Tense communication and disruptive behaviors during surgery have often been attributed to surgeons’ personality or hierarchies, while situational triggers for tense communication were neglected. Goals of this study were to assess situational triggers of tense communication in the operating room and to assess its impact on collaboration quality within the surgical team.

The prospective observational study was performed in two university hospitals in Europe. Trained external observers assessed communication in 137 elective abdominal operations led by 30 different main surgeons. Objective observations were related to perceived collaboration quality by all members of the surgical team. A total of 340 tense communication episodes were observed (= 0.57 per hour); mean tensions in surgeries with tensions was 1.21 per hour. Individual surgeons accounted for 24% of the variation in tensions, while situational aspects accounted for 76% of variation. A total of 72% of tensions were triggered by coordination problems; 21.2% by task-related problems and 9.1% by other issues. More tensions were related to lower perceived teamwork quality for all team members except main surgeons. Coordination-triggered tensions significantly lowered teamwork quality for second surgeons, scrub technicians and circulators.

Although individual surgeons differ in their tense communication, situational aspects during the operation had a much more important influence on the occurrence of tensions, mostly triggered by coordination problems. Because tensions negatively impact team collaboration, surgical teams may profit from improving collaboration, for instance through training, or through reflexivity.

Citation: Keller S, Tschan F, Semmer NK, Timm-Holzer E, Zimmermann J, Candinas D, et al. (2019) “Disruptive behavior” in the operating room: A prospective observational study of triggers and effects of tense communication episodes in surgical teams. PLoS ONE 14(12): e0226437. https://doi.org/10.1371/journal.pone.0226437

Editor: Frank Lau, Louisiana State University, UNITED STATES

Received: August 25, 2019; Accepted: November 26, 2019; Published: December 12, 2019

Copyright: © 2019 Keller et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Data cannot be shared publicly because they contain features that may remove the anonymity of the surgeons who participated in the study. Data are available from the CTU (Clinical Trial Unit) of the Faculty of Medicine of the Bern University and Bern University Hospital (Switzerland). The following e-mail address can be used to contact the CTU for data request: [email protected], for researchers who meet the criteria for access to confidential data.

Funding: This work was supported by a grant from the Swiss National Science Foundation (grant number 156882) to FT, GB, and DC. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

The picture of the cursing surgeon that throws instruments across the operating room (OR) is still present in many minds. Although such extreme events are nowadays rare [1], tense communication has not disappeared from modern ORs [2,3]. Several studies observed between one and four moments of tense communication per surgical procedure [4,5]; Jones and colleagues [1] reported that 2.8% of all communications in the OR were tense.

Expressed tensions can be described as communications emitted in a negative affective tone; they include the expression of dissatisfaction and incivilities of different intensity. Tensions thus range from subtle negative behaviors to overt aggressive communication such as yelling [6] or using condescending language [7]. A characteristic of these behaviors is that they do not necessarily imply an intent to harm: the instigator may even be unaware of potential harm caused [8]. Note that this definition is broader than definitions that limit disruptive behavior as socially and professionally inappropriate conduct [9]. These behaviors can, however, be characterized as signs of impaired “relational coordination” [10].

Uncivil behaviors are by no means restricted to surgeons, or to supervisors in general [11]. Nevertheless, attending surgeons are more likely than other professional groups to express tensions in the OR [12–15]. A common explanation regards personality or personal communication style of the surgeons as the main reason for tense communication [15], corresponding with studies in other medical teams, where physician personality is one of the most frequently mentioned reason for tensions [2,9,16]. Such attributions perceive tensions to be primarily caused by specific individuals, often known as “bad apples” [9,15] or bullies [17]. Other studies suggest that anxiety, depression, aggressiveness or prior victimization of surgeons increase the likelihood of dysfunctional behavior [9]. The prominent role in the OR hierarchy [7] and the surgeon’s status in the organization potentially contribute to the persistence of these behaviors as well [9,18,19].

