociotechnical systems. Understanding
patient cognition or the whys behind patient behaviors can support (1) the design of human-centered technologies
for patient–clinician synchronous and asynchronous information sharing, (2) patient–clinician
teaming in decision making, (3) informed and engaged patient navigation of the care
continuum, and (4) patient assessment and adoption of information, resources, and
tools for self-management of health and wellness.
To make the above a reality, patients need to take part in the research process. There
is no question that researchers have embraced the necessity of including patients
to generate and evaluate patient-facing solutions. Some examples include patient evaluation
of an application to support self-care during gastrointestinal cancer,[59] language for medication instructions,[60] and information websites.[61] A study to train patients to create a continuity of care document to share with
clinicians across systems reduced duplication of laboratory tests.[62] Further, Riggare[63] writes about the need for and the value of persons with lived experiences with illness
engaging in research. Twitter stakeholder communities, including patients and clinicians,
such as #BTSM (brain tumor social media)[64] and #BCSM (breast cancer social media),[65] (among numerous other benefits) inform research ranging from bringing together ad
hoc collaborations and developing research and operational questions, to disseminating
research findings, not to mention serving as research mechanisms. Last but not least,
using the term “patients marshaled epistemic authority,” a recent publication highlights
how it is patients, through social media, that shed research attention on “long-haul
Covid.”[66] In addition, a 2020 panel at the Human Factors in Health Care Symposium highlighted
the unique contributions of researchers bringing their patient perspectives to research.[67] Ultimately, engaging patients as partners across the research, development, and
implementation process is key to designing solutions that account for patient knowledge.
Takeaways
Patient knowledge is integral to inform patient–clinician teaming. Currently, a research
gap exists on both sides of the information exchange equation between patients and
clinicians. We presented a theoretical characterization of patient knowledge's problem
space as critical to informing patient-centered care. Future research needs to address
the application and integration of patient knowledge as part of patient–clinician
teaming and decision making.
In the absence of theory applied to informatics solutions, we may mistakenly assume
that external behavior reflects an accurate understanding. However, there lies a potential
disconnect. Studying cognition can help identify gaps and inaccuracies in understanding,
thereby informing appropriately tailored informatics solutions. The study by Lippa
and Klein on patient cognition of diabetes self-care is an example of such research.[68] There is much work to be done—examining specific illnesses, settings, and patient
life contexts. We leave the reader with the following takeaways:
-
Patients are knowledge workers. Patients possess unique knowledge that clinicians need. By defining the patient as
a knowledge worker, we can begin to develop a much-needed taxonomy to inform both
research and practice.
-
Patient knowledge impacts clinical outcomes. Patient-held knowledge is clinically relevant, stressing the importance of developing
solutions to facilitate patient knowledge sharing and integration.
-
The field of human factors offers a perspective and approaches for characterizing
and eliciting the content of patient knowledge. Based on decades of examining cognition and behavior, the field of human factors
has approaches such as cognitive task analysis (CTA) applicable to characterizing
patient cognition, as demonstrated by studies by Lippa et al[18]
[69] and Holden et al.[34] Further, literature has highlighted the complementary nature of CTA and participatory
design.[42]
-
Engaging patients as partners in research is needed. The active participation of the general public (e.g., patients, citizens, etc.) as
partners in the research process (a shift toward citizen science) is key to developing
solutions that effectively account for patient knowledge.[70]
Clinical Relevance Statement
Clinical Relevance Statement
The clinical picture is incomplete without privileged and unique knowledge from patients
about their health, particularly for complex or chronic illness. Understanding patient
knowledge is necessary to inform patient- and provider-centered HIT solutions to support
information sharing and integration.
Multiple Choice Questions
Multiple Choice Questions
-
Patient knowledge is clinically relevant because:
-
Patients may possess unique and privileged information.
-
Patients may have historical information regarding their illness.
-
Patients may be the only constant across the continuum of care.
-
All of the above.
Correct Answer: The correct answer is option d.
-
Cognitive science can offer theoretical perspectives to studying patient knowledge
because:
-
Cognitive scientists primarily conduct laboratory research.
-
Cognitive scientists have long studied knowledge acquisition.
-
It focuses on human behavior.
-
It has long focused on patient cognition in real-world settings.
Correct Answer: The correct answer is option b.
