Band 7, Heft 3 - Special issue: Situativity: A Family of Social Cognitive Theories for Clinical Reasoning and Error; Issue Editors: Steven Durning, Eric Holmboe, Mark L. Graber

The diagnostic error crisis suggests a shift in how we view clinical reasoning and may be vital for transforming how we view clinical encounters. Building upon the literature, we propose clinical reasoning and error are context-specific and proceed to advance a family of theories that represent a model outlining the complex interplay of physician, patient, and environmental factors driving clinical reasoning and error. These contemporary social cognitive theories (i.e. embedded cognition, ecological psychology, situated cognition, and distributed cognition) can emphasize the dynamic interactions occurring amongst participants in particular settings. The situational determinants that contribute to diagnostic error are also explored.

Traditional teaching and assessment of clinical reasoning has focused on the individual clinician because of the preeminence of the information processing (IP) theory perspective. The clinician’s mind has been viewed as the main source of effective or ineffective reasoning, and other participants, the environment and their interactions have been largely ignored. A social cognitive theoretical lens could enhance our understanding of how reasoning and error and the environment are linked. Therefore, a new approach in which the clinical reasoning process is situated and examined within the context may be required. The theories of embodied cognition, ecological psychology, situated cognition (SitCog) and distributed cognition (DCog) offer new insights to help the teacher and assessor enhance the quality of clinical reasoning instruction and assessment. We describe the teaching and assessment implications of clinical reasoning and error through the lens of this family of theories. Direct observation in different contexts focused on individual and team performance, simulation (with or without enhancement of technology), stimulated recall, think-aloud, and modeling are examples of teaching and assessment strategies grounded in this family of social cognitive theories. Educators may consider the instructional design of learning environments and educational tools that promote a situated educational approach to the teaching and assessment of clinical reasoning.

Although assessing clinical reasoning is almost universally considered central to medical education it is not a straightforward issue. In the past decades, our insights into clinical reasoning as a phenomenon, and consequently the best ways to assess it, have undergone significant changes. In this article, we describe how the interplay between fundamental research, practical applications, and evaluative research has pushed the evolution of our thinking and our practices in assessing clinical reasoning.

The genealogy of graduate medical education in America begins at the bedside. However, today’s graduate medical trainees work in a training environment that is vastly different from medical training a century ago. The goal of the Graduate Medical Education Laboratory (GEL) Study, supported by the American Medical Association’s (AMA) “Reimagining Residency” initiative, is to determine the factors in the training environment that most contribute to resident well-being and developing diagnostic skills. We believe that increasing time at the bedside will improve clinical skill, increase professional fulfillment, and reduce workplace burnout. Our graduate medical education laboratory will test these ideas to understand which interventions can be shared among all training programs. Through the GEL Study, we aim to ensure resident readiness for practice as we understand, then optimize, the learning environment for trainees and staff.

In a recent study using direct observation of physicians, we demonstrated that physician-generated clinical documentation is vulnerable to error. In fact, we found that physicians consistently overrepresented their actions in certain areas of the medical record, such as the physical examination. Because of our experiences carrying out this study, we believe that certain investigations, particularly those evaluating physician behavior, should not rely on documentation alone. Investigators seeking to evaluate physician behavior should instead consider using observation to obtain objective information about occurrences in the patient-physician encounter. In this article, we describe our experiences using observation, and we offer investigators our perspectives related to study design and ethical questions to consider when performing similar work.

While much attention has been given to quantifying errors in diagnosis, how best to study the process of diagnosis is less clear. Focused ethnography as a methodology is particularly valuable for studying healthcare processes because it examines specific questions, situations or problems among a smaller group of individuals. In this paper, we review this approach and illustrate how we applied it to study diagnostic errors in hospitalized patients.

