Gersten Jonker

General discussion   187 9 equipped with learning analytic technology helps individual learners to get meaningful information on their performance and progress [113], but can also be used to compare workplace assessments before and after the introduction of an EPA curriculum [114]. This strain of research provides insight in the role of the learning environment and supervision in instruction and sequencing of activities, which may be used to optimize the efficiency of CBME. Programmatic assessment relies on frequent, subjective, low- stakes expert observations in the clinical workplace with quantitative and qualitative measures on all aspects of competence [7, 102-104, 106, 107, 115-117]. However, there is a tension between single workplace assessment’s formative purpose and summative feeling to trainees [118-120]. This tension may influence feedback and assessment seeking behavior as well as effect of formative assessments, and requires further attention. The psychometric tradition in assessment, emphasizing numeric data and reliability of measurements, hasbeendominant inanesthesiology [80, 81, 121-124]. Suchquantitative assessments often focus on one procedural skill, whereas qualitative assessments are valued in holistic workplace performance evaluation [114]. The context-dependence of competence, the dynamics of clinical practice, the subjectivity of supervisors’ judgments, and the failure to capture all aspects of performance in numbers are at odds with the strong tradition to objectively assess learners [98, 125, 126]. Recent studies in anesthesiology aim to reconcile both worlds [127-131]. For example, adding an entrustment-supervision scale increased the reliability of workplace assessments [128], as asking supervisors a prospective judgment on the supervision level is aligned with the provision and delegation of care [132]. Workplace assessment tools have different purposes and different reliabilities; a clever combination of tools needs to be sought that assesses all important facets of competence and adds up to a reliable summative judgment [130]. Additional research must establish the role of simulation-training and simulation-based assessment [67, 123] in preparing for and sequencing of clinical tasks. Research in this area should explore the relation between simulation and workplace learning and assessment. Much research aims at how to achieve the competence level that permits unsupervised practice. Less is known about how expertise continues to develop after entrustment and the role of remote or delayed supervision (and assessment) of activities entrusted to trainees. Entrustment decisions do not only confirm a level of competence, but also the right to act with a decreased level of supervision [133]. Whether this leads to privileging of trainees to act with less supervision in practice remains to be determined.