Effect of practice on standardised learning outcomes in simulation-based medical education William C McGaghie, 1 S Barry Issenberg, 2 Emil R Petrusa 3 & Ross J Scalese 2 OBJECTIVES This report synthesises a subset of 31 journal articles on high-fidelity simulation-based medical education containing 32 research studies drawn from a larger qualitative review published previously. These studies were selected because they present adequate data to allow for quantitative syn- thesis. We hypothesised an association between hours of practice in simulation-based medical education and standardised learning outcomes measured as weighted effect sizes. METHODS Journal articles were screened using 5 exclusion and inclusion criteria. Response data were extracted and 3 judges independently coded each study. Learning outcomes were standardised using a common metric, the average weighted effect size (AWES), due to the heterogeneity of response meas- ures in individual studies. ANOVA was used to evaluate AWES differences due to hours of practice on a high- fidelity medical simulator cast in 5 categories. The eta squared (g 2 ) statistic was used to assess the association between AWES and simulator practice hours. RESULTS There is a strong association (g 2 ¼ 0.46) between hours of practice on high-fidelity medical simulators and standardised learning outcomes. The association approximates a dose–response relation- ship. CONCLUSIONS Hours of high-fidelity simulator practice have a positive, functional relationship with standardised learning outcomes in medical educa- tion. More rigorous research methods and more stringent journal editorial policies are needed to advance this field of medical education research. KEYWORDS education, medical, undergradu- ate ⁄ *methods; teaching ⁄ *methods; United States; learning; feedback, psychological; teaching materials; clinical competence ⁄ *standards. Medical Education 2006; 40: 792–797 doi:10.1111/j.1365-2929.2006.02528.x INTRODUCTION A recently published qualitative, systematic review spanning 34 years and covering 670 peer-reviewed journal articles identified 10 features and uses of high-fidelity medical simulations in a range of spec- ialties, including anaesthesiology, cardiology and surgery, that lead to effective learning. 1 The features and uses, in order of importance, are listed here. 1 Feedback is provided during the learning experience. Knowledge of results about one’s performance stands out from the research legacy as the most important feature of simulation-based medical education in promoting effective learning. 2 Learners engage in repetitive practice. This involved focused and repetitive learner engagement in practice sessions where the intent is skill improvement, not just idle repetition. 3 The simulator is integrated into the medical curricu- lum. Simulation-based educational experiences are a routine feature of the normal educational schedule and are grounded in learner perform- ance evaluation. 4 Learners practise with increasing levels of difficulty. Learners engage in medical skills practice across a range of difficulty levels, beginning with basics and advancing to progressively higher difficulty levels based on objective measurements. simulation 1 Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA 2 Miller School of Medicine, University of Miami, Miami, Florida, USA 3 Duke University Medical Centre, Durham, North Carolina, USA Correspondence: William C McGaghie PhD, Office of Medical Education and Faculty Development, Northwestern University Feinberg School of Medicine, 3–130 Ward Building, 303 East Chicago Avenue, Chicago, Illinois 60611–3008, USA. Tel: 00 1 312 503 0174; Fax: 00 1 312 503 0840; E-mail: wcmc@northwestern.edu Ó Blackwell Publishing Ltd 2006. MEDICAL EDUCATION 2006; 40: 792–797 792