This paper uses novel qualitative research methods (phenomenology, ethnography and enactivism) to understand the cognitive processes through which radiologists interpret medical images to arrive at a diagnosis. From this perspective, diagnosis is not simply a matching of findings to retrieved mental images, but more properly an act of embodied or situated cognition, one that involves perception along with the actualization of professional memory and imagination and an expert-level understanding of the involved technology. Image interpretation involves a diverse set of factors, each of which is critical to arriving at the correct diagnostic interpretations, and conversely, may be the source of mis-interpretations and diagnostic error. Interpretation depends on the radiologist’s understanding of the imaging modality that was used, a deep appreciation of anatomy and comprehensive knowledge of relevant diseases and how they manifest in medical imaging. A range of personal and inter-personal factors may also come into play, including understanding the actions, values and goals of the patient, the imaging technicians and the clinicians and other medical professionals involved in the patient’s care. This multi-dimensional perspective provides novel insights regarding the cognitive aspects of diagnostic radiology and a novel framework for understanding how diagnostic errors arise in this process. Some of the findings of this research may have applications for diagnostic praxis in general, that is, beyond radiology diagnostics.

Background Clinical reasoning performance assessment is challenging because it is a complex, multi-dimensional construct. In addition, clinical reasoning performance can be impacted by contextual factors, leading to significant variation in performance. This phenomenon called context specificity has been described by social cognitive theories. Situated cognition theory, one of the social cognitive theories, posits that cognition emerges from the complex interplay of human beings with each other and the environment. It has been used as a valuable conceptual framework to explore context specificity in clinical reasoning and its assessment. We developed a conceptual model of clinical reasoning performance assessment based on situated cognition theory. In this paper, we use situated cognition theory and the conceptual model to explore how this lens alters the interpretation of articles or provides additional insights into the interactions between the assessee, patient, rater, environment, assessment method, and task. Methods We culled 17 articles from a systematic literature search of clinical reasoning performance assessment that explicitly or implicitly demonstrated a situated cognition perspective to provide an “enriched” sample with which to explore how contextual factors impact clinical reasoning performance assessment. Results We found evidence for dyadic, triadic, and quadratic interactions between different contextual factors, some of which led to dramatic changes in the assessment of clinical reasoning performance, even when knowledge requirements were not significantly different. Conclusions The analysis of the selected articles highlighted the value of a situated cognition perspective in understanding variations in clinical reasoning performance assessment. Prospective studies that evaluate the impact of modifying various contextual factors, while holding others constant, can provide deeper insights into the mechanisms by which context impacts clinical reasoning performance assessment.

Developing valid assessment approaches to clinical reasoning performance has been challenging. Situated cognition theory posits that cognition (e.g. clinical reasoning) emerges from interactions between the clinician and situational (contextual) factors and recognizes an opportunity to gain deeper insights into clinical reasoning performance and its assessment through the study of these interactions. The authors apply situated cognition theory to develop a conceptual model to better understand the assessment of clinical reasoning. The model highlights how the interactions between six contextual factors, including assessee, patient, rater, and environment, assessment method, and task, can impact the outcomes of clinical reasoning performance assessment. Exploring the impact of these interactions can provide insights into the nature of clinical reasoning and its assessment. Three significant implications of this model are: (1) credible clinical reasoning performance assessment requires broad sampling of learners by expert raters in diverse workplace-based contexts; (2) contextual factors should be more explicitly defined and explored; and (3) non-linear statistical models are at times necessary to reveal the complex interactions that can impact clinical reasoning performance assessment.

Background Depression is substantially underdiagnosed in primary care, despite recommendations for screening at every visit. We report a secondary analysis focused on depression of a recently completed study using unannounced standardized patients (USPs) to measure and improve provider behaviors, documentation, and subsequent claims for real patients. Methods Unannounced standardized patients presented incognito in 217 visits to 59 primary care providers in 22 New Jersey practices. We collected USP checklists, visit audio recordings, and provider notes after visits; provided feedback to practices and providers based on the first two visits per provider; and compared care and documentation behaviors in the visits before and after feedback. We obtained real patient claims from the study practices and a matched comparison group and compared the likelihood of visits including International Classification of Diseases, 10th Revision ( ICD-10 ) codes for depression before and after feedback between the study and comparison groups. Results Providers significantly improved in their rate of depression screening following feedback [adjusted odds ratio (AOR), 3.41; 95% confidence interval (CI), 1.52–7.65; p = 0.003]. Sometimes expected behaviors were documented when not performed. The proportion of claims by actual patients with depression-related ICD-10 codes increased significantly more from prefeedback to postfeedback in the study group than in matched control group (interaction AOR, 1.41; 95% CI, 1.32–1.50; p < 0.001). Conclusions Using USPs, we found significant performance issues in diagnosis of depression, as well as discrepancies in documentation that may reduce future diagnostic accuracy. Providing feedback based on a small number of USP encounters led to some improvements in clinical performance observed both directly and indirectly via claims.

