Miscommunication in the Robotic OR: A Weighted Correlational Analysis

INTRODUCTION: Miscommunication in the operating room (OR) occurs up to 8 times per hour and is a leading cause of patient harm. Environmental distractors, which interrupt communication, become especially disruptive in robotic surgeries due to dispersed team members, complex tasks, and loud equipment. Subsequently, robotic ORs commonly suffer from speech communication interference (SCI), defined as “group discourse which is obstructed according to the goals of the communication, or the physical and situational context of the exchange,” further increasing miscommunication and risk of patient harm. OBJECTIVE: To analyze the association of miscommunication (SCI) events in the robotic OR with the staff involved, evidence of SCI, the context, and associated events. METHODS: This is an analysis of 687 SCI events from 78 robotic surgeries, primarily in gynecology, urogynecology, urology, and general surgery. We used our SCI instrument, which has inter-rater reliability 0.87–0.98, to note roles of sender, receiver, and messenger. Senders delivered the original message, receivers were the intended recipient, and messengers relayed the message when needed. Evidence of SCI included intended receivers not responding, senders/messengers repeating a message, and receivers asking for clarification. Contextual variables included multi-tasking (such as charting, counting sponges, inserting instruments), loud noises, and overlapping conversations. Associated events included the surgeon leaning out of the robotic console to communicate as well as communication requiring a messenger. We performed a correlational analysis, weighted by the number of SCI events per case, of aggregate group-level data to assess the relationships between contextual, event, and staff variables. RESULTS: Correlation coefficients for contextual and event-based weighted correlates show the strongest relationships between overlapping conversations and no-responses (r=0.63) during an SCI; and the surgeon leaning out of the console and clarifications (r=0.77), messenger (r=0.87), and multitasking events (r=0.77) (Table 1). Contextual and staff-based weighted correlates showed strong relationships between the frequency of clarifications and multi-tasking and the involvement of all staff (Table 2). The scrub tech and the resident were correlated more with no response than the attending surgeon. Messenger involvement was highly correlated with all staff except the medical student. Finally, event- and staff-based weighted correlates showed the strongest relationships between the frequency of loud machines and the medical student (r=0.67); and the frequency of the leaning out of the console and the resident (r=0.75) attending surgeon (r=0.79), and scrub tech (r=0.72) (Table 2). CONCLUSIONS: Our analysis suggests that overlapping conversations in the OR interfere with hearing instructions from the console and lead to “no-response.” Furthermore, when bedside assistants are occupied with complex tasks, the surgeon often leans out of the console to clarify. When their message is not received, the resident, CRNA, and scrub tech often deliver the message. Subsequently, we found that loud machines are associated with higher rates of no-response. Clarification is associated with the medical student, likely because they are isolated and lack tacit knowledge. Instructions from the attending have less of an association with SCI than from other surgical team staff. All these miscommunications increase case length as well as the risk of misunderstanding and patient harm.