Context It is widely believed that the emotional climate of surgical team's work may affect patient outcome.
Objective To analyse the relationship between the emotional climate of work and indices of threat to patient outcome.
Design Interventional study.
Setting Operating rooms in a high-volume thoracic surgery centre from September 2007 to June 2008.
Participants Thoracic surgery operating room teams.
Intervention Two 90 min team-skills training sessions focused on findings from a standardised safety-culture survey administered to all participants and highlighting positive and problematic aspects of team skills, communication and leadership.
Main Outcome Measures Relationship of functional or less functional emotional climates of work to indices of threat to patient outcome.
Results A less functional emotional climate corresponded to more threat to outcome in the sterile surgical environment in the pre-intervention period (p<0.05), but not in the post-intervention or sustaining period of this study. This relationship did not exist in the anaesthesia or circulating environments of the operating room.
Conclusions The emotional climate of work in the sterile surgical environment appeared to be related to threat to patient outcome prior to, but not after, a team-training intervention. Further study of the relationship between the emotional climate of work and threat to patient outcome using reproducible methods is required.
Statistics from Altmetric.com
The concept of ‘climate’ in healthcare is commonly used to focus on the ‘safety climate’ of organisations. ‘Safety climate’ may be viewed as a manifestation of underlying ‘safety culture’.1 2 We have extended the concept of climate into the ‘emotional climate’ of work in order to better understand how the expression of collective emotions by groups of actors may have consequences for individual and collective behaviours and what we have called, ‘threat to patient outcome’.3
It has been speculated that the ‘emotional climate’ of work affects the adequacy of teamwork, communication and patient outcome.4–6 A tense ‘emotional climate’ has been noted by Lingard and colleagues to be both infectious and to potentially lead actors to withdraw.6 In a simulated aviation environment, the captain's communication style and experienced crew stress were related to crew performance.4 These findings suggest that ‘emotional climate’ may contribute to team performance and patient outcomes. To the authors' knowledge, there are no data using reproducible techniques that analyse these associations.
Elsewhere we have reported our findings from an observational study to evaluate the impact of a defined team skills and communication improvement intervention on measurements of ‘emotional climate’ and indices of team performance.3 This study demonstrated a statistically non-significant trend towards an improvement in emotional climate in the sterile surgical environment of a pod of thoracic surgery operating rooms in a high-volume centre following a teamwork training intervention.
Here, we report the relationship of a functional (engaged or appropriately tense) and less functional (disengaged or inappropriately tense) ‘emotional climate’ and threat to patient outcome in the sterile surgical, anaesthesia and circulating environments of the operating rooms in the previously described study.
This study was conducted in four thoracic surgery operating rooms at the Brigham and Women's Hospital, Harvard Medical School, in Boston, Massachusetts, from September 5, 2007 to June 30, 2008.
Enrolment and consent
Approval for the study was obtained from the Partners Human Research Committee. Information sessions were held to inform thoracic anaesthesiologists, nurses, technicians and surgeons of the study objectives, risks and benefits. The study was authorised to begin after 70% of staff who worked in the thoracic operating rooms voluntarily opted in by completing a de-identified safety-culture survey7 during organised information sessions. From that point forward, anyone not wishing to participate had the ability to opt out by informing the principal investigator. No subject opted out.
Study procedures and randomisation
This study was developed in consultation with a steering group consisting of two thoracic surgeons, two anaesthesiologists, two nurses and three safety specialists. Observers included two anaesthesiologists, two nurses and three safety specialists. Two surgeons participated in refining the observation template based on pilot observations. Although these surgeons were approved investigators and observers, they did not participate in the formal study observations due to scheduling conflicts.
The group of subjects observed consisted of the core members of the thoracic surgery group's surgeons, anaesthesiologists, nursing staff and technicians. Although the team configurations did change, the core group did not—all were dedicated to the thoracic surgery pod.
Data were recorded on a standardised observation form published in our previous manuscript.3 Definitions of the specific behaviours were printed on the back of each form as a reference and are listed in tables 1, 2 and box 1. Three environments of the operating room were observed simultaneously: the anaesthesia environment, the sterile surgical field including surgeons and scrub nurses, and the unsterile circulating environment adjacent to the sterile field in which circulating nurses, but not anaesthesiologists, worked.
