RESUMO
PURPOSE: Among various methods for estimating blood loss, the gravimetric method is the most accurate; however, its use in routine practice is complicated. Although several equations have been proposed for this purpose, there is no consensus on the most suitable. METHODS: A cross-sectional study was conducted in seven secondary and tertiary hospitals between March and July 2018 including all patients undergoing total hip arthroplasty or hip prosthesis replacement under general or regional anaesthesia. We compared blood loss estimates obtained using the gravimetric method (weighing gauzes and pads and measuring volumes of blood collected by suction during surgery) and using three different equations, two of which considered intravenous fluids (CRYS 3.5 and 1.5) and a third which did not (the traditional equation). Additionally, intraclass correlation coefficients (ICCs) and Bland-Altman plots were used. RESULTS: The mean blood loss estimated using the gravimetric method was 513.7 ± 421.7 mL, while estimates calculated using the CRYS 3.5, CRYS 1.5 and traditional equations were 737.2 ± 627.4, 420.8 ± 636.2 and 603.4 ± 386.3 mL, respectively. Comparing these results, we found low levels of agreement (based on ICCs), except when using the traditional equation (ICC: 0.517). The limits of agreement comparing external blood loss with the estimates from the equations ranged from - 1655.6 to 1459.2 in the case of the CRYS 1.5 equation to - 839.6 to 1008.4 in the case of the traditional equation. CONCLUSIONS: For use in clinical practice, haematological index-based equations, regardless of whether they consider fluids administered, do not show sufficiently strong correlations with gravimetric estimates of intraoperative blood loss.