Background Situated cognition theory argues that thinking is inextricably situated in a context. In clinical reasoning, this can lead to context specificity : a physician arriving at two different diagnoses for two patients with the same symptoms, findings, and diagnosis but different contextual factors (something beyond case content potentially influencing reasoning). This paper experimentally investigates the presence of and mechanisms behind context specificity by measuring differences in clinical reasoning performance in cases with and without contextual factors. Methods An experimental study was conducted in 2018–2019 with 39 resident and attending physicians in internal medicine. Participants viewed two outpatient clinic video cases (unstable angina and diabetes mellitus), one with distracting contextual factors and one without. After viewing each case, participants responded to six open-ended diagnostic items (e.g. problem list, leading diagnosis) and rated their cognitive load. Results Multivariate analysis of covariance (MANCOVA) results revealed significant differences in angina case performance with and without contextual factors [Pillai’s trace = 0.72, F = 12.4, df =(6, 29), p  <  0.001, ηp2=0.72$\eta _{\rm p}^2 = 0.72$], with follow-up univariate analyses indicating that participants performed statistically significantly worse in cases with contextual factors on five of six items. There were no significant differences in diabetes cases between conditions. There was no statistically significant difference in cognitive load between conditions. Conclusions Using typical presentations of common diagnoses, and contextual factors typical for clinical practice, we provide ecologically valid evidence for the theoretically predicted negative effects of context specificity (i.e. for the angina case), with large effect sizes, offering insight into the persistence of diagnostic error.

Objectives Diagnostic reasoning has been shown to be influenced by a prior similar patient case. However, it is unclear whether this process influences diagnostic error rates or whether clinicians at all experience levels are equally susceptible. The present study measured the influence of specific prior exposure and experience level on diagnostic accuracy. Methods To create the experience of prior exposure, participants (pre-clerkship medical students, emergency medicine residents, and faculty) first verified diagnoses of clinical vignettes. The influence of prior exposures was measured using equiprobable clinical vignettes; indicating two diagnoses. Participants diagnosed equiprobable cases that were: 1) matched to exposure cases (in one of three conditions: a) similar patient features, similar clinical features; b) dissimilar patient features, similar clinical features; c) similar patient features, dissimilar clinical features), or 2) not matched to any prior case (d) no exposure). A diagnosis consistent with a matched exposure case was scored correct. Cases with no prior exposure had no matched cases, hence validated the equiprobable design. Results Diagnosis A represented 47% of responses in condition d , but there was no influence of specific similarity of patient characteristics for Diagnosis A, F(3,712)=7.28, p=0.28 or Diagnosis B, F(3,712)=4.87, p=0.19. When re-scored based on matching both equiprobable diagnoses, accuracy was high, but favored faculty (n=40) 98%, and residents (n=39) 98% over medical students (n=32) 85%, F(2,712)=35.6, p<0.0001. Accuracy for medical students was 84, 87, 94, and 73% for conditions a–d, respectively, interaction F(2,712)=3.55, p<0.002. Conclusions The differential diagnosis of pre-clerkship medical students improved with prior exposure, but this was unrelated to specific case or patient features. The accuracy of medical residents and staff was not influenced by prior exposure.

Background The literature suggests that affect, higher-level cognitive processes (e.g. decision-making), and agency (the capacity to produce an effect) are important for reasoning; however, we do not know how these factors respond to context. Using situated cognition theory as a framework, and linguistic tools as a method, we explored the effects of context specificity [a physician seeing two patients with identical presentations (symptoms and findings), but coming to two different diagnoses], hypothesizing more linguistic markers of cognitive load in the presence of contextual factors (e.g. incorrect diagnostic suggestion). Methods In this comparative and exploratory study, 64 physicians each completed one case with contextual factors and one without. Transcribed think-aloud reflections were coded by Linguistic Inquiry and Word Count (LIWC) software for markers of affect, cognitive processes, and first-person pronouns. A repeated-measures multivariate analysis of variance was used to inferentially compare these LIWC categories between cases with and without contextual factors. This was followed by exploratory descriptive analysis of subcategories. Results As hypothesized, participants used more affective and cognitive process markers in cases with contextual factors and more I/me pronouns in cases without. These differences were statistically significant for cognitive processing words but not affective and pronominal words. Exploratory analysis revealed more negative emotions, cognitive processes of insight, and third-person pronouns in cases with contextual factors. Conclusions This study exposes linguistic differences arising from context specificity. These results demonstrate the value of a situated cognition view of patient encounters and reveal the utility of linguistic tools for examining clinical reasoning.