Threat to outcome scored as appropriate, inadequately managed or not applicable (items 7–10)
Physical environment supports staff
Equipment and materials support procedures
Staffing level supports safe care
Shared mental model is maintained
Situational awareness is maintained
Interruptions/distractions are effectively managed
If applicable, clinical support available when needed
If applicable, communication with other departments is coordinated
If applicable, handoffs are comprehensive
If applicable, in a crisis, an event manager is established
Prior to initiation of the project, 36 exploratory observations were performed to desensitise subjects to being observed and to develop and test operability of behavioural definitions to be used in the study. Definitions were refined in consultation with observers and the project steering group. Analysed communications and team skills were refined from an observational template that the hospital patient safety group had used for previous quality improvement initiatives. Definitions of threat to outcome were created based on information from a group at Kaiser Permanente involved in a study that was ongoing, analysing threat to outcome. This study was reported while the present work was being conducted.8 Once the definitions were finalised, 30 paired pilot observations of thoracic surgical procedures were done to ascertain whether inter-rater reliability and agreement were acceptable. One new observer joined the team mid-way through the study and he was required to do paired observations until his inter-rater reliability matched that of the core group.
A randomisation list of operating rooms was used to decide where to observe. Following randomization, a single observer entered the room and stood adjacent to the doorway in a position where he or she could observe the sterile operating field, anaesthesia environment and circulating environment. Patient vital signs present on entry into the operating room were recorded from the overhead display along with other demographic data. Following this, observers assessed the emotional climate as disengaged, engaged, appropriately tense or inappropriately tense in the three individual environments (table 1).
Observers then rated six behavioural team-skills elements across the three environments as adequate, mediocre or expected but not observed (table 2). Each instance of the behaviour was recorded as a tick mark. Observation periods lasted for 10 min. At the end of an observation period, observers identified any threat to outcome across the three environments seen during the previous 10 min. The items listed (box 1) could be checked as either ‘Appropriate’, that is, the particular item described was appropriately managed or ‘Inadequate’, that is, the item listed was inadequately managed (ie, a threat was present). For items 7–10 in box 1, a third alternative (to either Appropriate or Inadequate) was ‘Not applicable’.
Data sheets were changed at the end of every 10 min observation period or if the emotional climate changed and remained changed for 1 min. A new set of patient vital signs was obtained within the first minute of using a new data sheet. All data were entered into the project database in a de-identified manner.
Data were collected in three periods; pre-intervention (11 consecutive weeks following the beginning of the study), post-intervention (12 consecutive weeks following a training intervention) and a sustaining period. The intervention consisted of two 90 min multidisciplinary team-skills training sessions for surgeons, anaesthesiologists, technicians and nurses during which a team-training expert (AF) created an interactive forum to educate team members on the importance of communication, and to role play interactive and communication strategies designed to optimise safety. The session included a discussion of findings from a standardised safety-culture survey7 administered to all participants and highlighting positive and problematic aspects of team skills, communication and leadership.
For the purposes of data analysis, threat to outcome were calculated as the presence or absence of a specific threat marked on each data sheet. The emotional environment was categorised as functional (engaged or appropriately tense) and less functional (disengaged or inappropriately tense) in each of the three environments.
Patient characteristics and study results were calculated using proportions. Kappa coefficients were estimated to determine the degree of agreement between pairs of ratings on the same observation. Fisher's exact tests were used to determine differences in proportions between cumulative threat to outcome in a functional versus less functional emotional climate during each period. All analyses were performed using SAS software, version 9.2 (SAS Institute, Cary, North Carolina). Statistical significance was set at p<0.05.
There were 1712 eligible surgical cases during the study period. A total of 305 data sheets (91 pre-intervention, 165 post-intervention, 49 sustaining) were collected, representing 105 surgical cases and approximately 50 h of observation.
During the 30 initial paired observations, the Kappa coefficient and 95% CIs for climate in the sterile surgical, anaesthesia and circulating environments were respectively, 0.99 (0.85–1.00), 0.99 (0.86–1.00), 0.51 (0.15–0.86). During the initial paired observations, agreement on threat to outcome exceeded 90% with the exception of one element (situational awareness being maintained) for which there was 75% agreement.
A less functional emotional climate corresponded to more threat to outcome in the sterile surgical environment in the pre-intervention period (p<0.05), but not in the post-intervention, or sustaining periods of this study (tables 3–5).
A less functional emotional climate was not related to threat to outcome in the anaesthesia or circulating environment during the pre-intervention, post-intervention and sustaining periods of this study (tables 3–5).
