Automated Quantification of Blood Loss versus Visual Estimation in 274 Vaginal Deliveries.

OBJECTIVE: The aim of the study is to compare quantified blood loss measurement (QBL) using an automated system (Triton QBL, Menlo Park, CA) with visual blood loss estimation (EBL) during vaginal delivery. STUDY DESIGN: During 274 vaginal deliveries, both QBL and EBL were determined. The automated system batch weighs blood containing sponges, towels, pads, and other supplies and automatically subtracts their dry weights and also the measured amount of amniotic fluid. Each method was performed independently, and clinicians were blinded to the device's results. RESULTS: Median QBL (339 mL [217-515]) was significantly greater than median EBL (300 mL [200-350]; p < 0.0001). The Pearson's correlation between EBL and QBL was poor (r = 0.520) and the Bland-Altman's limits of agreement were wide (>900 mL). QBL measured blood loss >500 mL occurred in 73 (26.6%) patients compared with 14 (5.1%) patients using visual estimation (p < 0.0001). QBL ≥ 1,000 mL was recorded in 11 patients (4.0%), whereas only one patient had an EBL blood loss of 1,000 mL and none had EBL >1,000 mL (p = 0.002). CONCLUSION: Automated QBL recognizes more patients with excessive blood loss than visual estimation. To realize the value of QBL, clinicians must accept the inadequacy of visual estimation and implement protocols based on QBL values. Further studies of clinical outcomes related to QBL are needed. KEY POINTS: · QBL detects hemorrhage more frequently than visual estimation.. · Median QBL is significantly greater than median EBL.. · There is poor agreement between QBL and EBL..

Accurate Assessment of Blood Loss during Cesarean Delivery Improves Estimation of Postoperative Hemoglobin.

OBJECTIVE: To determine if accurate blood loss determination during cesarean delivery can improve the prediction of postoperative hemoglobin levels. STUDY DESIGN: This is a retrospective cohort study using visually estimated blood loss (traditional, n = 2,025) versus estimates using a mobile application that photographs sponges and canisters and calculates their hemoglobin content (device, n = 756). RESULTS: The correlation between the actual and predicted postoperative day 1 hemoglobin value (PPO1 Hgb) was better in the device group (R 2 = 0.519, correlation = 0.720) than in the traditional group (R 2 = 0.429, correlation = 0.655) (p = 0.005). For patients in the device group where the estimated blood loss was >1,000 mL (n = 53), the PPO1 Hgb was also better correlated with the actual value (R 2 = 0.319, correlation = 0.565) than the predictions using visually estimated blood loss for those patients in the device group whose visual estimation was >1,000 mL (n = 32) (R 2 = 0.035, correlation = 0.187) (p = 0.027). CONCLUSION: Implementation of a device that accurately measures blood loss allows for a better prediction of postoperative day 1 hemoglobin concentration than is possible using visual blood loss estimation. This improvement was seen in the entire patient group and was particularly prominent in patients with blood losses of > 1,000 mL.

Clinical Experience with the Implementation of Accurate Measurement of Blood Loss during Cesarean Delivery: Influences on Hemorrhage Recognition and Allogeneic Transfusion.

OBJECTIVE: This article compares hemorrhage recognition and transfusion using accurate, contemporaneous blood loss measurement versus visual estimation during cesarean deliveries. STUDY DESIGN: A retrospective cohort study using visually estimated blood loss (traditional, n = 2,025) versus estimates using a mobile application that photographs sponges and canisters and calculates their hemoglobin content (device, n = 756). RESULTS: Blood loss > 1,000 mL was recognized in 1.9% of traditional visual estimation patients, while measured blood loss of > 1,000 mL occurred in 8.2% of device patients (p < 0.0001). In both groups, this was accompanied by a greater decrease in transfusion-adjusted hemoglobin levels than occurred in patients without hemorrhage (p < 0.0001). Despite similar transfusion rates (1.6% in both groups), fewer red cell units were given to transfused patients in the device group (1.83 ± 0.58 versus 2.56 ± 1.68 units; p = 0.038). None of the patients in the device group received plasma or cryoprecipitate. Seven patients in the traditional group received these products (p = 0.088). Device use was associated with shorter hospital stays (4.0 ± 2.3 versus 4.4 ± 2.9 days; p = 0.0006). CONCLUSION: The device identified hemorrhages more frequently than visual estimation. Device-detected hemorrhages appeared clinically relevant. Blood product transfusion was reduced possibly due to earlier recognition and treatment, although further studies are needed to verify the conclusion.