Validation of a new method to assess estimated blood loss in the obstetric population undergoing cesarean delivery

Saoud, Fawzi; Stone, Amanda; Nutter, Anna; Hankins, Gary D.V.; Saade, George R.; Saad, Antonio F.

doi:10.1016/j.ajog.2019.06.022

Cited by 22 publications

(12 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This post-hoc cutoff of 500mL represents a potential clinical marker of need for transfusion. The tendency for human experts to underestimate blood loss is well documented, [33][34][35][36] corroborated by our findings, and may result in delayed recognition of life-threatening hemorrhage.…”

Section: Benchmark Performance Of Human Expertssupporting

confidence: 83%

Expert Surgeons and Deep Learning Models Can Predict the Outcome of Surgical Hemorrhage from One Minute of Video

Pangal

Kugener

Zhu

et al. 2022

Preprint

View full text Add to dashboard Cite

Background: Major vascular injury resulting in uncontrolled bleeding is a catastrophic and often fatal complication of minimally invasive surgery. At the outset of these events, surgeons do not know how much blood will be lost or whether they will successfully control the hemorrhage (achieve hemostasis). We evaluate the ability of a deep learning neural network (DNN) to predict hemostasis control ability using the first minute of surgical video and compare model performance with human experts viewing the same video. Methods: The publicly available SOCAL dataset contains 147 videos of attending and resident surgeons managing hemorrhage in a validated, high-fidelity cadaveric simulator. Videos are labeled with outcome and blood loss (mL). The first minute of 20 videos was shown to four, blinded, fellowship trained skull-base neurosurgery instructors, and to SOCALNet (a DNN trained on SOCAL videos). SOCALNet architecture included a convolutional network (ResNet) identifying spatial features and a recurrent network identifying temporal features (LSTM). Experts independently assessed surgeon skill, predicted outcome and blood loss (mL). Outcome and blood loss predictions were compared with SOCALNet. Results: Expert inter-rater reliability was 0.95. Experts correctly predicted 14/20 trials (Sensitivity: 82%, Specificity: 55%, Positive Predictive Value (PPV): 69%, Negative Predictive Value (NPV): 71%). SOCALNet correctly predicted 17/20 trials (Sensitivity 100%, Specificity 66%, PPV 79%, NPV 100%) and correctly identified all successful attempts. Expert predictions of the highest and lowest skill surgeons and expert predictions reported with maximum confidence were more accurate. Experts systematically underestimated blood loss (mean error -131 mL, RMSE 350 mL, R2 0.70) and fewer than half of expert predictions identified blood loss > 500mL (47.5%, 19/40). SOCALNet had superior performance (mean error -57 mL, RMSE 295mL, R2 0.74) and detected most episodes of blood loss > 500mL (80%, 8/10). In validation experiments, SOCALNet evaluation of a critical on-screen surgical maneuver and high/low-skill composite videos were concordant with expert evaluation. Conclusion: Using only the first minute of video, experts and SOCALNet can predict outcome and blood loss during surgical hemorrhage. Experts systematically underestimated blood loss, and SOCALNet had no false negatives. DNNs can provide accurate, meaningful assessments of surgical video. We call for the creation of datasets of surgical adverse events for quality improvement research.

show abstract

Section: Benchmark Performance Of Human Expertssupporting

confidence: 83%

Expert Surgeons and Deep Learning Models Can Predict the Outcome of Surgical Hemorrhage from One Minute of Video

Pangal

Kugener

Zhu

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…An average of 15 surgical sponges were used per case (mean ¼ 15.1 AE 4.9; median ¼ 14.5; [12][13][14][15][16][17]). The gravimetric QBL on the sponges was substantially higher than the hemoglobin assayed amount of blood on the sponges in every case (mean absolute difference ¼ 480 AE 182 mL).…”

Section: Resultsmentioning

confidence: 99%

“… 7 16 Using the change in hemoglobin from the preoperative level to the value following delivery, Saoud et al confirmed the accuracy of the colorimetric method. 17 Katz et al recently demonstrated that quantifying blood loss using the colorimetric method resulted in improved identification of postpartum hemorrhage, changes in patient management and cost savings. 18 Further studies regarding the utility of accurate QBL using the colorimetric method and its effect on reduction of maternal morbidity and mortality are needed.…”

