Acquisition of clinical competence: Added value of clerkship real-life contextual experience

Abstract

Background: Medical students’ limited access to patients induces a shift of learning activities from clinical wards to classrooms.

Aim: Identify clinical competencies specifically acquired during real-life contextual clerkship added to case-based tutorials, by a prospective, controlled study.

Methods: Students entering our eight-week internal medicine (IM) clerkship attended paper case-based tutorials about 10 common presenting complaints and were assigned to an IM specialty ward. For each tutorial case, two groups of students were created: those assigned to a ward, the specialty of which was unrelated to the case (case-unrelated ward, CUW) and those assigned to a ward, the specialty of which was related to the case (case-related ward, CRW).

Results: Forty-one students (30 CUW and 11 CRW) volunteered for the study. Both groups had similar previous experiences and pre-clerkship exam scores. The CRW students collected more relevant clinical information from the patient (69% vs. 55% of expected items, p = 0.001) and elaborated charts of better quality (47% vs. 39% of expected items, p = 0.05). Clinical-knowledge mean score was similar (70%) in both groups (p = 0.92).

Conclusions: While paper-case tutorials did provide students with clinical knowledge, real contextual experience brought additional, specific competencies. This supports the preservation of clinical exposure with supervision and feedback.

Introduction

The acquisition of clinical competence is key to clinical practice. Among other competencies, clinical reasoning is particularly important in the diagnostic and decision processes as well as in patient management. The ability to collect relevant information from the patient, to recognize, organize and integrate clinical data, to elaborate relevant diagnostic hypotheses and to synthesize this process in a patient's chart reflects main features of clinical reasoning.

Several interventions exist to train students’ acquisition of clinical competence and patient management. For example, case-based tutorials in small groups (Barrows 1983; Kassirer 1983) are widely used to train the iterative process of reasoning, starting from a patient's problem and bringing students to ask for additional information and to justify its purpose. Students learn the content related to the clinical problem and the process of reasoning simultaneously in the context of a tutorial paper case. This teaching format seems to increase students’ ability to solve clinical cases, particularly when a feedback on their reasoning process is provided (Nendaz et al. 2011).

To increase clinical experience, clinical clerkships are traditionally organized after a period of theoretical activities to immerse students in stationary or ambulatory clinical contexts. Two systematic reviews (Littlewood et al. 2005; Dornan et al. 2006) showed that early immersion of medical students in a clinical setting helps their integration in a medical environment, their understanding of the medical profession and of the health care system and their ability to contextualize their knowledge. However, the assessment of its impact on their performance and their ability to manage the patient yielded contradictory results and needs further research. Other studies suggested that the number of real patient cases encountered is not necessarily a predictor of acquired clinical competence, if this experience is not associated with high-quality supervision and feedback (Ericsson 2004; Wimmers et al. 2006; Fung et al. 2007).

Although the clerkship tradition is well established, some concerns are raised because of the reduced availability of patients to students in clinical settings. In hospital settings, patient turn-over increases, due to shorter hospital stays, and the number of hospital beds tends to diminish. In ambulatory settings, logistical difficulties and concerns for efficiency might prevail over educational issues. Furthermore, social changes may lead some patients to refuse participation in students’ learning if adequate conditions are not fulfilled (Benson et al. 2005; Mol et al. 2011). Limited access to patients may induce a shift from the clinical ward to the classroom, with the assumption and in the hope that paper-case resolution during tutorials might overcome the lack of direct patient contact in a real environment.

In this prospective, controlled study we aimed at identifying – for a set of internal medicine (IM) paper cases worked up during case-base tutorials – clinical competencies specifically acquired by students exposed to a specialty ward experience related to these cases, as compared to students exposed to a specialty ward not directly related to those cases.

Table

Practice points

Interactive paper-case tutorials promote the acquisition of clinical knowledge A real clinical environment facilitates the acquisition of additional competencies pertaining to the relevance of the information collected during patient encounter and of the differential diagnosis recorded in patient chart Despite reduced patient availability, the exposure to real clinical practice cannot be replaced with paper-case tutorials

Methods

Setting

The undergraduate curriculum of the Faculty of Medicine of the University of Geneva consists of a six-year, integrated, problem-based learning program (Vu et al. 1997). The clinical curriculum in years four and five is designed to provide an integrated approach to students’ acquisition of theoretical clinical knowledge and clinical competencies. These two years consist of two four-week and four eight-week clerkship rotations in different fields of medicine. During the eight-week IM clerkship, all medical students are exposed two days a week to selected patient problems by case-based clinical reasoning tutorials. Clinical reasoning tutorials typically start with the patient's presenting complaint. The students ask the tutor for additional information about the patient's history and physical examination to evaluate the different diagnostic hypotheses raised during a sequential process, until they come to a final working diagnosis. They then discuss appropriate management of the patient (Kassirer 1983; Nendaz et al. 2011).

