Sir,

We thank Dr Swampillai et al for their correspondence1 regarding our article.2 In our study, the main inclusion criterion was ophthalmic trainees with minimal surgical experience2 (as defined in the paper). No other essential or desirable criteria from the ophthalmology training selection process were tested as trainees recruited had already passed through all this process. As Swampillai et al rightly pointed out, the importance of stereopsis in achieving satisfactory skill in ophthalmic surgery still remains debated.3 There are various gradations of stereopsis impairment, and until a clear relationship between these and surgical skills performance is defined their influence on data can only be speculated. There is also a range of other potential extraneous factors that could potentially influence surgical performance, some described, for example, sleep deprivation,4 and likely many more that have not been examined formally. It was for this combination of reasons that during the study, outset inclusion and exclusion criteria were defined as they were.

Defining the surgical learning curves will become central as the use of simulators broadens. We thank Swampillai et al for highlighting their observation and pretraining description. Importantly, our study showed that there were statistically significant differences in the results between the different tasks, thus the learning curves are likely to vary significantly depending on the task selection. Without more detailed quantitative analysis of how the simulator scores vary during this pre-training process, along with its effects thereafter, we feel the validity of this methodology cannot be assumed and cannot be uniformly applied to all tasks.

It is likely that as training systems progress that the definition of competency of performance for an individual will not just include being able to perform a task to a high level, but also demonstrate that this can be done consistently (with low variability). The authors feel that this is a very important question and significantly more work will be required in this field to best define methods for increasing consistency of performance.