Minding the gap: towards a shared clinical reasoning lexicon across the pre-clerkship/clerkship transition

Figures & data

Table 1. Clinical reasoning concepts included in the survey (from Gold et al., PLOS one, 17(8): e0273250).

Clinical Reasoning Concept	Definition	Example(s)
Semantic qualifier	Abstract, often binary terms that help narrow or specify the meaning of a symptom, sign, pathologic process or disease	Acute vs. chronic, moderate vs. severe, or unilateral vs. bilateral
Problem representation	Clinician’s synthesis of key discriminating aspects of the history, exam, and data; often expressed as a summary statement)	60 year-old male man with history of type 2 diabetes and hypertension with acute onset of left-sided chest pain and diaphoresis
Illness scripts	Clinician’s mental representations of clinical findings, underlying pathology, diagnostic and treatment approaches, and prognosis associated with diseases	The representative findings for croup are: age 18 months to 3 years-old with barky cough, and stridor, and presentation from October to March
Dual processing theory	Description of cognitive processes as the interplay of non-analytic and analytical reasoning approaches, described as System 1 (fast, pattern recognition, experience-based) vs. System 2 (slow, deliberate, rational)	An expert sees a swollen calf in a post-op patient and diagnoses a DVT (system 1). A novice learner looks at a swollen leg and needs to consider a variety of causes, associated with pathophysiology to consider DVT as a diagnosis (system 2).
Heuristics	Mental shortcuts or rules of thumb used subconsciously in approaching a problem	A physician immediately thinks of influenza in a patient presenting with fever during flu season
Bayesian reasoning	Calculating the post-test probability using the pre-test probability and the likelihood ratio.	With a pretest probability of 25% for pulmonary embolus and a negative D-dimer test with a LR of 0.1, the post-test probability of PE is 3%
Cognitive bias	Dispositions or preferences that can affect judgments and decisions in a subconscious manner	Narrowly focusing on a single feature (sore throat) to support the diagnostic hypothesis (e.g., Strep pharyngitis), even if other or new information refutes it (cough, lack of exudate, lack of fever)

Table 2. Importance of teaching clinical reasoning concepts in pre-clerkship clinical courses as rated by pre-clerkship clinical course directors and clerkship leaders (1-not at all important, 2-moderately or somewhat important, 3-extremely important).

	Pre-Clerkship Mean (SD) N	Pre-Clerkship Median (range)	Clerkship Mean (SD) N	Clerkship Median (range)	P value
Use of semantic qualifiers	2.72 (0.49) 100	3 (2)	2.37 (0.72) 208	2 (3)	<0.001
Use of problem representations	2.87 (0.42) 100	3 (3)	2.48 (0.69) 207	3 (3)	<0.001
Use of illness scripts	2.60 (0.62) 99	3 (3)	2.37 (0.73) 206	2 (3)	0.01
Dual-processing theory	2.38 (0.72) 100	2.5 (3)	1.86 (0.98) 207	2 (3)	<0.001
Use and limitations of heuristics	2.33 (0.70) 100	2 (3)	2.02 (0.86) 207	2 (3)	0.003
Use of Bayesian reasoning	2.27 (0.62) 100	2 (3)	2.16 (0.89) 207	2 (3)	0.84
Cognitive bias	2.69 (0.56) 100	3 (3)	2.28 (0.81) 208	2 (3)	<0.001

Table 3. Importance of teaching clinical reasoning concepts in clerkships as rated by pre-clerkship clinical course directors and clerkship leaders (1-not at all important, 2- moderately or somewhat important, 3-extremely important).

	Pre-Clerkship Mean (SD) N	Pre-Clerkship Median (range)	Clerkship Mean (SD) N	Clerkship Median (range)	P value
Use of semantic qualifiers	2.88 (0.42) 90	3 (3)	2.36 (0.73) 211	2 (3)	<0.001
Use of problem representations	2.88 (0.49) 91	3 (3)	2.69 (0.60) 211	3 (3)	<0.001
Use of illness scripts	2.81 (0.60) 91	3 (3)	2.62 (0.58) 211	3 (3)	<0.001
Dual-processing theory	2.64 (0.69) 91	3 (3)	2.04 (0.87) 211	2 (3)	<0.001
Use and limitations of heuristics	2.76 (0.58) 91	3 (3)	2.20 (0.77) 208	2 (3)	<0.001
Use of Bayesian reasoning	2.71 (0.56) 91	3 (3)	2.04 (0.88) 210	2 (3)	<0.001
Cognitive bias	2.84 (0.52) 91	3 (3)	2.51 (0.67) 209	3 (3)	<0.001

Table 4. Pre-clerkship clinical skills course director and clerkship leader perceptions about the degree to which specific barriers impede the inclusion of clinical reasoning activities in their respective phases of medical school curricula (pre-clinical course and clerkship curricula) (1-not at all an impediment, 2-somewhat of an impediment, 3-a major impediment).

	Pre-Clerkship Mean (SD) N = 94	Pre-Clerkship Median (range)	Clerkship Mean (SD) N = 208	Clerkship Median (range)	P value
Lack of curricular time to add clinical reasoning content	2.15 (0.69)	2 (2)	2.13 (0.68)	2 (2)	0.86
Perception that clinical reasoning concepts cannot be taught	1.23 (0.50)	1 (2)	1.30 (0.51)	1 (2)	0.18
Perception that clinical reasoning concepts are too advanced for pre- clerkship/clerkship students (respectively)	1.54 (0.56)	2 (2)	1.25 (0.48)	1 (2)	<0.001
Lack of faculty with skills to teach clinical reasoning concepts	1.87 (0.74)	2 (2)	2.09 (0.77)	2 (2)	0.02

Table 5. Inclusion of clinical reasoning concepts in pre-clerkship clinical courses and clerkships as rated by pre-clerkship clinical course directors (n = 94) and clerkship affiliated faculty (n = 305), respectively.

	Pre-clerkship clinical skills course directors (N = 94)				Clerkship leaders (N = 217)
Concept	Not familiar with concept % (N)	Not at all included % (N)	Moderately included % (N)	Extensively included % (N)	Not sure or unfamiliar % (N)	Not covered in the overall clerkships curriculum% (N)	Covered elsewhere in the clerkships curriculum but not in my clerkship % (N)		Covered in my clerkship % (N)
Semantic Qualifiers	2% (2)	7% (7)	52% (49)	38% (36)	11.1% (24)	11.6% (25)	15.2% (33)		62.0% (134)
Problem Representations	1% (1)	1% (1)	36% (34)	62% (58)	8.8% (19)	5.1% (11)	10.1% (22)		76.4% (165)
Illness scripts	1% (1)	9% (8)	48% (45)	43% (40)	8.8% (19)	8.8% (19)	18.5% (40)		63.9% (138)
Dual processing theory	1% (1)	24% (23)	54% (51)	21% (20)	27.1% (59)	22.1% (50)	20.6% (45)		29.4 (54)
Heuristics	1% (1)	29% (27)	54% (51)	48% (15)	22.2% (48)	15.7% (34)	31.5% (68)		39.4% (85)
Bayesian Reasoning	0% (0)	29% (27)	48% (45)	23% (22)	15.7% (34)	31.5% (68)	31.5% (68)		32.9% (71)
Cognitive bias	0% (0)	12% (11)	49% (46)	39% (37)	18.1% (39)	11.6% (25)	18.1% (39)		52.3% (113)