In addition to the paper case-based tutorials, each student is assigned to one IM ward unit, on average three days a week. Each unit preferentially admits patients in one specific specialty of IM, e.g. cardiology and respiratory diseases. On the ward, students are exposed to real patient cases during case discussions and clinical rounds. They also admit patients and are observed by supervisors during their clinical work to obtain feedback. After their patient admission, students present the case to the residents and chief residents and receive in turn additional feedback. Afterwards, they write up the patient chart and they contribute to the prescription and decisions on patient management. In addition, they participate to case discussions with the specialty consultants and may at times practice related technical procedures under supervision (e.g. ascites paracentesis). To harmonize supervision, clinical supervisors in the units have been offered workshops on precepting methods, based on Neher's microskills (Neher and Stevens 2003; Nendaz et al. 2006).

In this natural setting, all students work on the same set of paper cases during the case-based clinical reasoning tutorials outside the ward units. As those cases are frequent or exemplary problems in IM, they also encounter some of them while working on the ward. However, due to the organization of the service, some students gain additional and direct practical experience with some cases of the tutorial case set, while other students gain practical experience with other cases of that set, depending on the specialty of the ward unit they are assigned to. Thus, for each paper case learned during tutorials, two groups of students can be identified: those who experience this case during the tutorials, but are then assigned to a ward unrelated to this kind of case (case-unrelated ward, CUW) and those who gain an additional, practical experience with this case because they are assigned to a ward where such patient cases are admitted (case-related ward, CRW). This particular setting allowed us to compare dimensions of clinical expertise brought to students who were exposed to a real practical experience in a specialty related to some cases (CRW), compared to students exposed to a practical experience in a specialty not directly related to those same cases (CUW, Figure 1). Except for the type of patient cases encountered on the ward, all students benefited from the same instructional conditions in all units, as described above.

Figure 1. Study design.

Subjects and cases

Five successive rotations of about 12 students entered our IM clerkship during the 2007–2008 academic year. For the purpose of our study, we scheduled to assess in each rotation two cases out of the set of the tutorial paper cases. This led to the study of 10 different cases (‘study cases’, Figure 1) representing common and frequent IM complaints and diagnoses. At the beginning of each clerkship rotation we identified students who volunteered for the study and randomly assigned them to one of the ward units. This assignment allowed us to identify, for each study case, the CRW and CUW students. For example, if the study cases of one rotation were, respectively, chest pain and jaundice, the students assigned, respectively, to the cardiology and gastroenterology units would belong to the CRW group, while the students in the other units would belong to the CUW group. Figure 1 summarizes the study design and Table 1 displays a complete list of the study-case complaints and their related specialty wards.

Table 1.  List of complaints used for this study and their related specialty wards

Complaints	Diagnoses	Related specialty wards
Syncope	Rhythm disorder	Cardiology
Chest pain	Pericarditis	Cardiology
Increased waist	Liver cirrhosis	Gastroenterology
Generalized lymphadenopathy	Lymphoma	Haematology
Fatigue and sweating	Lymphoma	Haematology
Fever and back pain	Spondylodiscitis	Infectious diseases
Hypertension	Hyperkalaemia and renal failure	Nephrology
Cough and chronic dyspnoea	Chronic obstructive pulmonary disease	Respiratory diseases
Acute dyspnoea	Pulmonary embolism	Respiratory diseases
Impaired vision and thirst	Diabetes	Endocrinology

Outcomes and analysis

At the end of each clerkship rotation, we assessed students’ performance with a direct standardized patient (SP)-based, videotaped clinical encounter. In each session, two complaints among those listed in Table 1 were used to assess the students. Each student encountered one single SP. The SP case scripts and related diagnoses were adjusted, so that they were different enough from the paper cases studied during the case-based tutorials. One investigator (MN) wrote the SP scripts and the SPs were trained with the help of the members of our Clinical Competence Center.

