Human Factors Approach to Assess Physician Workload in Elective Plastic Surgery Breast Procedures

In the US, Medicare and private insurers may use resource-based relative value scales for physician reimbursment for procedures. These payments comprise practice expenses, liability insurance, and the largest component, physician workload.¹ The relative value unit (RVU) is a consensus-derived indicator that can capture the workload for a given procedure.² Although there is disagreement as to how to quantify physician workload, higher work RVUs are assigned to reflect greater complexity. In previous studies to assess surgeon workload measures, operative time was one of the most consistently measured proxies to the corresponding assigned RVU.^1,3-5 When defining physician workload in the clinical setting of medical reimbursement, it is recognized that the assigned RVU does not take into account all activities and tasks that may be involved in the case, including teaching and distractions.⁴  

The operating room (OR) is a complex, dynamic, high-stress environment that involves intimate interactions among the OR environment, technology, and multidisciplinary staff within a system to complete various procedures. Multiple factors are recognized that can impact performance, flow, and time efficiency in the OR, including complications,⁶ concurrent surgeries,⁷ disruptions, and nonroutine events.^8,9 However, the role of these factors has been minimally studied in the context of the OR environment. Emerging work from the human factors field has shown that these same multiple factors similarly affect the surgeon and their workload.^10,11 It is critical to note that from a surgeon’s perspective, workload is often synonymous with caseload, which is reflective of the current literature; yet the field of human factors considers workload more broadly—the human ‘cost’ incurred to perform a task.¹² Underestimation of the true workload can be associated with increased demands on the services, increased physician fatigue, stress, and burnout.⁴ 

The International Ergonomics Association defines human factors as a scientific discipline that seeks to understand interactions among humans and the other elements of a system with the goal of optimizing human well-being and system performance.¹³ Therefore, incorporating human factors research into the surgical domain may assist in addressing workload knowledge gaps to better understand procedural factors, workflows, and workload for surgeons within the OR and provide an evidence base for targeted quality improvement to perform procedures safely and effectively. Despite the growing volume of research into human factors in the area of surgical workload, there is a dearth of available evidence quantifying intraoperative workload across surgical specialities and specific procedures.^14-17 

The American Society of Plastic Surgery reports breast procedures as the most frequently performed cosmetic procedures, with over 349,000 cases and over 220,000 reconstructive surgeries performed in the US in 2020.¹⁸ Over 250,000 new cases of invasive breast cancer were diagnosed in 2017, accompanied by a steady annual growth of reconstructive procedures that has resulted in a 35% increase in volume since 2000.¹⁸ In anticipation of continued patient demand for breast-specific plastic surgery procedures, this study aimed to prospectively quantify self-reported surgeon workload within plastic surgery and specific elective breast procedures in particular to identify potential patient or procedural influences on intraoperative workload. 

Setting and Participants

This study was conducted prospectively at a quaternary academic institution located in the midwestern US. As part of a larger study,¹⁴ attending surgeons employed within the Department of Surgery at the institution were recruited for the study. From May 2017 through July 2017, participating surgeons were asked to complete a workload questionnaire following each procedure within a 3-month period or until they completed at least 20 questionnaires. Participants were advised that taking part in this study was optional; responses would remain anonymous and would have no impact on their employment. All ORs were comparable in terms of environment, staffing, equipment, and size. The Institutional Review Board approved this study (Protocol #17-002691) and complied with the American Psychological Association Code of Ethics. For this subset analysis, survey responses were analyzed from plastic surgeons performing breast-related procedures, including breast reduction, reconstruction, augmentation, and gender-confirming mastectomy.  

Research Protocol

Plastic surgeons participating completed a demographic survey that included questions on their age, height, weight, and operative experience. Operative experience was evaluated as part of data analysis within this study. Study coordinators checked surgical schedules of participants and recorded the procedural type and sequential order of cases each day using Qualtrics XM survey software, v2017 (Qualtrics). Participants then received automated emails with a secure link to a questionnaire for each procedure and were asked to complete it as soon as possible following the procedure. 

Workload Questionnaire

The electronic workload questionnaire contained 8 items. Six items came from the validated National Aeronautics and Space Administration Task Load Index (NASA-TLX),¹² which included questions on mental demand, physical demand, temporal demand, performance, effort, and frustration. One item on distraction came from the validated Surg-TLX.¹⁹ The final item pertained to surgeon expectation on procedural difficulty going into the case, reported as “less difficult than expected,” “as expected,” or “more difficult than expected.” The NASA-TLX and Surg-TLX items were used to rate task demands on a 20-point scale (1 = low, 20 = high). 

