Estimation of blood loss is inaccurate and unreliable

To determine the characteristics associated with improved accuracy or reliability of estimating operative blood loss.

Operating room personnel at a tertiary care hospital evaluated 3 operative simulations and provided estimations of blood loss. The simulations utilized precise, known volumes of porcine blood and saline on tapes, sponges, and in suction containers. Low volume (50 mL), mid volume (300 mL), and high volume (900 mL) blood loss scenarios were used in this simulation. Information collected included the blood loss estimation, the participant's occupation, years of experience in the operating room, confidence level in estimating blood loss, and their opinion as to which group would provide the most accurate estimation.

Sixty practitioners participated: 17 anesthesia providers, 22 surgeons, and 21 nurses and technicians. Overall, estimations were significantly inaccurate: scenario 1, mean error 52%; scenario 2, mean error 61%; scenario 3, mean error 85%. Ninety-five percent of participants provided estimations that had >25% error in at least 1 scenario. Only 27% demonstrated consistency in over or under-estimating the blood loss. There was no association between specialty, years of experience, or confidence in ability with consistency or accuracy of estimating blood loss.

This study demonstrates that visual estimation of operative blood loss is unreliable and inaccurate. No provider specialty, level of experience, or self-assessment of ability was associated with improved estimation. Blood loss estimations are not a reliable metric to judge physician performance or patient outcomes. Consideration should be given to alternative reporting of operative blood loss to better direct perioperative care.