We labelled a functional emotional climate as one that was ‘engaged’ or ‘appropriately tense’. We labelled a less functional emotional climate as one that was ‘disengaged’ or ‘inappropriately tense’. We found that a functional emotional climate correlated with less threat to outcome and that a less functional emotional climate correlated with more threat to outcome in the sterile surgical environment in the pre-intervention period (tables 3–5).
Our study did not find a relationship between threat to outcome and emotional climate during the post-intervention or sustaining periods (tables 4 and 5). The intervention consisted of two 90 min multidisciplinary team-skills training sessions for all subjects in the thoracic surgery pod. Team-training sessions focused on areas identified by the safety-culture survey as having high likelihood of affecting the outcome measures and being open for improvement based on the safety-culture survey scores (open communication, response to error, staffing concerns, management support for safety, teamwork, handoffs and transitions). Sessions were conducted by one of the author's, a team-training expert (AF), and focused on optimal communication strategies. One explanation for the lack of relationship between emotional climate and threat to outcome is that the study was insufficiently powered to detect this change once the overall threat decreased as a result of the team-training intervention.3
The emotional climate was not related to threat to outcome in the anaesthesia or circulating environments (tables 3–5). A potential explanation is that, although threat to outcome is influenced by any person involved in patient care, the potential for and magnitude of threat is greatest at the sterile surgical field.
We used ‘threat’ to patient outcome as a surrogate of actual outcome. Recent work has suggested a relationship between team behaviour and direct measures of outcome.8 The sensitivity and specificity of surrogate measures of direct outcome such as ‘threat’ requires further investigation.
Our findings contribute to the growing body of evidence that a perturbed emotional climate is associated with poorer team performance, as has been suggested in other disciplines.4–6 Although this relationship is commonly believed to exist, to the authors' knowledge, this is the first study using a reproducible technique to directly investigate the relationship.
The emotional climate of acute-care medical work should be the subject of further study using reproducible techniques. Powering such studies will continue to be a challenge, and a refined instrument and an improved definition of emotional climate will be necessary to nuance the understanding of emotional climate and its relationship to outcomes.
This study has several limitations. Definitions of emotional climate and threat to patient outcome were developed during this study and the adequacy of these definitions has not been tested.
There were no surgeon observers for the formal portion of this study although surgeons did help refine the observation template based on pilot observations. It is possible that surgeon observers may have rated behaviours and the emotional climate differently to non-surgeon observers.
Although the intra-class correlations for the 30 initial paired observations of emotional climate were perfect for the surgical and anaesthesia environments (but not the circulating environment), the extent to which the emotional climate is directly observable is not clear and warrants further study. Anecdotally, many participants and observers felt that it was obvious when an emotional climate was perturbed—an idea central to formulating this study, but they found it difficult to describe the perturbation succinctly in words.
As reported elsewhere, the indices of inter-rater reliability were not identical and this may have affected our results.3 Inter-rater reliability was assessed prior to observations, but it was never re-checked and it is possible that during the course of the study inter-observer variation became greater.
Although agreement on threat to outcome during pilot observations was excellent, threat to outcome was recorded based on observer's judgments of a threat being present. The extent to which threats are directly observable is not clear and warrants further study.
We cannot exclude that individual behaviour improved when observers were in the operating room. The study took place from the beginning to the end of an academic year and we cannot exclude that the observed improvements were impacted by an increase in team skills by trainees. For these reasons, we caution readers not to make direct causal relationships between emotional climate and outcome. These and other confounding factors may have influenced the observed correlations.
Some of the behaviours taught during the team-training sessions were observed at a lower frequency after training, albeit often with greater adequacy when they did occur.3 This finding deserves further research to explain which behaviours were responsible for the effects of the team-training intervention reported previously.
A tertiary aim of this project was to consider relationships between the emotional climate of work and threat to outcome. For this reason, these results should be considered exploratory and warrant further investigation. In addition, further work is necessary to confirm that our method of measuring behaviours and emotional climate is indeed reproducible.
The emotional climate of work and its relationship to direct and surrogate measures of patient outcome should be the subject of further study using reproducible techniques and interventions. Methods for improving the emotional climate of operating room team-work may improve patient outcome.
Funding Partners HealthCare Inc., Boston, Massachusetts.
Competing interests During the period of this study, Allan Frankel was Director of Patient Safety at Partners Healthcare. He is currently a Principal at Pascal Metrics however the methods used in the study are unlike any that are currently being used and have no bearing on Pascal activity.
Ethics approval This study was conducted with the approval of the Partners Human Research Committee, Brigham and Women's Hospital/Massachusetts General Hospital.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.