How we created virtual patient cases for primary care-based learning

Abstract

Virtual patient (VP) cases have been defined as ‘interactivecomputer simulations of real-life clinical scenarios for thepurpose of medical training, education or assessment’ (Ellawayet al. 2006). VP cases are particularly valuable for theconstruction of clinical reasoning skills, which are notoriouslychallenging to teach (Balla & Edwards 1986).We present our preliminary experiences and early appli-cation of a selection of video-enriched VP cases designed tomeet some of the specific pedagogical challenges of under-graduate education in primary care. The VP cases givestudents opportunities to practise and refine unique clinicalreasoning skills typically deployed by general practitioners(GPs) (Balla & Edwards 1986).Tailored VP cases have been created for a variety ofspecialties ranging from geriatrics to paediatrics (Orton M Gill et al. 2010). Clinical reasoning differsbetween medical specialities; in particular, GPs draw onunique strategies (Heneghan et al. 2009), which may not beexplicitly taught. VP cases can be used to multiply patientencounters providing a bank of verbal and non-verbalexperiences (Cook & Triola 2009) which may accelerateacquisition of targeted clinical reasoning skills. VP scenariosare currently used in some areas of primary care education,such as maintenance of certification examinations (AmericanBoard of Family Medicine) (Sumner & Hagen 2008) and toquantify variability in primary care physician decision making(e.g. assessing qualified GPs in the treatment of depression instandardised VPs) (Hooper et al. 2008).Diagnostic strategies employed by GPs have recently beendefined (Heneghan et al. 2009), providing some clear learningoutcomes to instruct VP case design. Little is reported in theliterature about how GPs acquire their clinical reasoning skills,but experience is likely to have an important role (Heneghanet al. 2009). During the consultation, GPs generate promptdiagnostic hypotheses (Elstein et al. 1972). Fast hypothesisgeneration often depends on prior experience allowing ‘spotdiagnosis’or‘pattern recognition’. We chose to use realpatient videos as the basis of our VP cases to provide a virtual‘prior experience’ to enable these strategies. Genuine patientvideos also offer experience in the interpretation of patientnarrative such as a ‘classic history’ or patients ‘self-reporting’ofa diagnosis which GPs also draw on to make diagnoses(Heneghan et al. 2009).Video enrichment of VP cases has previously been shownto be beneficial. One study found that users develop enhancedcognitive and meta-cognitive processing when using video-enriched cases compared to text alone (Balslev et al. 2010).The advantages of using video cases to educate non-experts indiagnostic strategies for paediatrics were recently highlighted;patient video cases were found to ‘strongly stimulate clinicaldiagnostic reasoning processes’ (Balslev et al. 2010).GPs use further strategies such as ‘restricted rule-outs’ and‘probabilistic reasoning’ to refine their initial diagnosis(Heneghan et al. 2009). The strategy of ‘restricted rule-outs’involves synthesising a short list of serious diagnoses whichmust be excluded. Key questions then elicit symptomssuggestive of these diagnoses (red flags). For ‘probabilisticreasoning’, clinicians estimate a pre-test probability of apatient having a particular pathology and then adjust thisprobability in the light of test results. Diagnostic decision treesoftware can be used to create VP cases that give virtualexperience of these real-time clinical reasoning techniques.Follow-up of and reflection on patients is a powerfuldiscipline for the learning and construction of clinicalreasoning skills (Kassirer 2010), but this is limited by short