Abstract
The rise of robotic surgery has been accompanied by numerous educational challenges as surgeons and trainees learn skills unique to the robotic platform. Remote instruction is a solution to provide surgeons ongoing education when in-person teaching is not feasible. However, surgical instruction faces challenges from unclear communication. We aimed to describe, examine, and compare ambiguities in remote and in-person robotic instruction. We designed a simulation scenario in which a standardized learner performed tasks in robotic surgery while making pre-scripted errors. Instructors provided remote or in-person instruction to the standardized learner. We applied tools from discourse analysis to transcribe sessions, identify instructional instances, classify ambiguities, and select passages for further review. We used tests of proportions to compare ambiguities between the settings. We conducted four simulation sessions, including two remote and two in-person sessions, and identified 206 instructional instances. Within these, we found 964 occurrences of three common semantic ambiguities, or ambiguities arising from words alone. Instructors used visual tools - thus employing multimodality - to clarify semantic ambiguities in 32% of instructional instances. We identified a similar degree of referential ambiguity, or ambiguity for which context from multimodality did not provide clarifying information, during remote (60%) and in-person (48%) instructional instances (p = 0.08). We described, examined, and compared ambiguities in remote and in-person instruction for simulated robotic surgery. Based on the high prevalence of ambiguity in both settings, we recommend that robotic instructors decrease referential ambiguity. To do so, instructors can reduce semantic ambiguity, harness multimodality, or both.