Minimally invasive surgical procedures often include complex and cognitively demanding tasks requiring high levels of concentration for optimal performance [1]. Deviations from standard protocols and practice may result in patient harm. Several studies demonstrated that the operating room (OR) environment is frequently faced with several sources of distraction [2,3,4,5,6,7,8,9]. Distractions can break surgeons’ concentration, which may subsequently increase the likelihood of patient harm [10]. Distractions in the OR can be broadly grouped into those arising from auditory sources, such as pager rings and machine alarms, and those from cognitive sources, such as irrelevant conversation and device malfunction [11]. Our group previously performed a prospective cohort study of 132 consecutive patients undergoing elective laparoscopic general surgery using detailed observation of comprehensive audio–video intraoperative data and demonstrated that a median of 138 auditory distractions occurred per case and 64% of the cases had at least one cognitive distraction [7].

In simulated settings, both auditory and cognitive distractions were associated with impaired performance of surgical trainees. Szafranski et al. demonstrated that auditory distraction was associated with increased time to task completion, decreased economy of motion, and increased errors in surgical residents performing a simple laparoscopic task [12]. Further, Feuerbacher et al. demonstrated that cognitive distractions produced increased errors in surgical residents performing a laparoscopic cholecystectomy in a high fidelity simulator [13]. In addition to the deleterious performance effects, distraction can also lead to stress among the surgical team [14]. However, empiric evidence on how surgeons process and perceive distraction in real OR setting is limited. Surgeons may adapt and develop resilience to carry out operations safely despite facing a large number of distraction [11]. Investigations on how surgeons deal with distractions can be performed using human-factors analysis, a discipline from engineering that studies how humans interact with their surroundings. Human-factor research has shown promise in identifying system factors in the OR that can influence performance and safety. One such example is the Surgical Team Assessment Record (STAR) questionnaire used by de Leval et al., which addresses human-factors aspects of operation, including the situational factor of whether the surgeons felt distracted [15]. The present study aims to understand which intraoperative system factors are associated with surgeons’ perceived distraction during operations.

Materials and methods

Subjects, setting, and data source

We conducted a prospective cohort study in 265 consecutive patients undergoing elective laparoscopic gastrointestinal surgery from May 2015 to April 2017 at an academic center in Toronto, Canada, to identify sources of distractions in the OR and determine which factors were predictive of surgeons feeling distracted. The research ethics board at St. Michael’s Hospital approved this study (#16-243).

All elective laparoscopic gastrointestinal surgery procedures performed on adult patients (≥ 18 years old) involving one attending surgeon during the study period were considered for inclusion. The attending surgeon has fellowship training in minimally invasive surgery and worked in the same academic hospital for more than 10 years. Cases with invalid or missing informed consent from patients or the OR team members were excluded.

Baseline patient and procedure characteristics were obtained from admission and pre-operative anesthesia assessment notes through the electronic medical records. Patient-level characteristics included age, sex, Body Mass Index (BMI), a history of previous abdominal surgery, comorbidities, and Charlson Comorbidity Index. Procedure types (e.g., Roux-en-Y gastric bypass, cholecystectomy, and paraesophageal hernia repair) and duration in minutes were collected. Two trained and calibrated raters performed analyses using video–audio data obtained through a comprehensive intraoperative data recorder called the OR Black Box® (Surgical Safety Technologies, Toronto, Canada) in consecutive surgical operations to identify sources of distraction according to a modified version of the framework proposed by Sevdalis et al. [7, 16]. Acceptable inter-rater reliability between the two raters was achieved before the start of this study.

Outcomes

The primary outcome was the attending surgeon’s perceived distraction as answered in a self-reported human-factors questionnaire administered at the completion of each procedure. We modified the Surgical Team Assessment Record (de Leval et al.) questionnaire, which was previously used primarily in cardiovascular surgery, to better reflect pertinent organizational, situational, team, and personal factors in laparoscopic gastrointestinal surgery. One of the questions asked whether the surgeon felt that there were many distractions during the case and if he or she was distracted. The attending surgeon provided a “yes” or “no” answer. The collected data were entered and stored in a secure and encrypted database management system.

