What is effective supervision and how does it happen? A critical incident study David Cottrell, 1 Sue Kilminster, 2 Brian Jolly 3 & Janet Grant 4 Objectives To identify the key features of supervision from the perspectives of educational supervisors and specialist registrars. Design Critical incident study. Telephone interviews were conducted with selected informants representing a range of specialties. The sample comprised educational supervisors with an identified interest in supervision, specialist registrars and GP trainees in the Yorkshire region. Results Educational supervisors and specialist registrars were generally agreed on what constitutes effective supervision: direct supervision was seen as very important. Educational supervisors and specialist registrars had very different concerns in relation to ineffective supervision: specialist registrars were con- cerned with inadequate supervision whilst educational supervisors were concerned with failures in direct supervision and poorly performing trainees. Supervi- sion practices varied between specialties; in this study there seemed to be particular problems in anaesthesia, medicine and paediatrics. Conclusions Direct supervision and the quality of the supervisory relationship are key to effective supervision. There is a need for clear guidance on supervision and the establishment of appropriate procedures and mech- anisms to resolve difficulties relating to inadequate supervision for trainees and performing trainees. Insuf- ficient numbers of supervisors have received training in supervision. Keywords education, medical ⁄ *standards; patient care ⁄ *standards; *professional competence; mentors; interviews; England. Medical Education 2002;36:1042–1049 Introduction Clinical and educational supervision is essential at all levels of medical education 1–3 and has increasing importance for all doctors in the context of clinical governance and audit. Supervision is a complex activ- ity: it occurs in a variety of settings and has various functions and modes of delivery. It usually involves an interpersonal exchange. This complexity means that research into supervisory practice presents many meth- odological problems and adequate research methodol- ogies have yet to be established. Even the definition of supervision is unclear. Differ- ent authors place different degrees of emphasis on the relative importance of promoting professional develop- ment and ensuring patient safety. 4 This paper reports some of the findings of a recent UK Department of Health-funded project: Good supervision: guiding the clinical educator of the 21st century 5 and uses the definition of supervision generated by that study, as: ÔThe provision of guidance and feedback on matters of personal, professional and educational develop- ment in the context of the traineeÕs experience of providing safe and appropriate patient care.’ Despite the assumed importance of supervision in postgraduate medical training, empirical studies on supervision are limited and often based on non-medical clinical training, for example, nursing. There is more research examining one-to-one supervision than Ôon the job supervisionÕ. 4 Whether the findings of such research can be extrapolated to medical supervision is unclear. Kilminster and Jolly’s review 4 of the current literature 1 Academic Unit of Child and Adolescent Mental Health, School of Medicine, University of Leeds, UK 2 Medical Education Unit, School of Medicine, University of Leeds, UK 3 Centre for Medical and Health Sciences Education, Faculty of Medicine, Nursing and Health Sciences, Monash University, Victoria, Australia 4 Open University Centre for Education in Medicine, Open University, Milton Keynes, UK Correspondence: Professor David Cottrell, Academic Unit of Child & Adolescent Mental Health, School of Medicine, University of Leeds, 12A Clarendon Road, Leeds LS2 9NN, UK. Tel.: 00 44 113 295 1760; Fax: 00 44 113 295 1761; E-mail: d.j.cottrell@leeds.ac.uk Clinical and communication skills 1042 Ó Blackwell Science Ltd MEDICAL EDUCATION 2002;36:1042–1049