Patient and Procedural Factors

Following the survey portion of the study, patient and procedural factors were gathered from the electronic health record (EHR). At the institution of study, patients may opt out of any use of their EHR information for research purposes; therefore, only 67 of 72 completed surveys have corresponding patient and procedural data because 5 patients did not consent to research participation. A total of 58 surveys that were completed were related to breast procedures. Patient and procedural factors collected from the EHR included age, gender, body mass index (BMI), American Society of Anaesthesiologists (ASA) category, procedure performed, and skin-to-skin procedural duration. 

Procedures were further broken into the following categories: (1) breast reduction; (2) implant- or tissue expander–based procedures, including augmentation/reconstruction and implant removal; (3) pedicled flap-based breast reconstruction; (4) broader range of revision and implant exchange performed simultaneously with other procedures, such as fat grafting, capsule surgery, and contralateral symmetrization; (5) gender-affirming mastectomy procedures; (6) free flap–based breast reconstruction; and (7) minor breast procedures, including micro-fat grafting only, nipple reconstruction, and excision of benign lesions.

Data Analysis

Data analyses were performed in JMP Pro v13 (SAS Institute Inc). Normally distributed variables (eg, surgical duration and survey responses) were reported as means and standard deviations (SD). Nonnormally distributed variables (eg, surgeon and patient characteristics) were reported as medians with interquartile ranges (IQRs). Comparisons across categorical variables (eg, procedure types, expectation) for normally distributed variables required analysis of variances (ANOVAs), while Kruskal-Wallace tests were used for comparisons of nonnormally distributed variables. Post hoc Mann-Whitney tests using Bonferroni correction were performed following ANOVAs to identify group differences. Pearson correlations were performed to identify associations between and among patient, procedural, and survey data. A P < .05 was considered significant for all comparisons. 

Four attending surgeons who performed elective plastic surgery breast procedures participated in this study. During the study period, surgeons completed surveys for 58 unique plastic surgery breast cases. Approximately 51 of 58 (90%) survey responses were completed within 24 hours of being sent. 

Collected procedures (N = 58) included a range of immediate and delayed reconstructions, revisions, and elective primary cosmetic procedures. The mean (SD) age of patient was 49 (15.2) years. The mean (SD) BMI was 28 (4.8) kg/m². Implant-based surgery (ie, implant or tissue expander alone, implant with fat grafting/revisional implant-based surgery) accounted for 62% of all the breast procedures. Three abdominal-based free flaps (5%) and 1 pedicled flap (2%) were included the cohort. Procedures were performed in combination with general surgery in 8 of 58 (14%) cases and consisted of implant/expander only (n = 5), oncoplastic breast reduction (n = 1), mastectomy (n = 1), and contralateral reduction (n = 1) procedures. A total of 40 patients (69%) underwent bilateral procedures. Immediate reconstruction was performed in 5 cases (9%).

The median (IQR) length of all procedures was 128 (58-149) minutes (Table). Surgical duration differed significantly by procedure type (P < .001), with the longest procedure being free flap breast reconstruction (median [IQR], 600 [550-1170] minutes), followed by pedicled flap (median, 316 minutes). Implant-only based procedures were the shortest (median [IQR], 94 [51-134] minutes). The median (IQR) time for bilateral procedures was 134 (108-232) minutes compared with 56 (34-88) minutes for unilateral procedures; this difference was statistically significant (P < .001). Patient BMI had a moderate positive correlation with increased duration of surgical procedures (R = 0.44; P = .001); however, there was no correlation between length of surgery and patient age. Comparison of BMI and procedure type did demonstrate a difference between all breast procedure categories (P = .02). 

Modified NASA-TLX Workload

Across all procedures, the levels of effort, mental demand, physical demand, and temporal demand were rated high. Surgical duration had moderate positive correlations with mental demand (R = 0.54; P < .001), physical demand (R = 0.52; P < .001), effort (R = 0.44; P = .001), temporal demand (R = 0.41; P = .002), frustration (R = 0.31; P = .02), and distractions (R = 0.44; P < .001). However, there was no significant correlation observed between surgical duration and perceived performance. 