Background The cognitive pathways that lead to an accurate diagnosis and efficient management plan can touch on various clinical reasoning tasks (1). These tasks can be employed at any point during the clinical reasoning process and though the four distinct categories of framing, diagnosis, management, and reflection provide some insight into how these tasks map onto clinical reasoning, much is still unknown about the task-based clinical reasoning process. For example, when and how are these tasks typically used? And more importantly, do these clinical reasoning task processes evolve when patient encounters become complex and/or challenging (i.e. with contextual factors)? Methods We examine these questions through the lens of situated cognition, context specificity, and cognitive load theory. Sixty think-aloud transcripts from 30 physicians who participated in two separate cases – one with a contextual factor and one without – were coded for 26 clinical reasoning tasks (1). These tasks were organized temporally, i.e. when they emerged in their think-aloud process. Frequencies of each of the 26 tasks were aggregated, categorized, and visualized in order to analyze task category sequences. Results We found that (a) as expected, clinical tasks follow a general sequence, (b) contextual factors can distort this emerging sequence, and (c) the presence of contextual factors prompts more experienced physicians to clinically reason similar to that of less experienced physicians. Conclusions These findings add to the existing literature on context specificity in clinical reasoning and can be used to strengthen teaching and assessment of clinical reasoning.

Objectives Diagnostic error is a growing concern in U.S. healthcare. There is mounting evidence that errors may not always be due to knowledge gaps, but also to context specificity : a physician seeing two identical patient presentations from a content perspective (e.g., history, labs) yet arriving at two distinct diagnoses. This study used the lens of situated cognition theory – which views clinical reasoning as interconnected with surrounding contextual factors – to design and test an instructional module to mitigate the negative effects of context specificity. We hypothesized that experimental participants would perform better on the outcome measure than those in the control group. Methods This study divided 39 resident and attending physicians into an experimental group receiving an interactive computer training and “think-aloud” exercise and a control group, comparing their clinical reasoning. Clinical reasoning performance in a simulated unstable angina case with contextual factors (i.e., diagnostic suggestion) was determined using performance on a post-encounter form (PEF) as the outcome measure. The participants who received the training and did the reflection were compared to those who did not using descriptive statistics and a multivariate analysis of covariance (MANCOVA). Results Descriptive statistics suggested slightly better performance for the experimental group, but MANCOVA results revealed no statistically significant differences (Pillai’s Trace=0.20, F=1.9, df =[4, 29], p=0.15). Conclusions While differences were not statistically significant, this study suggests the potential utility of strategies that provide education and awareness of contextual factors and space for reflective practice.

Objectives Uncertainty is common in clinical reasoning given the dynamic processes required to come to a diagnosis. Though some uncertainty is expected during clinical encounters, it can have detrimental effects on clinical reasoning. Likewise, evidence has established the potentially detrimental effects of the presence of distracting contextual factors (i.e., factors other than case content needed to establish a diagnosis) in a clinical encounter on clinical reasoning. The purpose of this study was to examine how linguistic markers of uncertainty overlap with different clinical reasoning tasks and how distracting contextual factors might affect physicians’ clinical reasoning process. Methods In this descriptive exploratory study, physicians participated in a live or video recorded simulated clinical encounter depicting a patient with unstable angina with and without contextual factors. Transcribed think-aloud reflections were coded using Goldszmidt’s clinical reasoning task typology (26 tasks encompassing the domains of framing, diagnosis, management, and reflection) and then those coded categories were examined using linguistic markers of uncertainty (e.g., probably, possibly, etc.). Results Thirty physicians with varying levels of experience participated. Consistent with expectations, descriptive analysis revealed that physicians expressed more uncertainty in cases with distracting contextual factors compared to those without. Across the four domains of reasoning tasks, physicians expressed the most uncertainty in diagnosis and least in reflection. Conclusions These results highlight how linguistic markers of uncertainty can shed light on the role contextual factors might play in uncertainty which can lead to error and why it is essential to find ways of managing it.

