Summary of main results
Laparoscopic surgery is different from open surgery because of: increased need for hand-eye-co-ordination to perform tasks looking at a screen to compensate for not being able to operate under direct vision; increased need for manual dexterity to compensate for the use of long instruments, which can amplify any error in movement; fulcrum effect of the body wall, ie, when the surgeon moves his hand to the patient's right, the operating end of the instrument moves to the patient's left on the monitor (Gallagher 1999); the need for handling tissues carefully (to compensate for the lack of sensation of touch using hands); and the lack of three-dimensional images. Virtual reality training is one of the many methods of laparoscopic surgical training and is currently aimed at improving psychomotor skills (Gallagher 1999).
An increasing number of procedures are being performed laparoscopically. With the decreasing time to train surgeons because of European Working Time Directive (Chikwe 2004) and modernising medical careers (MMC) initiative by the Department of Health (Payne 2005), training structured to improve surgical skills in the least time with maximum efficiency is necessary. This is applicable to surgical trainees with no prior experience in laparoscopic surgery and in those who have started their laparoscopic career but have not achieved proficiency. Because of the shortened working hours, the trainees may be exposed to fewer surgical procedures and hence may lack experience. Thus, it is necessary to develop generic skills, such as suturing or cutting and also procedure-specific skills, such as cannulating the common bile duct. In the previous version of this review, we demonstrated that virtual reality training improved generic skills such as suturing or cutting (Gurusamy 2009). In this update, we have focused on the effect of virtual reality training on operative performance.
Eight trials with 109 surgical trainees were included in this review. The training regimens included training in basic tasks such as cutting, suturing, transfer of objects, or diathermy in five trials (Hamilton 2002; Seymour 2002; Grantcharov 2004; McClusky 2004; Hogle 2009) whereas they included dissection in anatomical models in addition to basic tasks in two trials (Ahlberg 2007; Larsen 2009). The details of training were not available in one trial (Sendag 2009).
Assessing the evidence, it appears that virtual reality training decreases the operating time compared with no supplementary training. However, the difference is approximately 10 minutes per procedure. Although a formal subgroup analysis was not performed, the decrease in operating time appears to be more pronounced in the trial in which the trainees were trained on anatomical models (in addition to basic tasks) (Larsen 2009) compared to the trials in which the surgical trainees were trained only in basic tasks (Analysis 1.1). However, it should be noted that only three trials reported this outcome (Grantcharov 2004; McClusky 2004; Larsen 2009), and we cannot be certain that the difference observed in the magnitude of the effect was due to the difference in the training regimen. Irrespective of the reason for the difference in the magnitude of the effect, the difference in operating time is unlikely to benefit patients in a major way. Whether this difference will decrease the costs by increasing the number of procedures performed in the theatre list depends upon the type of procedures and the duration of the theatre list. Virtual reality training improves the operative performance compared with no supplementary training. As in the case of operating time, the magnitude of difference is greater in the trial in which the trainees were trained on anatomical models (in addition to basic tasks) compared to the trials in which the surgical trainees were trained only in basic tasks (Analysis 1.3). Again, we cannot be certain that the difference observed in the magnitude of the effect was due to the difference in the training regimen as this outcome was reported by two trials only (Hogle 2009; Larsen 2009). There is considerable uncertainty as to the magnitude of the improvement and what this improvement means to the patient and to the healthcare funder. One would generally equate better operative performance with better patient outcomes. However, there is currently no evidence to demonstrate the correlation between better operative performance and better patient outcomes. The likely reason for this is the sensitivity of the issue. Another reason may be that the performance can change (either improve because of more meticulous surgery or decrease because of the stress of the assessment) when a formal assessment is made, which makes the issue quite a difficult one to prove. However, if we agree to the common logical notion that patient outcomes are likely to be better if the operative performance improves, virtual reality training may improve patient outcomes by improving operative performance. While the longevity of this difference in operative performance between virtual reality-trained surgeons and those with no supplementary training is an unknown quantity, as the surgeons with no supplementary training may catch up with their virtual reality-trained counterparts as they gain more surgical experience, the difference during the learning curve in surgical training may benefit patients.
The version of the software used was reported in only one trial (Ahlberg 2007). It was therefore not possible to determine whether the magnitude of effect was greater with later versions of the software compared to the earlier versions. Because of the limited number of trials, it was not possible to perform a subgroup analysis based on the levels of experience of the surgical trainees, so it was not possible to determine whether the magnitude of effect was greater in surgical trainees with higher number of procedures performed under supervision.