However, explanations that focus on personal predispositions are likely to underestimate contextual factors that trigger tense episodes. Stress, as well as production and time-pressure may further increase tense communication [9,10,14]. Many tensions are indeed related to overrunning or changing scheduled surgeries and to availability of theatre time, staff or equipment [5,20–22]. During surgical procedures, unexpected intra-operative complications, but also divergent opinions related to safety have been found to trigger tensions [15], as have task-execution and collaboration problems [4,14]. Thus, focusing solely on personality and person aspects when explaining tensions in the OR is not sufficient, and research on tensions in the OR requires observational studies looking beyond personal aspects of team members as reasons for tensions [7] and impaired relational coordination in general [10].

The primary goal of this prospective observational study was to assess tense communication episodes in the OR, with a focus on its triggers, and to distinguish between personal and situational influences on the occurrence of tense communication in the OR.

The second goal of this study focusses on the potentially negative consequences of tense communication for the quality of teamwork in the OR [23].

Tense communication can elicit negative emotions, which may decrease attentional resources [24], shift the attention to the perpetrator [25], and thereby hamper individual performance in the OR [26]. Tensions have negative effects for the OR team [9]. After tense episodes, team members minimize communication [27,28] and prosocial behavior [28]; and even mild display of negative emotions can impair speaking up [29,30]. In addition, tensions may impair team learning [31]. Thus, tensions most likely have a negative impact on the quality of collaboration within surgical teams.

However, the negative effects of tensions depend on the situation that triggered the tension [32], as people take the context into account when interpreting social communication [33]. For example, if a surgeon is faced with a very difficult aspect of the surgical task and complains with high tension about an apparatus that is not functioning, team members may attribute the tension to the surgeon’s stress. They thus may experience such a tension as less threatening, as compared to a situation in which the surgeon reacts angrily because he or she is not satisfied with the cooperation, for example by how instruments are handed. This implies that the negative effect of tensions on collaboration quality may be different if the tense communication is triggered by collaboration problems (where someone could be blamed), as opposed to problems due to the difficulty of the task at hand or to malfunctions that no one present is responsible for [32].

In general, more conflict or tensions in teams are related to lower performance, but also to lower satisfaction of team members [34–36]. In surgical teams, tensions potentially impair collaboration quality, and an association between tensions and worse patient outcomes or medical errors [2,7,9] is assumed [26].

Second goal of this study was to assess effects of observed tensions during surgeries on the perceived collaboration quality within the surgical team.

This prospective observational study was part of a larger study that received ethical approval by the Institutional Review Board of the canton of Bern (KEK-BE #161/2014; leading ethics committee) and the canton of Vaud (CER-VD #2016–00991). No patient data was used in this study. Participating surgical team members were extensively informed at dedicated presentations and received written information about the study. Consent of surgical team members to be observed was assured by an opt-out procedure–participants could at any time refuse the presence of the observer in the operating room. Oral consent was obtained from team members each time before filling in the questionnaire. Participants expressed their consent by accepting to answer to the questionnaire. No written record of participant name was collected.

In two university hospitals, 137 abdominal operations, led by 30 different main surgeons, were observed. Inclusion criteria were availability of the observers and elective surgical procedure. All team members in the OR were aware of the presence of the observers and were informed that communication and distractions were observed.

The surgical procedures were observed by trained observers (industrial psychologists) using a validated observational system for assessing communication and distractions in the OR [37]. Interobserver reliability was assessed by Cohen’s kappa, separately for each observational code; all Cohens kappa were above 0.70. The observers were present in the OR, sitting about 150 cm away from the operating table. Tablet computers were used for coding, and the coding software [38] automatically timestamped observed events. Observation time was between the WHO timeout checklist (immediately prior to incision) and the last stitch.

A tense communication was coded if a verbal message of a member of the surgical team was expressed in an annoyed or angry tone [4,39]. This definition is broad and ranges from the expression of mild annoyance not targeted at other team members to open insults. If a tension was observed, observers used a comment function on the tablet to describe the content of the communication and the circumstances. These open descriptions were content-coded by two independent coders (SK and FT) [40]; disagreements for this coding were resolved in discussion.

