Are the "Visual Haptics" Conferred via Robotics Non-inferior to Actual Haptics?

Objectives: The precision, dexterity and visualization provided by the most widely used robotic surgery system (da Vinci Xi System, Intuitive Surgical, Sunnyvale CA) have been realized despite its lack of haptic feedback to the surgeon. While critics have bemoaned the absence of haptic feedback as the Achilles heel of the technology, experienced robotic surgeons often report development of “visual haptics” which allow them to discern the properties of various tissue types accurately via the high definition 3D image available at the surgeon console. To date, no studies have compared surgeons’ ability to discern the nature of objects using their hands versus robotic instruments. The objective of this study was to determine whether surgeons’ ability to discern balloon contents using robotic technology was non-inferior to discernment via surgeons’ hands.

Methods: This was a non-randomized parallel group study based on a model that included four identical balloons each with different contents – air, water, petroleum jelly, or a firm substance (Figure 1). One group of surgeons (Group A) discerned the balloon contents by feeling with their hands, and the other group (Group B) did so via the robotic technology. Study surgeon activities were carried out in a one-on-one basis for both groups so as to eliminate the chance of any surgeons learning balloon contents through observation of other surgeons. Groups A and B were compared in terms of age, and years since residency graduation. We used the Farrington-Manning test of non-inferiority. We expected 100% correct balloon discernment by hands-on surgeons, and we set 80% correct discernment by robotic surgeons (i.e. a risk difference of 0.2) as our definition of non-inferiority. A correct response was a binary outcome for each of the four balloons of varying density.

Results: A total of 57 "hands-on" surgeons and 19 robotic surgeons participated in the study. All surgeons were board certified in their respective fields. The mean number of years since residency graduation and mean age of Group A and Group B were similar ( p-values 0.17 and 0.16 respectively). As expected, 100% of Group A (the “hands-on” surgeons) discerned all 4 balloons correctly. Of the 19 robotic surgeons, one made 2 discernment errors. Therefore the actual risk difference between groups was 2.6% which easily met our definition of non-inferiority (p= < .0001).

Conclusions: Object discernment via “visual haptics” provided by the da Vinci Xi System was non-inferior to actual haptic discernment by “hands-on” surgeons.