All students received the study-case presenting complaint five minutes before the SP encounter. Then they had 20 minutes with the SP to elicit relevant history information, perform a targeted physical examination and conclude the encounter. Immediately after the session, the students wrote a post-encounter patient summary chart mentioning the main working diagnosis, the differential diagnosis and the management decisions (further tests, treatment and triage). They also took a 16-item clinical knowledge test related to the case. The stem of all questions consisted of short patient descriptions: the students were asked to provide their related diagnostic hypotheses or additional clinical information to be collected from the patient.

The information collected by each student during the encounter was recorded by the SPs in a pre-established checklist of relevant clinical information to be collected (25–30 items). In addition, one investigator (AR) reviewed all videotaped encounters and filled in the same checklists. Baseline agreement between SPs and the investigator was 88% (kappa 0.75). Discordances were resolved by visioning the videotapes again with another investigator as needed.

The summary chart content was assessed by using a predefined checklist (15–20 items) of expected diagnostic hypotheses in the differential diagnosis, expected main diagnosis and expected management decisions (treatment, tests or triage). The reference checklists used to assess the relevance of the collected information and of the patient chart content were previously established by a group of experienced physicians working up the same cases. The investigator assessing the video recordings and the charts was blind to the students’ group.

During their clerkships, the students kept track of the patients they admitted in a log-book. They also indicated on a questionnaire provided after the assessment session the amount of previous experience acquired on the ward with the chief complaint used for the assessment.

We compared the characteristics of the students in the CUW and CRW groups, including their biographic data, previous pre-clinical and clinical experience and exam scores. Major outcomes consisted of the students’ checklist overall relevance score pertaining to the information collected from the patient, and their overall chart relevance score. Secondary outcomes included the relevance score of each component of the patient encounter (history and physical exam) and of the summary chart (main diagnostic accuracy, relevance of the differential diagnoses, management decision accuracy regarding tests ordered, treatment and triage).

We estimated that 14 students in each group allowed the detection of a difference in relevance scores of 8% (SD 0.08) with a power of 80% when using two-tailed t-tests with an alpha value of 0.05. To take into account the fact that students in the CUW group were distributed in various specialty ward units, we aimed at recruiting about two CUW students for one CRW student. Relevance scores were compared by t-tests after assessment of normality of data and equality of variances. We used ANOVAs to identify factors predicting score relevance. SPSS 15.0 (Chicago, Inc.) was used for the analyses.

A complete ethical review was not requested by our institution and the study was approved by our Clerkship Director and IM Head. We were careful to conform to the ethical standards of the Declaration of Helsinki. At the beginning of each clerkship rotation, students were informed about the nature and objectives of the study and were offered the possibility to participate. Volunteers signed their agreement to be included in the study. Confidentiality was assured and individual data were anonymized for analysis.

Results

Forty-one students (30 in the CUW group and 11 in the CRW group) were assessed by one SP-based clinical encounter and the knowledge test. Both groups presented similar characteristics regarding their previous experience or course grades, the number of preceding clerkship rotations made, and the number of patients encountered during their present IM clerkship (Table 2). As expected, 27% (95% CI 11–43) of CUW-group students and 64% (95% CI 36–92; p = 0.02) of the CRW-group students had encountered previously at least one real patient case similar to the complaints of the study cases. Both groups achieved similar clinical knowledge scores on the written test related to the study cases (Table 2).

Table 2.  Subject characteristics

	CUW N = 30	CRW N = 11	p
Female (% students, 95% CI)	53 (36–70)	73 (43–90)	0.31
Age, (mean, SD)	22.7 (1.4)	22.7 (1.5)	0.95
Previous years completed in the same medical school (% students, 95% CI)	90 (74–97)	91 (62–98)	0.71
End-of-clerkship mark^a (mean, SD)	2.7 (0.4)	2.9 (0.3)	0.24
End-of-clerkship case presentation mark^b (mean, SD)	5.5 (0.5)	5.5 (0.5)	0.88
N patient admissions made during clerkship (mean, SD)	16 (4.4)	15 (2.5)	0.61
N clerkship rotations done before IM (mean, SD)	3.6 (2.7)	3.4 (2.4)	0.83
Introduction to clinical years unit mark^c (mean, SD)	5.5 (0.7)	5.5 (0.7)	0.85
End of preclinical year exam per cent score (mean, SD)	68.6 (7.6)	72.6 (6.3)	0.13
Clinical knowledge test about the study cases, per cent-score (mean, SD)	70 (0.10)	70 (0.11)	0.92
Study cases already encountered at least once (% students, 95% CI)	27 (11–43)	64 (36–92)	0.02

Notes: CUW: students attributed to case-unrelated specialty ward. CRW: students attributed to case-related specialty ward.