Teamwork is fundamental for high-quality clinical reasoning and diagnosis, and many different individuals are involved in the diagnostic process. However, there are substantial gaps in how these individuals work as members of teams and, often, work is done in parallel, rather than in an integrated, collaborative fashion. In order to understand how individuals work together to create knowledge in the clinical context, it is important to consider social cognitive theories, including situated cognition and distributed cognition. In this article, the authors describe existing gaps and then describe these theories as well as common structures of teams in health care and then provide ideas for future study and improvement.

Objectives While the need to address patients’ social determinants of health (SDoH) is widely recognized, less is known about physicians’ actual clinical problem-solving when it comes to SDoH. Do physicians include SDoH in their assessment strategy? Are SDoH incorporated into their diagnostic thinking and if so, do they document as part of their clinical reasoning? And do physicians directly address SDoH in their “solution” (treatment plan)? Methods We used Unannounced Standardized Patients (USPs) to assess internal medicine residents’ clinical problem solving in response to a patient with asthma exacerbation and concern that her moldy apartment is contributing to symptoms – a case designed to represent a clear and direct link between a social determinant and patient health. Residents’ clinical practices were assessed through a post-visit checklist and systematic chart review. Patterns of clinical problem solving were identified and then explored, in depth, through review of USP comments and history of present illness (HPI) and treatment plan documentation. Results Residents fell into three groups when it came to clinical problem-solving around a housing trigger for asthma: those who failed to ask about housing and therefore did not uncover mold as a potential trigger ( neglectors – 21%; 14/68); those who asked about housing in negative ways that prevented disclosure and response ( negative elicitors – 24%, 16/68); and those who elicited and explored the mold issue ( full elicitors – 56%; 28/68). Of the full elicitors 53% took no further action, 26% only documented the mold; and 21% provided resources/referral. In-depth review of USP comments/explanations and residents’ notes (HPI, treatment plan) revealed possible influences on clinical problem solving. Failure to ask about housing was associated with both contextual factors (rushed visit) and interpersonal skills (not fully engaging with patient) and with possible differences in attention (“known” vs. unknown/new triggers, usual symptoms vs. changes, not attending to relocation, etc.,). Use of close-ended questions often made it difficult for the patient to share mold concerns. Negative responses to sharing of housing information led to missing mold entirely or to the patient not realizing that the physician agreed with her concerns about mold. Residents who fully elicited the mold situation but did not take action seemed to either lack knowledge or feel that action on SDoH was outside their realm of responsibility. Those that took direct action to help the patient address mold appeared to be motivated by an enhanced sense of urgency. Conclusions Findings provide unique insight into residents’ problem solving processes including external influences (e.g., time, distractions), the role of core communication and interpersonal skills (eliciting information, creating opportunities for patients to voice concerns, sharing clinical thinking with patients), how traditional cognitive biases operate in practice (premature closure, tunneling, and ascertainment bias), and the ways in which beliefs about expectancies and scope of practice may color clinical problem-solving strategies for addressing SDoH.

Objectives In the medical community of practice, the resident is situated in systems of professional and cultural activities. How diagnostic competence manifests in their sociocultural context is examined through situativity and systems theories. A holistic model is proposed that could examine diagnostic competence across micro (individual), meso (activity), and macro (cultural) systems. Methods Two short scenarios are presented resulting from observations of emergency medicine residents and their supervising physicians. These scenarios are analyzed using a trans-theoretical model of situativity and systems theories to understand how diagnostic competence manifests in practice (activity system). Results Assessment of diagnostic competence in a sociocultural context may require assessment of responses to contextual factors that seem immaterial to clinical reasoning. This information may signal that the resident also has the skills to identify appropriate information channels within an activity system and can accurately collect and prioritize clinical information within those channels. Therefore, the formal assessment of competent clinical reasoning performance, as a situated practice, may benefit from delineating how much of the context of an activity system is required to be competently managed and synthesized across the competency milestones. Conclusions The examination of diagnostic competence as a sociocultural practice can provide a unique and holistic examination of clinical reasoning performance and assessment.