フランス語版SURG-TLXの検証：手術ワークロード評価のための信頼性が高くタスク固有のツール
Georges Pfister1, Emeric Saguin2
1Orthopedic Departement, Percy Military Teaching Hospital, 101, avenue Henri Barbusse, 92140 Clamart, France.
Background:
Assessing intraoperative workload is essential for understanding performance variability, preventing errors, and supporting occupational well-being. The Surgery Task Load Index (SURG-TLX) offers greater contextual relevance for surgery than general tools such as the NASA-TLX, but no validated French version has been available.
Objective:
To translate the SURG-TLX into French and evaluate its psychometric properties in a real-world operating-room setting.
Methods:
Twenty operating-room professionals each rated two distinct procedures (Time 1, T1; Time 2, T2), yielding 40 evaluations. After each operation, participants completed the French SURG-TLX, the NASA-TLX, and a 1-10 perceived difficulty rating. Analyses included descriptive statistics, Pearson correlations (convergent/construct validity), paired t-tests, Bland-Altman plots, intraclass correlation coefficient (ICC[3,k]) for test-retest reliability, and linear mixed-effects models with participant as a random intercept.
Results:
Mean weighted SURG-TLX scores indicated moderate workload (T1: 9.28±4.27; T2: 8.37±2.98 on a 0-40 scale). Convergent validity was supported by strong correlations with NASA-TLX (T1: r=0.67, P<0.001; T2: r=0.81, P<0.001). Construct validity was confirmed by the association with perceived difficulty across all procedures (r=0.69, P<0.001). The paired comparison showed no T1-T2 difference (t(19)=0.94, P=0.36, d=0.21). Test-retest reliability was moderate (ICC[3,k]=0.48; 95% CI -0.12-0.76). In mixed-effects models, perceived difficulty predicted SURG-TLX (β=1.20±0.22, P<0.001), whereas time and demographics were non-significant.
Conclusion:
The French SURG-TLX shows solid preliminary validity, moderate temporal reliability, and sensitivity to task complexity in real surgical practice.
