Estimating Blood Loss

    loading  Checking for direct PDF access through Ovid

Excerpt

Surgical hemostasis is a central component of patient blood management,1 for which there is now a growing list of pharmacologic agents with the potential to be effective in reducing perioperative blood loss.2 The most straightforward means to demonstrate efficacy of such interventions in clinical trials is in the reduction of mortality or in allogeneic blood transfusions, which can be high hurdles that may be difficult to achieve, depending on the clinical trial design, patient setting, sample size cohorts, etc. At one extreme of this spectrum is the clinical randomization of an antifibrinolytic in significant head injury (CRASH-2) study, which enrolled a very large number of trauma patients in order to be able to demonstrate improved mortality in patients treated with adjuvant tranexamic acid (TXA) therapy.3 At the other end of the spectrum for efficacy is the demonstration of reduced perioperative bleeding (ie, improved surgical hemostasis) in patients undergoing hip replacement surgery who were treated with TXA.4 In this instance, perisurgical blood loss may be only an indirect metric for demonstrating improved clinical patient outcomes, although there was an indication that allogeneic blood transfusions were reduced. Unfortunately, methods for quantifying estimated surgical blood loss (EBL) have long been felt to be inaccurate and, therefore, were variable and poorly correlated with other clinical outcomes.5
In this issue, Lopez-Picado et al6 have used data from their previously published clinical trial4 to assess direct measures (external blood loss) and indirect calculations for EBL (using 4 previously published formulas plus a fifth formula developed by the authors). Notably, although the majority of formulas reaffirmed a reduction in blood loss with TXA administration during total hip arthroplasty compared to placebo, calculation using the Gross formula resulted in no significant difference in EBL between treatment groups on postoperative day 4. Additionally, as the amount of blood loss increased, the agreement in EBL among formulas decreased, highlighting the difficulty in accurately quantifying large-volume EBL. Thus, the authors concluded that these metrics are useful for evaluating the efficacy of interventions aiming to decrease blood loss but do not ensure that the values themselves are sufficiently accurate for absolute measuring.
What does this mean for the potential role(s) of EBL for use as a clinical metric? Estimating blood loss has a number of potentially important clinical roles (Table). But in each of these instances, the method for measuring EBL must be sufficiently accurate and validated in order to be accepted as a quality or patient care metric.
One approach is the use of validated bleeding severity scales such as for clinical studies investigating hemostatic agents.7 These are based on a clinician-reported scale that is based on the essential elements and criteria to evaluate a clinician-reported scale of the US Food and Drug Administration.8 Alternative approaches to EBL include gravimetric, colorimetric, photometric, spectrophotometric, and feature extraction imaging.9–11
Ironically, the least important role for quantifying EBL may be as an indicator for the need for red blood cell transfusion therapy. This is because while concurrent estimates of blood loss are relevant, they should be accompanied by other clinical parameters also important for transfusion decisions, including patient vital signs and serial hemoglobin level determinations. However, as a post hoc measure of hemostasis, EBL can be used as a metric for comparison of surgical or anesthetic techniques. For example, determinations of reduced bleeding under regional anesthesia compared to general anesthesia12–15 may influence choice of anesthetic technique and recommendations for standardized clinical pathways for routine surgery. Similarly, EBL may be used as part of peer performance review in surgical techniques for hospital credentialing.
    loading  Loading Related Articles