Statistical analysis

Descriptive statistics including mean (standard deviation (SD)) or median (interquartile range (IQR)) for continuous variables and frequency analysis (percentages) for categorical variables were performed on patient-level characteristics and procedure-level characteristics. To compare cases with perceived distraction and those without, we performed t test or Mann–Whitney “U” test for continuous variables and Fisher’s exact test for categorical variables. We investigated whether patient-level and intraoperative factors were associated with the attending surgeon feeling distracted during surgical operations. All pairwise associations between each explanatory variable were assessed (Pearson’s r and Spearman’s r for non-parametric data) to ensure none were overly correlated (r ≥ 0.8). Univariable logistic regression was used to investigate the relationship between patient-level and procedure-level explanatory variables and the primary outcome, including age, sex, BMI, a history of previous abdominal surgery, ASA classification, Charlson Comorbidity Index, procedure type, procedure duration, door opening, loud noise, machine alarm, external communication, teaching, management of next case, device malfunction, absent or late member, and irrelevant conversation. Explanatory factor with p ≤ 0.2 on univariable testing was entered into a multivariable logistic model. The accepted statistical practice of considering no more than one explanatory variable for every 5–10 events (response variable) was employed in our model [17]. Goodness of fit was examined for the multivariable logistic model using the Hosmer–Lemeshow test and ‘c’ statistics. Internal validity of the model variable selection was assessed using tenfold cross-validation method. Confidence intervals and p values are all two sided. Analyses were conducted using SAS 9.4 (SAS Institute Inc., Cary, NC) and R software version 3.4.2 (R foundation, Vienna, Austria).

Results

Subjects

Between May 1, 2015 and April 30, 2017, 265 patients met our inclusion criteria and consented to participate in the present study (Table 1). Patients had a mean age of 50 years (SD, 14) and predominantly female sex (76%). More than half of the patients had a history of previous abdominal surgery (61%) and a Charlson Comorbidity Index Score of 1 or above (59%). The most frequently performed procedure was Roux-en-Y gastric bypass (43%), followed by sleeve gastrectomy (13%) and exploratory laparoscopy (13%). Body Mass Index and a history of previous abdominal surgery were missing in two and one patients, respectively.

Table 1 Patient and procedure characteristics
Full size table

Outcomes

Frequency and characteristics of distraction sources

Intraoperative sources of distraction were grouped into cognitive and auditory types (Table 2). More than half (54%) of the cases had at least one teaching activity. In 91 cases (34%), there was at least one absent or malfunctioning device that was required for procedure, such as surgical instruments, laparoscopic consoles, and patient warmers. Any conversation involving the attending surgeon not related to patient care, surgery, or teaching for an extended time period occurred in 72 cases (27%). Auditory sources of distraction were reported as distributions of count per case, including the median and interquartile range. The OR door opened and closed a median of 41 times per case (IQR 32–54). Machine alarm and loud noise occurred medians of 68 times (IQR 45–95) and 15 times (IQR 6–27) per case, respectively. External communication, including telephone and pager ring, was heard a median of 6 times per case.

Table 2 Frequency and characteristics of intraoperative distraction source
Full size table

Comparison of cases with and without perceived distraction

Out of 265 procedures performed, the attending surgeon reported perceived distraction in 120 cases (45%) (Table 3). A larger proportion of patients had previous abdominal surgery in cases where the attending surgeon felt distracted compared to the cases without perceived distraction (69% vs. 54%, p = 0.022). In cases with perceived distraction, the cases were longer in procedure duration (a median of 103 min (IQR 81–118) vs. 87 min (IQR 66–106), p < 0.001) and the OR door opened more frequently (a median of 46 times (IQR 34–62) vs. 39 times (30–48), p < 0.001). A larger proportion of cases had any irrelevant conversation in the cases with perceived distraction (35% vs. 21%, p = 0.012).

Table 3 Comparison of patient-level and intraoperative factors by perceived distraction
Full size table

Multivariable logistic regression analysis

One patient was excluded from multivariable analysis due to missing data on history of previous abdominal surgery (Table 4). Cases with any irrelevant conversation (odds ratio 2.14, 95% CI 1.16–3.94, p = 0.015) and patient history of previous abdominal surgery (odds ratio 2.2, 95% CI 1.18–4.1, p = 0.013) were independently associated with attending surgeon feeling distracted after adjusting for relevant patient-level and intraoperative factors. Goodness of fit for the multivariable model was adequate (‘c’ statistics = 0.73, insignificant Hosmer and Lemeshow test). A tenfold cross-validation demonstrated robust internal validity of variable selection strategy (Table S1).

Table 4 Multivariable logistic regression analysis modeling perception of distraction
Full size table

Discussion

In this prospective study of 265 consecutive patients undergoing elective laparoscopic gastrointestinal cases using detailed analysis of intraoperative audio–video data, we reported that auditory and cognitive distractions occurred frequently in the OR. Further, the attending surgeon in our study reported perceived distraction in 45% of the cases. The cases with perceived distraction were more likely to have irrelevant conversations, more frequent OR door opening, longer procedure duration, and patients with previous abdominal surgery, compared to the cases without perceived distraction. In multivariable regression model, irrelevant conversation (odds ratio 2.14, 95% CI 1.16–3.94, p = 0.015) and patient history of previous abdominal surgery (odds ratio 2.2, 95% CI 1.18–4.1, p = 0.013) were associated with perceived distraction after risk adjustment.