Virtual reality training also appears to decrease the operating time and improve the operative performance when compared with box-trainer training. Again the magnitude and the impact of these differences between virtual reality training and box-trainer training is not known. In the United Kingdom, training is largely based on box-trainers (rather than virtual reality trainers) in addition to the standard laparoscopic training model of apprenticeships. A survey of satisfaction of the trainees in the virtual reality group and box-trainer group (Madan 2005) found that the majority of the trainees preferred box-trainers to virtual reality trainer and a significant number felt that the virtual reality training model was not realistic. This study did not have a haptic feedback interface. Haptic feedback is a tactile feedback technology which mimics the sense of touch by applying forces or vibrations to the user. None of the trials included in this review employed a haptic feedback facility. In the next few years, haptic feedback is likely to form an integral part of the virtual reality simulator and it is likely that the trainee satisfaction will increase with the better simulation. However, the degree of fidelity or realism does not alter the effectiveness in training (Grober 2004). This might explain the reason for the effectiveness of virtual reality training in spite of being a low-fidelity model. However, improving the fidelity may increase trainee satisfaction and the enthusiasm to learn on virtual reality models.
Some potential advantages of virtual reality over box-trainer trainer include:
1. Two-handed tasks need to be followed closely using a second person for training. In virtual reality trainers which follow the instrument tips, there is no need for the second person. In box-trainers with a fixed video camera, a distance has to be chosen so that the task can be viewed closely. The introduction of instruments cannot be followed. This violates the rule of keeping the business end of the instrument under vision always, which is particularly important in those who are beginning their laparoscopic career.
2. One of the other major problems with box-trainer training is the 'trainer' time. An expert is necessary for evaluation and feedback in box trainer training. In virtual reality training, the computer evaluates every movement of the trainee and provides feedback after completion of the task (eg, reports the number of movements, distance moved by each hand, traces the path of movements, etc). These can even be used for monitoring the improvement in skills. Thus, the virtual reality software can act as a 'virtual tutor' and a regular training session every week is feasible. However, this advantage of a virtual reality trainer over a box-trainer has been questioned by some since it is not easy, even for experts, to distinguish reliably that a task was completed without problems (Greco 2010). So if experts cannot come to an agreement about successful completion of a task, it is not possible to program this into a virtual reality trainer in order to determine whether the task was completed without problems.
The potential advantages of box-trainer training over virtual reality training include:
1. Cheaper cost of the model which enables training multiple trainees simultaneously in short training courses.
2. Better realism (use of real tissue and presence of haptic feedback) compared to currently evaluated virtual reality models (Madan 2005).
The recent hybrid simulators with camera trackers to follow the instruments (Botden 2007) combine some of the advantages of virtual reality training and box-trainer training. Further research is needed into whether such hybrid simulators are better than virtual reality trainers.
Recent advances in virtual reality technology has made it possible to import images into virtual reality software from external sources (Jaselskis 2013). It is possible to reconstruct the three-dimensional images if the x, y, z co-ordinates and colour information of each pixel is available (Cyberware 2013). In the near future, it might be possible to import these three-dimensional images into the virtual reality software. Once these images are imported into the virtual reality software, it should be possible to manipulate the images. That would mean that it is possible to train the surgical trainees in three-dimensional reconstructions of actual patients rather than train in their component skills only. Training on numerous such models with anatomic variations can also help with the improvement of decision-making skills and procedure-specific skills.
Overall completeness and applicability of evidence
The results of this review are applicable only to surgical trainees with limited laparoscopic experience and confine to the types of virtual reality training used in the trials. The evidence only shows that the operating time is decreased and the operative performance is increased by virtual reality training when compared with no supplementary training or with box-trainer training. There is currently no evidence that virtual reality training improved patient outcomes.
Quality of the evidence
All the trials were at high risk of bias. While blinding of outcome assessors was performed in three of the four trials that reported this outcome (Hamilton 2002; Hogle 2009; Larsen 2009), the lack of blinding of the participants can result in bias. However, very few of the trials were able to provide data for our meta-analyses, which were based on few participants. Accordingly, we cannot exclude random errors. Overall the quality of evidence is very low as indicated in Summary of findings for the main comparison. Nevertheless, this is the best evidence that is currently available.
Potential biases in the review process
Although study selection and data collection were performed in a non-blinded manner, the potential for bias and errors is largely reduced by the use of two independent data extractors. We were unable to assess publication bias by funnel plot but went through the trial registers to identify the trials. Thus there is unlikely to be publication bias but there appears to be evidence of reporting bias since many of the trials did not report the common outcomes that are likely to be measured during the conduct of the trial. The inclusion of data from such trials may result in a change in conclusions. We imputed the standard deviation when it was not available from the studies. A sensitivity analysis demonstrated that the impact of such imputation was low. The alternative was to exclude the information from the trial which would have made the interpretation of data even more difficult.
Agreements and disagreements with other studies or reviews
The previous version of this review (Gurusamy 2009) found that virtual reality training could supplement the standard laparoscopic surgical training model of apprenticeship, and was at least as effective as box trainer training in supplementing standard laparoscopic training. Most reviews on this topic including the ones mentioned in the Background section arrive at similar conclusions.