Introduction

Surgery is a bimanual profession; it involves adequate visualisation, manipulation, and mobilisation of structures held within enclosed spaces. Execution of procedural tasks efficiently is the hallmark of a fully qualified surgeon. Research to date has focused on the acquisition of dexterity in the non-dominant hand of general surgeons, a skillset accentuated through the introduction of minimally invasive surgery into standard practice. The bimanual control required for visualisation and manipulation of structures in the abdomen has been the subject of much of the research in this area, with evidence on the superiority of left handed-people noted in dexterous tasks, often theorised to arise from an adaptation to a right-hand dominant environment [4]. Efforts to mitigate the potential impact of hand dominance on surgical performance have resulted in the introduction of robotic assistance to help surgeons overcome limitations of the non-dominant hand [7].

Orthopaedic surgery is lateralised by the nature of the operating field, including the differences in surgical approaches based on the operative side during limb and spinal surgeries, and surgeon handedness. Handedness has been demonstrated to impact patient outcomes in a variety of subspecialties; with hand dominance impacting the quality of pedicle screw placement [19], medialisation of hip implantations [8, 12], and patient outcomes following total knee arthroplasties [1).

Fig. 1
figure 1

Non-dominant camera alignment between groups. Experts demonstrated consistently better camera alignment with a narrower standard deviation than both novice and intermediate groups

Full size image

Between-group performance

One-way ANOVA evaluating between-subjects effects indicated gras** efficiency in the dominant hand demonstrated a significant difference (N 44% \(\pm 17\) vs. I 55% \(\pm\) 23 vs. E 55% \(\pm\) 14, p = 0.013). Between-group analysis noted a significant difference in length of the camera path during the ND task attempt (N 122.3 cm \(\pm\) 162.4 cm vs. I 79.2 cm \(\pm\) 32.4 cm vs. E 83.3 cm \(\pm\) 35.9 cm, p = 0.006) between groups.

Participant characteristics

Hand dominance

Significant difference between left (LHD) and right (RHD) hand dominant participants was noted in two parameters. Length of grasper path was noted to be significantly longer in the LHD dominant group (N 409.1 cm \(\pm\) 232.6 cm vs. I 325.1 cm \(\pm\) 81.6 cm, p = 0.032). Similarly, grasper path length travelled with grasper open was also found to be significantly longer in this participant cohort (N 248.2 cm \(\pm\) 167.8 cm vs. I 197.7 cm \(\pm\) 93.3 cm, p = 0.034).

Gender

There were no females in the experienced group, reflective of current trends in the proportion of gender representation across surgical specialties at the time of writing. Performances from both genders in both intermediate and novice groups were compared for analysis. Females were found to take significantly longer to complete the exercise with the grasper in their non-dominant hand (Female 263.58 s \(\pm\) 192.34 vs. Male 136.47 s \(\pm\) 46.01, p = 0.001) (Fig. 2). Sub-analysis demonstrated this difference as significant only in the novice group, with no difference found between genders in the intermediate group.

Fig. 2
figure 2

Time taken to complete between genders. Males took significantly less time than females across both novice and intermediate groups

Full size image

Multivariate analysis

Between-subject effects using multivariate analysis were carried out to assess the impact of each component on measured outcomes. The ND camera path travelled across gender was significantly different in the female group when controlling for experience (p = 0.017).

Hand dominance when controlling for gender was significant for ND gras** efficiency (p = 0.028). Level of experience when accounting for hand dominance was found to be significant, with the experienced group demonstrating higher levels of proficiency across ND gras** efficiency scores than the intermediate group (p = 0.04) (Table 3.).

Table 3 Group results
Full size table

Discussion

The highlight of this study is the finding which demonstrates grasper efficiency in the ND hand improves with increasing arthroscopic experience. This indicates that acquisition of bimanual control skills likely occurs with increased exposure to arthroscopy during conventional orthopaedic training. Thus any future implementation of simulation-based training which has been demonstrated to improve surgical skills in Orthopaedic trainees should consider the integration of bi-manual specific tasks to decrease laterality which was found in the inexperienced cohort.

Previous studies validating arthroscopic simulators available commercially have attempted to identify epitomical surgical skills reflective of expert performance. Limited availability of valid surgical simulation-based metrics to analyse performance has led to conclusions derived from endpoints with limited generalisability including; task duration [1]; surrogate skill markers to evaluate proficiency [10], and exclusion of datapoints which were incongruent with expected outcomes occasionally demonstrated [15]. This study evaluated each metric outputted from the “remove the stars” module, with exercise duration not demonstrating any significant difference between groups in either dominant or non-dominant attempts, highlighting the limitation of this endpoint as a valid metric to differentiate between levels of experience. Gras** efficiency and camera path length, however, were demonstrated to reflect levels of experience between groups; demonstrating both the ability of the simulator to readily identify between levels of experience, and the potential use of these metrics to identify acquisition of skill through Orthopaedic training.

The impact of baseline characteristics on simulated performance has been to date unreported in arthroscopic simulation-based studies. Both gender and handedness demonstrated a significant impact on simulator derived outcomes in this study. Handedness has been addressed previously in surgical training, with research to date focused primarily on the left-handed trainee [2, 13]. Up to 15% orthopaedic surgeons identify as left hand dominant [16], however, variability in the operating field in orthopaedic surgery dictated by patient pathology highlights the onus on orthopaedic surgeons and trainees to reduce laterality regardless of hand dominance to promote positive patient outcomes [7].

Gender was similarly found to impact performance. A disproportionate level of female participants had a significantly longer time taken to completion recorded during the attempt completed in the non-dominant attempt reaching a level of significance in the novice group. This finding was no longer significant in the intermediate group. While it should be noted that the significance of these findings should be noted in the context of the disparity in gender distribution across the groups, this finding echoes previous research related to gender in laparoscopic practice which found right-handed makes performed superiorly in the simulated module compared to left-handed males and all females included in the study [3], and outlines potential considerations in the implementation of future surgical curricula. The novice group consisted primarily of medical students with a variable interest in surgery as a future career. This may have impacted performance, as interest in surgery has been found to impact simulated surgical performances with a degree of heterogeneity. with some studies indicating an intention to enter surgical careers as an advantage in simulated performance [6, 11].

It is worth noting that despite no significant difference between D and ND attempts across all parameters except for gras** efficiency between the intermediate and expert groups, the expert group reported greater levels of comfort using their dominant hand indicating a degree of laterality remains beyond what simulator software derived metrics can identify. Authors feel this highlights the potential for simulation to promote continued professional development in line with technological advances and prevent surgical skill decay in consultant orthopaedic surgeons.

This study has several limitations; the disproportionate representation of females limits the generalisability of the results; however, the study demographics are reflective of current surgical training and is representative of the study population. The novice group contained the most balanced representation in gender, with females consisting of 40% of the group, similarly reflecting current medical school trends, finding a significant difference in simulator-derived metrics at this level. The ceiling effect is a recognised limitation in simulation-based studies [14]. Laterality in the expert group may not be detectable in current simulator software yet have a clinical impact on patient outcomes, and as such should be considered a potential confounder.

Conclusion

Arthroscopic simulation-based training is a valuable learning tool for orthopaedic training. This study demonstrated that experienced orthopaedic surgeons have a greater degree of ambidexterity than intermediate or novice groups, hypothesised by authors to be conferred through conventional orthopaedic training. Consequently, incorporation of simulation-based orthopaedic training in future training curricula require integrated bimanual-dedicated modules to mirror real-world practice, as well as facilitated learning opportunities for the development of bimanual skills in everyday practice.