Can virtual reality be used to measure and train surgical skills?

The quantitative literature on the use of virtual environments to measure and train a variety of surgical skills is critically reviewed. We selected works from the years 1995-2000. Theoretical perspectives, such as those of Saltzman (1979), Bernstein (1967) and Schmidt (1975) and techniques, such as hierarchical task analysis, are presented and contrasted with the largely atheoretical approach of the practitioners of virtual surgery. It is concluded that the quantitative work discussed provides few findings of value to practising surgeons. This may be due in part to the lack of consideration paid to fundamental issues in the learning of motor skills, such as whether motor skills learning is most effective with varying training conditions and the distinction between purely motoric aspects and knowledge of procedures. Possible ways forward for surgical training are outlined. It is suggested that the theoretical perspectives and techniques available in the area of motor behaviour should be incorporated into future experimental studies of surgery in virtual environments.