^aEnd-of clerkship assessment: scale 1–3, 3 being the best mark.

^bEnd-of clerkship case presentation assessment: scale 1–6, 6 being the best mark.

^cAssessment made at the end of the Introduction to Clinical Reasoning Unit, taking place at the beginning of the clinical years: scale 1–6, 6 being the best mark (van Gessel et al. 2003).

During the SP clinical encounters, the CRW students collected more relevant clinical information than the CUW students (total checklist scores 69% vs. 55%, p = 0.001, Table 3). This difference seemed essentially due to history-taking items.

Table 3.  Difference between groups about the relevance of collected information and of patient chart

	CUW (N = 30)		CRW (N = 11)		p
	% Score	95% CI	% Score	95% CI
Relevance of collected information (% of expected items)	55	51–59	69	62–76	0.001
History	58	53–63	73	65–81	0.005
Physical examination	51	45–57	63	50–77	0.06
Relevance of patient chart content (% of expected items)	39	34–43	47	40–54	0.05
Main diagnosis	49	34–64	62	37–87	0.34
Differential diagnosis	38	26–49	59	37–82	0.05
Investigations	43	37–50	37	26–48	0.29
Treatment	15	8–23	30	5–56	0.10
Triage	63	46–80	91	74–100	0.08

Note: CUW: students attributed to case-unrelated specialty ward. CRW: students attributed to case-related specialty ward.

The total chart relevance score (Table 3) was better for the CRW students (total chart scores 47% vs. 39%, p = 0.05). Analyses of each component suggested that the relevance of the differential diagnosis (scores 59% vs. 38%, p = 0.05) accounted mostly for this difference.

We assessed the influence of other factors on the relevance of data collection, such as gender, type of study case, sequence of clerkship rotations, age, knowledge and previous practical examination scores, total number of patient admissions made during the IM clerkship and place of training during the previous years in the curriculum. Multivariate analysis confirmed that the study group (CUW vs. CRW) remained the only independent predictor of the relevance of collected information (B coefficient 14.3, 95% CI 9.1–50.0, p = 0.004, R² of the model = 0.31).

Discussion

The specificities of clinical competence brought by real-life clinical context are not well described. This study showed that medical students who experienced a set of cases both in structured case-based tutorials and in related real ward context collected more relevant patient information and wrote more accurate charts than students who were also exposed to these cases in structured case-based tutorials but spent their clerkship in a specialty ward unrelated to these cases. These findings suggest that at this stage of training clinical immersion help students develop a deeper knowledge about a set of patient cases, enabling them to broaden their repertoire of differential diagnoses and the related information to collect during the patient encounter. This happened despite the attendance to case-based tutorials that brought seemingly the same amount of case-related clinical knowledge and clinical experience similar in terms of duration of clinical clerkship and total number of patients encountered.

Due to the increased difficulty to give students access to patients, some advocate the possibility to decrease the duration of clinical contacts between students and patients and to limit the clinical approach to tutorials and few clinical rotations, in the hope that a kind of generic competence would emerge from these experiences. This study does not support such considerations and reinforces instead the importance of immersion in real clinical context for the acquisition of clinical competence (Norman 2005). Generic competence is not supported in this study, since students spend their clinical rotation in a ward, the specialty of which was unrelated to the study cases did not perform as well as those in a ward, the specialty of which was related to the study cases.

