Peripheral Blood Hematocrit in Critically Ill Surgical Patients: An Imprecise Surrogate of True Red Blood Cell Volume

RBC Inventory-Management System Based on XGBoost Model

It is difficult to predict RBC consumption accurately. This paper aims to use big data to establish a XGBoost Model to understand the trend of RBC accurately, and forecast the demand in time. XGBoost, which implements machine learning algorithms under the Gradient Boosting framework can provide a parallel tree boosting. The daily RBC usage and inventory (May 2014-September 2017) were investigated, and rules for RBC usage were analysed. All data were divided into training sets and testing sets. A XGBoost Model was established to predict the future RBC demand for durations ranging from a day to a week. In addition, the alert range was added to the predicted value to ensure RBC demand of emergency patients and surgical accidents. The gap between RBC usage and inventory was fluctuant, and had no obvious rule. The maximum residual inventory of a certain blood group was up to 700 units one day, while the minimum was nearly 0 units. Upon comparing MAE (mean absolute error):A:10.69, B:11.19, O:10.93, and AB:5.91, respectively, the XGBoost Model was found to have a predictive advantage over other state-of-the-art approaches. It showed the model could fit the trend of daily RBC usage. An alert range could manage the demand of emergency patients or surgical accidents. The model had been built to predict RBC demand, and the alert range of RBC inventory is designed to increase the safety of inventory management.

The use of biomarkers to describe plasma-, red cell-, and blood volume from a simple blood test

Plasma volume and red cell mass are key health markers used to monitor numerous disease states, such as heart failure, kidney disease, or sepsis. Nevertheless, there is currently no practically applicable method to easily measure absolute plasma or red cell volumes in a clinical setting. Here, a novel marker for plasma volume and red cell mass was developed through analysis of the observed variability caused by plasma volume shifts in common biochemical measures, selected based on their propensity to present with low variations over time. Once a month for 6 months, serum and whole blood samples were collected from 33 active males. Concurrently, the CO-rebreathing method was applied to determine target levels of hemoglobin mass (HbM) and blood volumes. The variability of 18 common chemistry markers and 27 Full Blood Count variables was investigated and matched to the observed plasma volume variation. After the removal of between-subject variations using a Bayesian model, multivariate analysis identified two sets of 8 and 15 biomarkers explaining 68% and 69% of plasma volume variance, respectively. The final multiparametric model contains a weighting function to allow for isolated abnormalities in single biomarkers. This proof-of-concept investigation describes a novel approach to estimate absolute vascular volumes, with a simple blood test. Despite the physiological instability of critically ill patients, it is hypothesized the model, with its multiparametric approach and weighting function, maintains the capacity to describe vascular volumes. This model has potential to transform volume management in clinical settings. Am. J. Hematol. 92:62-67, 2017. © 2016 Wiley Periodicals, Inc.

Mid-regional pro-adrenomedullin (MR-proADM), a marker of positive fluid balance in critically ill patients: results of the ENVOL study

Background

The optimal control of blood volume without fluid overload is a main challenge in the daily care of intensive care unit (ICU) patients. Accordingly this study focused on the identification of biomarkers to help characterize fluid overload status.

Methods

Sixty-seven patients were studied from ICU admission to day 7 (D₇). Blood and urine samples were taken daily and sodium and water balance strictly calculated resulting in a total cumulative assessment of ∆Na⁺ and ∆H₂O. Furthermore, plasmatic biomarkers (cortisol, epinephrine, norepinephrine, renin, angiotensin II, aldosterone, pro-endothelin, copeptine, atrial natriuretic peptide, erythropoietin, mid-regional pro-adrenomedullin (MR-proADM)) and Sequential Organ Failure Assessment (SOFA) scores were measured at D₂, D₅ and D₇. Blood volumes were measured with ⁵¹Cr fixed on red blood cells at D₂ and D₇.

Results

The ∆Na⁺ or ∆H₂O were increased in all patients but never related to blood volumes at D₂ nor D₇. Total blood volumes were at normal values with constantly low red blood cell volumes and normal or decreased plasmatic volume. Weight, plasmatic proteins, and hemoglobin were weakly related to ∆Na⁺ or ∆H₂O. Amongst all tested biomarkers, only MR-proADM was related to sodium and fluid overload. This biomarker was also a predictor of SOFA scores.

Conclusions

Plasmatic concentration in MR-proADM seems to be a good surrogate for evaluation of ∆Na⁺ or ∆H₂O and predicts sodium and extracellular fluid overload.