Coders noted the person who initiated a tense communication and who, if anyone, was the target towards whom the tense communication was directed. Codes for initiators referred to all team members, that is, main surgeon, second surgeon, resident, scrub technician, anesthetist, circulator, plus non-team members (e.g., external visitors). For the tensions that were aimed at specific team members (targets of tension), the same codes were used with the exception that surgeons (main, second or resident) were not distinguished; furthermore, a code for tensions targeted at the whole team (room), and a code for undirected tensions were added.

Based on the observers’ summaries of what was communicated during a tension episode, triggers of tensions were coded. Note that more than one trigger was possible for some tensions. Code development was based on previous research [39,41,42] on the one hand, and inductively derived from the material on the other hand. Our examples refer mainly to surgeons because the surgeons initiated most of the tense communication (see results section). Task-related tension trigger was coded if tensions were triggered by disagreements about the procedure, by problems related to the procedure (e.g. the surgeon expresses anger about the complicated anatomy of the patient), by the work organization (e.g. a surgeon complains about shift changes in the anesthesia team), by delays that cannot be attributed to a specific person (e.g. a device is not working correctly) and by too much noise or distractions (e.g. a surgeon complains about a high noise level in the OR. Coordination-related tension-trigger was coded if the trigger was related to collaboration difficulties (e.g. a surgeon demands that the scrub technician hands the instruments faster; a surgeon expresses anger about being handed the wrong type of thread), triggered by an inappropriate or inept act of another person (e.g. the surgeon shouts at a resident who obstructs her view; a nurse tells a resident in an angry tone that he really needs to find that clamp), or triggered by a perceived lack of competence of a team member (e.g. the surgeon angrily tells the scrub technician that she should know by now that he always uses a specific type of thread). Other tension triggers was coded for tensions without an obvious relationship to the current case (e.g. the surgeon blames a resident for discharging another patient without full information), and interpersonal tensions that do not have a clear trigger but indicate a tense relationship between two individuals (e.g. a scrub technician and a surgeon have difficulties collaborating and show signals of interpersonal disaffection).

Before leaving the OR, each member of the surgical team (all surgeons, anesthetists, scrub technicians, and circulators) were asked by the observers to individually fill out a short questionnaire related to the procedure. Up to three questions assessed the perceived quality of teamwork within, or with, the surgical team. Main and second surgeons answered tree questions, assessing collaboration quality with the other surgeons, the scrub technician and anesthetists separately. An example is “How was the collaboration among the surgeons”. For the surgeons, the three questions were combined into a scale. Cronbach’s alpha for this scale was 0.69 for main surgeons, and 0.82 for second surgeons, indicating acceptable to high internal consistency. For scrub technicians, the question was “How was the collaboration within the surgical team”; and for anesthetists and circulators it read “How was the collaboration with the surgical team”. Answers were on a seven-point Likert scale from 1 (very bad) to 7 (very good). For the perception of teamwork quality by the team as a whole, values of the individual members were averaged. As different team members may have very different perspectives, this index cannot be expected to be unidimensional, and indeed, Cronbach’s alpha was 0.50. Even though agreement about teamwork quality among surgical team members is not very high, an average value nevertheless can be regarded as an overall indicator.

Primary outcome of this study was triggers of tense communication episodes during surgical procedures. Triggers included task-related, coordination-related, and other triggers. Secondary outcome was teamwork quality as a function of tense communication, as well as the relationship of tensions triggered by task versus coordination issues with teamwork quality.

We used descriptive statistics (counts, per cent, mean, standard deviations) for descriptive results. We used t-test to assess differences between hospitals; χ2 test to assess differences between surgeons with regard to absence or presence of tense communications, and analysis of variance to assess differences in the mean number of tensions across different surgeons.

To assess effects of tensions on perceived collaboration quality, we used random intercept multilevel multiple regression analyses. Multilevel regression is appropriate because surgeons performed several surgeries. The main surgeons therefore were treated as predictor on the second (higher) level; all other variables, notably the situational triggers, were predictors on level 1 (surgery). Results were adjusted for hospital and for duration of the surgery. Multilevel regression was also used to calculate the variance partitioning value as the percentage of variance associated with individual surgeons or with the context (surgery) when estimating the mean number of tensions. IBM SPSS Statistics for Windows, Version 25.0, released 2017 in Armonk, NY by IBM, was used for data analyses; P