What then does the clinical context bring in addition? The specific factors in the clinical environment that enable students’ acquisition of clinical competence are still open to further investigation but some hypotheses may be considered. Learning in a relevant context has been widely recognized as an enabling condition in some teaching formats, such as problem-based learning (Barrows 1983). However, according to Koens (Koens et al. 2005), clinical context may encompass several levels of personal involvement. The first level is cognitive, when the student has, for example, to explain the patient's symptoms. The second one is physical, when the student is on the ward and sees the patient in his hospital bed. The third one is more affective, when the student takes some responsibility for the patient. Real clinical context offers all these levels of involvement, provided that students have some opportunities to take responsibility under supervision, which was the case at our institution. Tutorials that make students solve a clinical case on paper offer the first level of involvement, while observation of patients on a ward may offer the second dimension. In this study, the access to all three levels might represent one explanation to our results, which reinforces the importance of practicing in context with deliberate practice, supervision and feedback.

In addition, the more frequent, practical contacts with similar, real patient cases under direct supervision were another potential condition to students’ learning (Ericsson 2004; Norman 2005). If this did not apparently bring more clinical knowledge, it may have brought an additional, unconscious, internalized knowledge used when facing a patient (or an SP) with similar complaints. This knowledge might be of deeper level, more anchored in pathophysiology, as suggested by the higher relevance of differential diagnoses and related information to collect during the encounter. This finding echoes the theories about the dual-process clinical reasoning brought by research in cognitive psychology (Ericsson 2004; Nendaz et al. 2005; Norman 2005; Nendaz et al. 2006; Bordage 2007; Schmidt and Rikers 2007; Elstein 2009): one component of the process includes an analytical, hypothesis-driven collection of the data from the patient and a verification of the hypotheses. This hypothetico-deductive process is the one trained during case-based tutorials in our study. However, when experience grows, a less analytical approach based on the recognition of a more global clinical picture emerges. This recognition allows the physician to raise diagnostic hypotheses to be verified by targeted data collection. Although not at an ultimate stage of immediate recognition of the whole patients’ problems, the CRW group of this study may have been on their way to better recognize the general clinical picture, raise better grounded differential diagnoses and thus collect clinical data of higher relevance.

This study presents limitations. First, it was conducted in a single institution and in a hospital-based environment, potentially preventing generalization of results. However, as our study focused more on mechanisms than conditions of learning in a clinical environment, our results may still be of general interest. Second, because one clinical unit in our institution could only accommodate one student at a time, the recruitment of students in the CRW group was slow, limiting their number in this group, as well as the number of assessment cases by student. For feasibility reason, each student only encountered one SP case, which is insufficient to draw conclusions on individuals but was deemed as satisfactory at the group level. Moreover, the case-related knowledge test was short, with a potentially limited reliability. Thus, our negative results on some outcomes must be interpreted with caution. Third, as we compared students who had already acquired clinical experience with clinical cases during the tutorials by using a hypothesis-driven reasoning approach, the magnitude of an additional effect of the clinical environment was probably more difficult to detect. A comparison with lectures or more theoretical formats than tutorials would possibly have brought more visible effects but the lessons learned from such findings would have been more trivial.

In conclusion, while structured, interactive tutorials focused on case resolution and training of clinical reasoning do provide clinical knowledge, adding additional experiences about the same cases in a related clinical environment still adds dimensions pertaining to clinical performance. The mechanisms of these findings may be related to the affective dimension brought by clinical context, when students are personally involved in patient management, and to the ability to better recognize the global clinical pictures due to more frequent case exposure, which may bring deeper knowledge and understanding of pathophysiology. These hypotheses should be tested by further investigation.

Our data do not support institutional considerations that suggest overcoming the diminished patient availability by replacing contextual experiences in real environments by tutorials or minimal clinical exposure, in the hope that a generic clinical competence would be acquired. Other strategies aiming at maintaining maximal clinical exposure with supervision and feedback should be further developed, such as the use of other structures than the traditional hospital-based environment. These structures should, however, be able to provide intended, deliberate and coordinated exposure of students to both theoretical and practical experiences, with supervision and feedback. This approach may turn out more efficient than strategies relying solely on the length of training and the number of patients to encounter.

Acknowledgements

We thank the students who volunteered for this study and the physicians who contributed to create the gold standards for the cases used. We acknowledge Dr Florence Demaurex and Mrs Jacqueline Laurin of the Clinical Competence Center for their help in assessing the students.

Conflicts of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

This study was partially funded by the University of Geneva Funds Edmond Safra, the Research Fund of the Department of Internal Medicine of the University Hospital and the Faculty of Medicine of Geneva (unrestricted grant from AstraZeneca Switzerland) and the Foundation Gabriella Giorgi-Cavaglieri.