Accuracy of visually estimated blood loss in the operating room: a simulation study

Abstract Background Visually estimated blood loss (VEBL) is the most widely used method for assessing intraoperative hemorrhage, yet it is highly subjective and inaccurate. Prior studies have focused on surgeons and anesthesiologists; data on the broader operating room workforce are limited. The primary aim was to quantify VEBL accuracy, defined as the difference between estimated and actual blood volume, in a multidisciplinary operating room cohort. The secondary, exploratory aim was to examine whether accuracy was associated with routine use of VEBL or with years of clinical experience. Methods Fifty-five operating room personnel, including attending surgeons, surgical residents, anesthesiologists, anesthesia residents, nurse anesthetists, and nurses, estimated blood volume at 8 stations displaying surgical materials prepared with diluted expired packed red blood cells (20–300 mL). The primary outcome was the difference between estimated and actual blood volume at each station. Two secondary comparisons were made: providers who routinely perform VEBL versus registered nurses, and attending versus resident physicians. Between-group differences were compared using two-sample t tests and reported as mean differences with 95% confidence intervals; p-values for the 18 subgroup comparisons were adjusted using the Benjamini-Hochberg false discovery rate procedure. Results Accuracy varied widely, with overestimation the predominant error. The mean per-participant difference was + 124.1 mL (standard deviation 129.4), and 46 of 55 participants (83.6%) overestimated on average. Stations 1 and 3 (both 300 mL; station 3 an irrigation-contaminated canister) showed marked overestimation, whereas station 4 (225 mL, chux pad) was dominated by underestimation. Neither routine use of VEBL nor years of clinical experience was associated with a consistent difference in accuracy (all-stations mean difference 11.14 mL, 95% confidence interval − 69.88 to 92.16, and 49.31 mL, -62.59 to 161.20, respectively). After adjustment for multiple comparisons, no station-level difference remained statistically significant (smallest adjusted p = 0.252). Conclusions VEBL accuracy was low across all provider roles studied, including nursing staff. Experience did not appear to improve accuracy, suggesting that structured VEBL training may be beneficial. Given the small subgroups and number of comparisons, subgroup findings are hypothesis-generating. Because VEBL is used alongside laboratory hemoglobin and hemodynamic status in resuscitation decisions, improved accuracy would strengthen that process.