Evaluating composite workload and subscales by procedure type identified significant differences across mental demand, physical demand, temporal demand, effort, and distraction (P < .05; Figure 1). Free flap breast reconstructions were reported as having the highest workload across all subscales and composite NASA-TLX workload score (mean [SD], 84.7 [14.0]). In contrast, minor procedures, such as nipple reconstruction, excision of benign lesions, and fat grafting only breast procedures had consistently the lowest reported workload scores across all subscales. 

Comparing workload by various case characteristics yielded some differences. Bilateral and unilateral procedures differed significantly. When compared with unilateral procedures, bilateral procedures required significantly higher composite workload overall (P = .002), including higher mental demand (P = .01), physical demand (P = .002), temporal demand (P < .0001), effort (P = .004), frustration (P = .03), and distractions (P = .001) (Figure 2). Furthermore, case timing (immediate vs delayed) demonstrated minimal effects on surgeons’ perceived workload; the mean (SD) distraction score of 9.75 (7.9) was significantly higher than that for delayed cases for immediate cases (7.0 [7.3]; P = .04).

There were no significant correlations between BMI, patient age, and composite NASA-TLX workload scores or subscales. There was a strong positive correlation of the degree of distractions on increased mental workload during an operation that was statistically significant (R = 0.68; P < .001). Similarly, the degree of perceived distractions experienced during a procedure was positively correlated with physical demand (R = 0.61; P < .001), effort (R = 0.65; P < .001), temporal demand (R = 0.78; P < .001), and frustration (R = 0.57; P < .001). However, increased level of distraction did not significantly impact perceived performance (P > .05).

Workload Threshold

Among the collected procedures, 27.6% of procedures scored above the threshold for composite workload (ie, score of 50 out of 100). Persistent mental and physical demands greater than the threshold may translate into errors in performance and increased risk of musculoskeletal injury. Adapting the threshold to individual subscales (ie, threshold set at 10 out of 20), the surgeons’ perceived mental demand exceeded the threshold 41.4% of the time for plastic surgery breast procedures. Surgeons also perceived their physical demand to be exceeding the threshold (ie, score of 10 or higher) during 40% of their procedures. Breast reduction (5 of 6, 83%), free flap (3 of 3, 100%), gender-affirming mastectomy (2 of 2, 100%), and pedicled flap (1 of 1, 100%) procedures were regularly rated above the midpoint threshold for composite workload. Figure 3 summarizes the frequency of procedure types that exceeded the threshold for mental and physical demand. 

Expected Difficulty Level

Across the analyzed plastic surgery breast procedures, 58% of surgeons felt that the procedure had gone as expected and 16% reported the procedure to be less difficult than initially expected, while 26% felt the procedures were more difficult than anticipated. Figure 4 summarizes the distribution of expected difficulty level based on breast procedure type. Procedures that were perceived as more difficult had a significantly greater median length of procedure (149.5 minutes) than those that were as difficult as expected (105 minutes) and those that were less difficult (99 minutes; P = .018). However, no relationships were identified between expected difficulty level and timing of the procedure (delayed vs immediate) (P > .05) or procedure type (P > .05). Furthermore, patient factors (BMI, age, ASA) did not differ by expected difficulty level.

The complex OR work system can require high workloads, both physically and cognitively, for the surgical team. To achieve successful flow and completion of a procedure, various factors, such as patient characteristics, case complexity, surgeon and team dynamics, use of technology, and the OR environment, must interact cohesively. An individual’s workload is a broad term derived from human factors and behavioral psychology that refers to the human capacity to perform a task as matched to the task demands.²⁰ Factors that increase workload (eg, complexity of the procedure, complex patient factors, time pressure, use of technology, conflicting priorities) may increase mental demand, which may in turn affect the operative team’s ability to anticipate or adapt to unexpected events and impair performance.^11,21 

The NASA-TLX questionnaire is a multidimensional validated tool that subjectively quantifies the perceived demands of a task on the individual.¹² NASA-TLX has been used in a variety of high-risk industries including aviation and healthcare, where it was further adapted for surgery.^12,16,22 Through employment of NASA-TLX and similar workload questionnaires, human factors researchers have demonstrated workload differences in single-port versus 4-port laparoscopy,^23,24 across surgical team members,^23,25 and across surgical specialties.²³ Mental and physical workloads are highest among operating surgeons and residents compared with other OR team members,^21,26 and it is expected to increase as innovative procedures continue to be developed and operative complexity increases. The evaluation of mental effort and physical workload in plastic surgery has been limited, both in simulations and in OR settings. Although there is limited published literature in in this field within surgical specialties, the study of human factors in surgery has gained recognition and continues to grow. 

