While Patient Blood Management (PBM) is becoming a new standard worldwide with evidence-based and expert consensus guidelines to promote the three PBM pillars, measuring the performance of these efforts remains a challenge. PBM programmes need to measure transfusion reductions and patient-specific outcomes, and report these to drive change.
Here is one such impactful report supporting the feasibility and associated transfusion reductions of a PBM programme. This is an interesting retrospective quality improvement report from Dr. Spahn and co-authors of the University Hospital of Zurich who investigated the impact of their PBM programme on blood transfusion practices, costs and sustainability. Blood transfusion practices for each department were monitored and feedback was provided on a regular basis regarding adherence to set transfusion thresholds based on a Hb 12.7 or Factor V activity <20%.
The authors included over 200 000 patients; about 60 000 from the pre-PBM programme year (2012-13) as a baseline, 44 000 from year one of the PBM programme (2014) and over 110 000 from the time period 2015-17. The feasibility, sustainability and associated decreased allogeneic blood transfusion (by 35%) with a corresponding reduction in costs is reported.
The fact that the authors did not find a change in in-hospital mortality and did not include an analysis of any adverse outcomes are limitations in part due to the retrospective nature of the report. Prospective well-designed multicentre studies with patient-centred outcome measures are needed to drive this message home – PBM programmes improve care, decrease complications and decrease costs.
The authors analysed data from a retrospective cohort of transfused patients from the Recipient Epidemiology and Donor Evaluation Study-III (REDS-III) using stratified Cox regression models to estimate associations between blood donor characteristics (age, body mass index, haemoglobin, smoking status), hospital mortality and post-transfusion length of stay (LOS).
Statistical analyses were adjusted for recipient factors, including total number of transfusions. A total of 93 726 patients receiving 428 461 RBC units were analysed. There were no associations between blood donor characteristics and hospital mortality. Receipt of RBC units from donors <20 years of age was associated with a shorter hospital LOS (P < 0.001) whereas no assocition was observed between LOS other donor characteristics. No associations between blood donor factors and in-hospital mortality was found. The shorter hospital LOS in patients transfused RBCs from younger donors requires confirmation.
Published: Nov 2019
Harenberg J, Beyer-Westendorf J, Crowther M, et al.
Harenberg et al. have developped a rapid colorimetric point-of-care test using patient urine sample that is able to detect direct oral anticoagulants. In this multicentre study (880 evaluable patients), factor Xa inhibitors (apixaban, edoxaban, rivaroxaban) and the thrombin inhibitor dabigatran were detected with a high sensitivity, specificity and accuracy (>95% correct positive and negative predictive results).
This test may be of great help in daily clinical practice, as demonstrated by the authors. In a patient presenting with a major haemorrhage, a negative DOAC Dipstick test result indicates that DOACs are highly unlikely to contribute to the bleeding event. If the DOAC Dipstick test result is positive, a specific antidote can be considered immediately or after plasma DOAC levels have been determined.
If a patient treated with a DOAC requires an urgent major operative procedure, a negative DOAC Dipstick test shows that it is unlikely that the patient has significant DOAC concentrations in blood, and the operative procedure may be performed as soon as indicated. If the DOAC Dipstick test is positive, the operation may need to be postponed or additional blood tests should be performed to guide the decision process.
Indeed, the test does not enable precise quantification of DOAC levels; the rating scale for both for the factor Xa and thrombin inhibitor pads contains the following possibilities: negative, positive +, and positive ++ . Nevertheless, this test is really promising and the study meets international standards. For sure, it will be integrated into laboratory and clinical decision-making processes and will improve patient care.
The authors sought to determine whether transfusion of platelets or fresh frozen plasma, in addition to RBCs, is associated with an increased risk of mortality and infection after cardiac surgery. Patients who underwent valve surgery and/or coronary artery bypass grafting at 2 centres performing cardiac surgery were retrospectively studied.
After stratifying patients based on propensity score matching, rates of mortality and infection between patients who received RBCs, FFP or platelets were compared with subjects not receiving these blood products. Outcomes were compared between patients who received any of the 3 blood products and patients who received no transfusions at all. Multivariable logistic regression was used to assess associations between transfusion and outcomes.
Of 8238 patients,109 (1.3%) died, 812 (9.9%) experienced infection, and 4937 (59.9%) received at least one type of blood product. Transfusion of any blood type was associated with higher rates of mortality (2.0% vs. 0.18%; P < 0.01) and infection (13.3% vs. 4.8%; P < 0.01). Each of the 3 blood products was independently associated with an increase in mortality per unit transfused (RBC, OR 1.18, 95% [CI], 1.14-1.22; FFP, OR 1.24, 95% CI, 1.18-1.30; platelets, OR 1.12, 95% CI, 0.07-1.18).