Our study characterized auditory and cognitive sources of distractions in the OR using video–audio intraoperative data in a structured way. These distractions were identified and time stamped along the procedure timeline using our novel data recording system [7]. This allowed for accurate characterization of distractions without reliance on written patient records, which often lack such detail, or on human observers to be present in the OR [18]. For example, irrelevant conversation was identified when a member of the OR team was engaged in a prolonged discussion about non-clinical matter, whereas teaching activity was determined when a prolonged discussion or skill demonstration relevant to the case being performed took place. Our reported incidences of intraoperative distractions were similar to previously published studies. For instance, Healey et al. identified an average of 33 door openings per case [2]. Further, 15 teaching activities in 48 cases and 4 teaching activities in 19 cases were reported in two separate publications [6, 19]. These results suggested that the video–audio data could be feasibly used to characterize intraoperative sources of distractions. The strength of our study was in the comprehensiveness of distraction classification and the consecutive nature of prospective cohort in a 2-year time period.

We adopted a human-factors self-reported questionnaire to assess surgeon’s perceived distraction. Human-factors engineering is a discipline that studies interactions among people, tools, and the environment within a system [15]. Its aim is to allow the system to be optimally designed and calibrated based on the strengths and limitations of human performance [20]. In the present study, the attending surgeon expressed perceived distraction in 45% of the cases. In the risk-adjusted multivariable analysis, patient history of previous abdominal surgery and irrelevant conversations were independently associated with perceived distraction. This finding can be extrapolated to suggest that the attending surgeon in this study has developed resilience to process several sources of distraction as non-distracting factors during procedures. The attending surgeon in this study is an experienced surgeon who has worked in the same hospital for more than 10 years. A systematic review of the effects of distraction on surgical performance in simulated settings demonstrated that inexperienced surgeons had difficulty adapting to distracting environments, but experienced surgeons showed resiliency mechanisms that were protective from certain distractors [11]. Patient history of previous abdominal surgery may have created altered anatomy and adhesions, often resulting in more complex and longer procedures [21]. The combination of a more challenging case and longer procedure durations may have resulted in the attending surgeon perceiving distraction. Unlike patient history of previous abdominal surgery, irrelevant conversation in the OR is a modifiable factor associated with perceived distraction. In other studies, such as Weigl et al., irrelevant conversation was associated with increased feeling of distraction [22]. Further, when sources of distraction were stratified by the severity of distraction using a nine-point ordinal scale, Healey et al. [23], Sevdalis et al. [19], and Wheelock et al. [24] independently showed that irrelevant communication was ranked as high-intensity distraction, often resulting in operation flow interruptions [2]. Based on the principles of human-factors engineering, irrelevant conversation in this study could be identified as a modifiable environmental factor that could influence human performance. Thus, efforts must be targeted in creating system-level interventions to decrease irrelevant conversations in the OR, especially when more challenging cases are taking place, in order to optimize surgical performance and subsequently, patient outcomes.

There were potential limitations to our study. Our study was based on a single surgeon experience, which limited the generalizability of our results. Our video–audio data capturing platform is now installed in several centers around the world and several surgeons of various demographic backgrounds, surgical specialties, and hospital settings are participating in on-going prospective studies. We plan to investigate validity of our results with a larger, multicenter cohort. Second, as with any study design where an intervention is implemented to observe individuals, there is a concern for possible intentional alteration of behavior and performance, a phenomenon known as the Hawthorne effect. In order to minimize this possible effect, a 1-year pilot phase was conducted to allow surgical teams to familiarize with the study design and environment. Lastly, perception of distraction is a complex construct, and the validity evidence for using our questionnaire to measure this construct is still preliminary.

In conclusion, this prospective cohort study of consecutive patients undergoing elective laparoscopic gastrointestinal surgery involving one attending surgeon demonstrated that while auditory and cognitive sources of distractions frequently occurred in the OR, the surgeon perceived distraction when cases had irrelevant conversations. Future research will continue to examine how surgeons interact with sources of distraction in the OR in order to create system-level intervention strategies to mitigate the potential risk associated with increased stress and cognitive workload due to perceived distraction. The “sterile cockpit” concept utilized in aviation, in which non-essential communication is prohibited during critical phases of flight, is an example of an implementable intervention that may be applied to surgery to lessen the impact of perceived distraction during high-risk phases of surgical procedures [6, 25, 26]. Moreover, future investigations would be aided by incorporating prospective data collection using comprehensive video–audio recording devices during cases, such as the OR Black Box® platform, to allow more in-depth analysis into the associations between distraction and surgical performance [1, 18, 27].