Trial registration

ClinicalTrials.gov: NCT01858675 in May 13, 2013.

Electronic supplementary material

The online version of this article (doi:10.1186/s13054-016-1540-x) contains supplementary material, which is available to authorized users.

Calculation of the Residual Blood Volume after Acute, Non-Ongoing Hemorrhage Using Serial Hematocrit Measurements and the Volume of Isotonic Fluid Infused: Theoretical Hypothesis Generating Study

Fluid resuscitation, hemostasis, and transfusion is essential in care of hemorrhagic shock. Although estimation of the residual blood volume is crucial, the standard measuring methods are impractical or unsafe. Vital signs, central venous or pulmonary artery pressures are inaccurate. We hypothesized that the residual blood volume for acute, non-ongoing hemorrhage was calculable using serial hematocrit measurements and the volume of isotonic solution infused. Blood volume is the sum of volumes of red blood cells and plasma. For acute, non-ongoing hemorrhage, red blood cell volume would not change. A certain portion of the isotonic fluid would increase plasma volume. Mathematically, we suggest that the residual blood volume after acute, non-ongoing hemorrhage might be calculated as 0·25N/[(Hct1/Hct2)-1], where Hct1 and Hct2 are the initial and subsequent hematocrits, respectively, and N is the volume of isotonic solution infused. In vivo validation and modification is needed before clinical application of this model.

Blood volume analysis by radioisotopic dilution techniques: state of the art

In the last years, there has been a growing recognition of the importance of blood volume abnormalities in the pathophysiology of several conditions and, consequently, a growing interest of accurate and rapid volume status assessment. Accordingly, there has been a surge of interest in blood volume analysis by radioisotopic dilution technique. However, there are still some controversies about this technique, such as the use of the f-cell ratio, the errors associated with the method and the reference values. This review aims to revise and discuss the theoretical and methodological aspects of this technique and also to discuss their controversies. Furthermore, it is questioned whether red cell volume or plasma volume can be accurately estimated once the other quantity has been measured or should red cell volume and plasma volume be directly measured. As a conclusion, blood volume analysis by radioisotopic dilution technique is still valid and very useful.

A Theoretical Mathematical Model to Estimate Blood Volume in Clinical Practice

Perioperative intravenous (IV) fluid management is controversial. Fluid therapy is guided by inaccurate algorithms and changes in the patient's vital signs that are nonspecific for changes to the patient's blood volume (BV). Anesthetic agents, patient comorbidities, and surgical techniques interact and further confound clinical assessment of volume status. Through adaptation of existing acute normovolemic hemodilution algorithms, it may be possible to predict patient's BV by measuring hematocrit (HcT) before and after hemodilution. Our proposed mathematical model requires the following four data points to estimate a patient's total BV: ideal BV, baseline HcT, a known fluid bolus (FB), and a second HcT following the FB. To test our method, we obtained 10 ideal and 10 actual subject BV data measures from 9 unique subjects derived from a commercially used Food and Drug Administration-approved, semi-automated, BV analyzer. With these data, we calculated the theoretical BV change following a FB. Using the four required data points, we predicted BVs (BVp) and compared our predictions with the actual BV (BVa) measures provided by the data set. The BVp calculated using our model highly correlated with the BVa provided by the BV analyzer data set (df = 8, r = .99). Our calculations suggest that, with accurate HcT measurement, this method shows promise for the identification of abnormal BV states such as hyper- and hypovolemia and may prove to be a reliable method for titrating IV fluid.

Brain-type natriuretic peptide and right ventricular end-diastolic volume index measurements are imprecise estimates of circulating blood volume in critically ill subjects

BACKGROUND

Surrogate indicators have often been used to estimate intravascular volume to guide fluid management. Brain-type natriuretic peptide (BNP) has been used as a noninvasive adjunct in the diagnosis of fluid overload and as a marker of response to therapy, especially in individuals with congestive heart failure. Similarly, right ventricular end-diastolic volume index (RVEDVI) measurements represent another parameter used to guide fluid resuscitation. The aim of this study was to evaluate whether BNP and RVEDVI are clinically valuable parameters that can distinguish among hypovolemia, euvolemia, and hypervolemia, as measured by blood volume (BV) analysis in critically ill surgical subjects.