Breast procedures are among the most common elective procedure performed in plastic surgery, but there is a broad range in complexity from minor revision surgery to much longer and demanding microsurgical reconstructions, as reflected within this study. This pilot study helped to identify some factors during routine elective procedures in an academic setting that had an impact on physician workload. Some of these factors may have been perceived as predictable, such as increased time and workload for bilateral cases, increased BMI having a positive correlation with longer operative times, and higher demands across multiple subscales for flap reconstruction than for fat grafting and minor revision procedures. However, evaluating the subscales of cognitive workload and effort can help distinguish the workload ranges between categories of cases. The sensitivity of the instrument to different levels of complex plastic surgery procedures validates its use in the specialty. Furthermore, the assessment of composite workload scores similarly reflected graded measures of case complexity as guided by the RVU reimbursement weighting for elective breast procedures. Interestingly, the literature has demonstrated that operative times best correlated with RVU values in other surgical specialities.^2-4

In the current study, nearly 25% of cases were considered more difficult than expected. This distribution is similar to that found in a previous study by Lowndes et al (2018) that evaluated workload among surgeons across various surgical subspecialties and demonstrated greater than expected difficulty level in 22% of cases.¹⁴ Longer operative times were also independently associated with higher composite NASA-TLX scores, higher mental and physical demands, and a greater degree of perceived distractions; all of these differences were statistically significant. Perceived distractions were significantly positively correlated to greater mental demands, physical demands, and frustration during the case. 

This study highlighted that there are multiple interacting factors that may impact surgeon workload, including those working in an academic setting where there may be expectations to provide resident education and assist with secondary tasks or surgical cases with other teams. This pilot study was limited by lack of more in-depth analysis of potential additional contributors to the workload scores or distractions, such as sequence of the cases, having a regular or unfamiliar team, changeover of staff, outside calls, and resident/fellow contributions. Other limitations in this pilot study included the small number of participating surgeons, the limited range of experience levels, and the exclusion of other members of the team members or trainees. Although the response rate was satisfactory, the use of questionnaires for this study was subjective and voluntary and therefore subject to bias. 

Although this study was limited by the sample size from the questionnaires completed, the study covered the breadth of complexity and procedure type to represent a broad breast practice among plastic surgeons. The trends identified in these findings for commonly performed routine elective procedures, therefore, have generalizability to other institutions.

Considerations for future work include the evaluation of workload in a larger multicenter prospective study with a broad range of faculty seniority and assessment of residents and fellows. The study could be broadened to encompass other areas of plastic and reconstructive surgery subspecialty areas. Building upon this foundation of knowledge, further investigation to correlate perceived surgeon workload and clinical patient outcomes would be considered as part of future research and may have potential impact on clinical practice.    

The impact of cognitive workload and human factors in plastic surgery procedures is limited in the published literature, and this study provided some initial insight into perceived workload of the primary surgeon during elective breast procedures. Identification of these factors may be important influential considerations when scheduling cases  to reduce distractions  and moderate workloads for surgeons balancing the demands of simultaneous cases while improving surgical workflow and operative times. Identifying ways to mitigate surgeons’ workload and mental demand has been shown to help to improve performance and allow more available mental resources to manage secondary tasks, including decision-making. Quantifying surgeon workload across a specialty may identify specific quality improvement opportunities to perform procedures more safely and more effectively. 

Affiliations: ¹Division of Plastic Surgery, Mayo Clinic, Rochester, Minnesota; ²Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, Minnesota; ³Department of Neurological Sciences, University of Nebraska Medical Center, Omaha, Nebraska; ⁴Department of Health Sciences Research, Mayo Clinic, Rochester, Minnesota

Correspondence: Susan Hallbeck, PhD; hallbeck.susan@mayo.edu

Presentation: This paper was presented at the American Society of Plastic Surgery, The Meeting, Resident Section, San Diego, USA 2019. 

Ethics: The Institutional Review Board approved this study (Protocol #17-002691).

Disclosures: All authors confirm that they have no financial disclosure or conflicts of interest to declare