Section: Strengths and Limitationsmentioning

confidence: 99%

Limitations of Gravimetric Quantitative Blood Loss during Cesarean Delivery

Thurer

Doctorvaladan

Carvalho

et al. 2022

AJP Rep

View full text Add to dashboard Cite

Objective This study examined the accuracy, sources of error, and limitations of gravimetric quantification of blood loss (QBL) during cesarean delivery. Study Design Blood loss determined by assays of the hemoglobin content on surgical sponges and in suction canisters was compared with QBL in 50 parturients. Results QBL was moderately correlated to the actual blood loss (r = 0.564; p < 0.001). Compared with the reference assay, QBL overestimated blood loss for 44 patients (88%). QBL deviated from the assayed blood loss by more than 250 mL in 34 patients (68%) and by more than 500 mL in 16 cases (32%). Assayed blood loss was more than 1,000 mL in four patients. For three of these patients, QBL was more than 1,000 mL (sensitivity = 75%). QBL was more than 1,000 mL in 12 patients. While three of these had an assayed blood loss of more than 1,000 mL, 9 of the 46 patients with blood losses of less than 1,000 mL by the assay (20%) were incorrectly identified as having postpartum hemorrhage by QBL (false positives). The specificity of quantitative QBL for detection of blood loss more than or equal to 1,000 mL was 80.4%. Conclusion QBL was only moderately correlated with the reference assay. While overestimation was more common than underestimation, both occurred. Moreover, QBL was particularly inaccurate when substantial bleeding occurred. Key Points

show abstract

“…Major societies have focused on educating their members, on the awareness, prevention, recognition, monitoring and management of peripartum hemorrhage [4]. Perinatal care providers have evaluated numerous strategies of quantifying blood loss accurately and are using them in clinical practice in assessing blood loss and guiding transfusions during obstetric hemorrhage [5][6][7]. However, one of the major obstacles confronting clinicians is the lack of a simple, user friendly point of care (POC) coagulation monitoring device, which provides timely and accurate results to guide clinical management of unexpected obstetric hemorrhage.…”

Section: Introductionmentioning

confidence: 99%

Efficacy of sonorheometry point of the care device in determining low fibrinogen levels in pregnant blood: an invitro dilution and reconstitution study

Kodali

Karuppiah

Bharadwaj

et al. 2021

J Clin Monit Comput

View full text Add to dashboard Cite

Quantra® Hemostasis Analyzer is a Point of the care device that uses ultrasound technology to assess clot formation. In this study, we establish how Quantra® system performs compared to conventional coagulation tests at low levels of fibrinogen in the blood obtained from pregnant women. 24 mL blood was obtained from each healthy parturient. Blood was analyzed for Quantra® variables (Q): Clot time (CT), Clot stiffness (CS), platelet contribution to CS (PCS), fibrinogen contribution to CS (FCS), and conventional coagulation (CL) tests: PT, aPTT, INR, Factor VIII and fibrinogen. 6 ml blood were centrifuged to obtain pregnant plasma. 30 mL of saline was added to 10 mL of blood to simulate crystalloid resuscitation (DB) and was evaluated for Q and CL. Fractions of pregnant plasma, or nonpregnant plasma (Blood Bank) was added to DB to obtain 15% and 30% clotting factor enriched samples. 4 ml of DB was added to 4 ml of original blood (1:1) to obtain the final sample (resus). Each of the samples were analyzed for Q and CL parameters. Regression analysis and Receiving Characteristics Curves were used to study the relationship between Quantra variables and CL tests. There were remarkably high linear correlations between Fibrinogen and CS (R = 0.93, P < 0.001), fibrinogen and FCS (R = 0.77, P < 0.001). An FCS value 2.45 (sensitivity of 79.2 and specificity of 97.3%), and CS value 10.85 hPa (sensitivity of 83% and specificity of 100%) predicted fibrinogen of 200 mg/dL. This study demonstrates a good correlation between Quantra® CS, FCS and serum fibrinogen. Clinical Trial Number: NCT04301193.

show abstract

Validation of a new method to assess estimated blood loss in the obstetric population undergoing cesarean delivery

Cited by 22 publications

References 10 publications

Expert Surgeons and Deep Learning Models Can Predict the Outcome of Surgical Hemorrhage from One Minute of Video

Expert Surgeons and Deep Learning Models Can Predict the Outcome of Surgical Hemorrhage from One Minute of Video

Limitations of Gravimetric Quantitative Blood Loss during Cesarean Delivery

Efficacy of sonorheometry point of the care device in determining low fibrinogen levels in pregnant blood: an invitro dilution and reconstitution study

Contact Info

Product

Resources

About