METHODS

This observational study was part of a prospective, randomized controlled trial. Subjects with pulmonary artery catheters for the treatment of traumatic injuries, severe sepsis/septic shock, cardiovascular collapse, adult respiratory distress syndrome, and postsurgical care were studied. Circulating BV was measured by a radioisotope dilution technique using the BVA-100 Analyzer (Daxor Corporation, New York, NY) within the first 24 hours of acute resuscitation. BV results were reported as percent deviation from the patient's ideal BV based on height and percent deviation from optimum weight. Hypovolemia was defined as less than 0%, euvolemia was defined as 0% to +16%, and hypervolemia was defined as greater than +16% deviation from ideal BV. RVEDVI was measured by continuous cardiac output pulmonary artery catheters (Edwards Lifesciences, Irvine, CA). BNP and RVEDVI measurements obtained with BV analysis were evaluated with Fisher's exact test and regression analysis.

RESULTS

In 81 subjects, there was no difference in BV status between those with BNP of 500 pg/mL or greater and BNP of less than 500 pg/mL (p = 0.82) or in those with RVEDVI of 140 mL/m or greater and RVEDVI of less than 140 mL/m (p = 0.43). No linear relationship existed between BV and these parameters.

CONCLUSION

In critically ill surgical patients, BNP and RVEDVI were not associated with intravascular volume status, although they may be useful as indices that reflect increased cardiac preload.

LEVEL OF EVIDENCE

Diagnostic study, level III.

Blood volume analysis can distinguish true anemia from hemodilution in critically ill patients

BACKGROUND

Peripheral hematocrit (pHct) is traditionally used as a marker for blood loss. In critically ill patients who are fluid resuscitated, pHct may not adequately represent red blood cell volume (RBCV). We hypothesize that the use of pHct alone may overestimate anemia, potentially leading to unnecessary interventions.

METHODS

Patients admitted to the intensive care unit underwent blood volume analysis. Serial blood samples were collected after injection of I-albumin. Samples were then processed by the Blood Volume Analyzer-100. RBCV and total blood volume (TBV) were calculated using the directly measured plasma volume (PV) and pHct. A computed normalized hematocrit (nHct) adjusts pHct to the patient's ideal blood volume.

RESULTS

Thirty-six patients (21 men), aged 49.8 years ± 18.4 years, Acute Physiology And Chronic Health Evaluation II score 14.9 ± 8.1, and injury severity score 29.4 ± 12.4 had 84 blood volume analyses performed on 3 consecutive days. Using ratios of TBV compared with ideal TBV, patients were stratified into three separate groups: hypovolemic (16 of 84), normovolemic (23 of 84), and hypervolemic (45 of 84). Mean differences between pHct and nHct in each group were 4.5% ± 3.1% (p≤0.01), 0.0% ± 1.2% (p=0.85), and -6.5% ± 4.1% (p≤0.01), respectively. pHct, when compared with nHct, diagnosed anemia (Hct <30) nearly equal within the hypovolemic and normovolemic groups. However, pHct overdiagnosed anemia in 46.7% of hypervolemic patients.

CONCLUSION

Use of blood volume analysis in critically ill patients may help to distinguish true anemia from hemodilution, potentially preventing unnecessary interventions.

A prospective randomized trial using blood volume analysis in addition to pulmonary artery catheter, compared with pulmonary artery catheter alone, to guide shock resuscitation in critically ill surgical patients

Measurement of blood volume (BV) may guide fluid and red blood cell management in critically ill patients when capillary leak from shock and fluid resuscitation makes assessment of intravascular volume difficult. This is a prospective randomized trial of critically ill surgical patients with septic shock, severe sepsis, severe respiratory failure, and/or cardiovascular collapse. The control group received fluid management based on pulmonary artery catheter parameters and red blood cell transfusions based on hematocrit values. The BV group received fluid and red blood cell transfusions based on BV analyses in addition to pulmonary artery catheter parameters. Blood volume was measured using the radioisotope tracer technique with iodine 131-labeled albumin. This allowed direct measurement of plasma volume and calculation of the red blood cell volume. The control group was blinded to the BV results. There were statistically significantly more times when the control group (compared with the BV group) demonstrated hypervolemia (48% vs. 37%) and red blood cell deficiency (33% vs. 16%). There was a delay in red blood cell transfusions administered to the control group by 1.5 +/- 2 days at which time the abnormality became clinically evident. Blood volume analyses provided additional information to the clinicians resulting in a change in treatment in 44% of the time to patients randomized to the BV group. The mortality rates were significantly different between the two groups (8% for the BV group and 24% in the control group; P = 0.03). Blood volume measurements allowed the physicians to promptly treat physiologic disturbances in both red blood cell volume and plasma volume, resulting in improved survival.