Association Between Surgeon Stress and Major Surgical Complications

Abstract

Importance: Surgeon stress can influence technical and nontechnical skills, but the consequences for patient outcomes remain unknown.

Objective: To investigate whether surgeon physiological stress, as assessed by sympathovagal balance, is associated with postoperative complications.

Design, setting, and participants: This multicenter prospective cohort study included 14 surgical departments involving 7 specialties within 4 university hospitals in Lyon, France. Exclusion criteria consisted of patient age younger than 18 years, palliative surgery, incomplete operative time-stamping data, procedures with a duration of less than 20 minutes, and invalid surgeon heart rate variability (HRV) data. Data were accrued between November 1, 2020, and December 31, 2021, with 30-day follow-up completed on May 8, 2022. Analyses were performed from January 1 to May 31, 2024.

Exposure: Sympathovagal balance of the attending surgeon in the first 5 minutes of surgery.

Main outcomes and measures: Major surgical complications, extended intensive care unit stay, and mortality within 30 days, after adjustment via mixed-effects multivariable logistic regression for surgeon age, professional status, the time of incision, the random effect of the surgeon, and a composite risk score incorporating patient comorbidities and surgery characteristics. Sympathovagal balance was quantified by the low frequency to high frequency (LF:HF) ratio derived from HRV data measured by chest monitors worn intraoperatively. The LF:HF ratio was normalized at the surgeon level to the median value observed for each surgeon during the study period to control for baseline differences.

Results: A total of 793 surgical procedures performed by 38 attending surgeons were included in the analysis. Median patient age was 62 (IQR, 47-72) years, and 412 (52.0%) were female, with a median of 2 (IQR, 1-4) comorbidities. Median surgeon age was 46 (IQR, 39-52) years, 39 (78.9%) were male, and 22 (57.9%) were professors. Median surgeon heart rate was 88 (IQR, 77-99) beats per minute. Median surgeon LF:HF ratio was 7.16 (IQR, 4.52-10.72) before and 1.00 (IQR, 0.71-1.32) after normalization. Increased surgeon sympathovagal balance during the first 5 minutes of surgery was associated with significantly reduced major surgical complications (adjusted odds ratio [AOR], 0.63; 95% CI, 0.41-0.98; P = .04), though not with reduced intensive care unit stay (AOR, 0.34; 95% CI, 0.11-1.01; P = .05) or mortality (AOR, 0.18; 95% CI, 0.03-1.03; P = .05).

Conclusions and relevance: Increased surgeon stress at the beginning of a procedure was associated with improved clinical patient outcomes. The results are illustrative of the complex relationship between physiological stress and performance, identify a novel association between measurable surgeon human factors and patient outcomes, and may highlight opportunities to improve patient care.

Figure 1.. Flowchart of Included and Excluded Cases

Exclusion criteria were applied to optimize the reliability of heart rate variability (HRV) spectral analysis and the calculation of the low frequency to high frequency ratio. Polar data indicate HRV data obtained from chest belts (Polar H10; Polar Electro) for 5 minutes after the surgical incision. The final cohort included a mean (SD) of 21 (5) operations per surgeon (median, 16.5 [range, 3-66]).

Figure 2.. Surgeon Low Frequency to High Frequency (LF:HF) Ratio Normalization

Each box plot represents a single surgeon. Box plots with the same color between the panels represent the same surgeon. A, Box plots of the raw LF:HF ratio values for each surgeon across all cases completed during the study, arranged in order of ascending median value. B, Box plots of the normalized LF:HF ratio values for the same surgeons in the same order after normalizing by dividing each raw LF:HF ratio by the individual surgeon’s median value across all of their cases. Boxes indicate the 25th percentile, median, and 75th percentile. Dots represent data greater than 1.5 times the IQR from the median. Whiskers show the minimum and maximum ranges.

Figure 3.. Adjusted Association Between Normalized Low Frequency to High Frequency (LF:HF) Ratio and Surgical Outcomes

Effect plots demonstrating the association between surgeon normalized LF:HF ratio and the probability of the indicated outcome after adjusting for surgeon age, surgeon professional status, time of day of the initial incision, and the logarithm of the associated patient risk score. Continuous covariates that are not the primary exposure are centered at the cohort mean, while categorical ones are set to the reference value. The curve does not incorporate the random effect of the performing surgeon. The chosen range represents approximately 2 SDs (1 SD = 0.51) above and below the median (1.00) of the normalized LF:HF ratio. Shaded areas represent the 95% CI. ICU indicates intensive care unit.