Transfusing 3 units of any of the 3 blood products was associated with a dose-dependent increase in the incidence of mortality (OR 1.88, 95% CI, 1.70-2.08) and infection (OR 1.50, 95% CI, 1.43-1.57). Transfusion of RBCs, FFP, or platelets is an independent risk factor of mortality and infection, and combination of the 3 blood products is associated with adverse outcomes in a dose-dependent manner after cardiac surgery.
The World Health Organization recommends that pregnant women should take supplements containing 30-60 mg elemental iron (depending on anaemia prevalence in the area) and 400 μg folic acid daily to prevent maternal anaemia, puerperal sepsis, low birthweight and preterm birth. This supplementation regimen was also recommended in recent NATA guidelines for the management of anaemia and haematinic deficiencies in pregnancy and in the post-partum period (Muñoz M et al., Tranfus Med 2018;28:22-39).
In this double-blind randomised controlled trial, the authors tested two hepcidin-guided screen-and-treat approaches (hepcidin was tested every week and women received 30 mg/day or 60 mg/day if hepcidin 2.5 µg/L, until the next hepcidin measurement) against the standard-of-care 60 mg per day regimen. All women received folic acid 400 µg/day during the study period (12 weeks). Hepcidin-guided iron supplementation offered no advantages over the standard WHO recommendations in terms of efficacy (haemoglobin, ferritin) and safety (gastrointestinal side effects, infections) but carried a considerable increment in health-care costs.
Oral iron-related side effects are frequently reported by pregnant women, which may lead to poor adherence and undermine antenatal iron supplementation programmes. The unusually high adherence (86%) observed in this study may reflect the influence of sensitization and fieldworker encouragement and the fact that participants were aware that adherence was being monitored. As expected, the prevalence of illnesses and side effects was lowest in the 30 mg screen-and-treat group. This lower daily dose iron supplementation has been recommended for low anaemia settings (WHO 2019). Alternate day iron supplementation (60 mg dose) may increase adherence (Stoffel NU et al., Lancet Haematol 2017;4:e524-e533), and is also supported by NATA guidelines (Muñoz M et al., Tranfus Med 2018;28:22-39).
The use of parenteral iron formulations which can deliver ≥1000 mg elemental iron in a single seating could be an alternative for iron supplementation in pregnancy (starting in the second trimester), but it will need evidence of cost-effectiveness and safety in low-income settings, together with development of infrastructure, to overcome barriers to implementation. In addition, the European Medicines Agency (2013) considered that, during pregnancy, allergic reactions are of particular concern as they can put both the mother and unborn child at risk, and therefore intravenous iron should not be used during pregnancy unless clearly necessary.
The available evidence suggests that efforts should be directed towards developing newer oral iron supplements with better side-effect profiles. In this regard, the efficacy, tolerability and safety of sucrosomial iron administration for preventing iron deficiency and anaemia in pregnancy (14 mg/day or 28 mg/day, from gestation week 12-14 until postpartum week 6) has been reported recently (Parisi F et al., J Matern Fetal Neonatal Med 2017;30:1787-1792).
Tranexamic acid administered within 3 hours of injury in patients with head trauma reduces death.
I know there has been criticism of the study design, but the results of this trial should change practice as it a low-cost intervention with no safety issues.
Chaudhury et al. retrospectively analysed the outcomes of 769 consecutive adult patients with pulmonary embolism (PE) (as diagnosed by computed tomography PE protocol) for the 18 months prior to and after initiation of a multidisciplinary Pulmonary Embolism Response Team (PERT) in their large tertiary care centre.
PERT-era patients had significantly lower rates of bleeding, shorter time-to-therapeutic anticoagulation, and decreased use of inferior vena cava filters. Most importantly, there was a significant decrease in 30-day/inpatient mortality, which was more pronounced in intermediate- and high-risk patients (mortality 10.0 vs. 5.3%; P = 0.02).
The authors provide evidence that availability of multidisciplinary PERT is associated with improved outcomes including 30-day mortality, with higher severity patients appearing to derive the most benefit. In addition, there was no evidence of overuse of invasive techniques or harm demonstrated in the PERT-era patients.
Bleeding after cardiac surgery is a common and serious complication leading to transfusion of multiple blood products, resulting in increased morbidity and mortality. Transfusion rates in cardiac surgery have declined only modestly, remaining at 50% or greater in high-risk patients.
The Society of Cardiovascular Anesthesiologists has formed the Blood Conservation in Cardiac Surgery Working Group to organize, summarize, and disseminate the available best-practice knowledge in patient blood management in cardiac surgery. This work includes summary statements and algorithms designed by the working group and recommendations for patient blood management practices.
Orthotopic liver transplants have come a long way over the past several decades. While we have solved a myriad of problems leading to significantly improved outcomes, bleeding remains a problem.
This randomised controlled trial tested a thromboelastometry-based algorithm against standard practices. Intraoperative transfusion requirements were reduced in the thromboelastometry group, specifically FFP transfusions.
Orthotopic liver transplants are life-saving procedures associated with significant risk of bleeding and clotting. This study highlights a method where these risks may be further reduced.
Paediatric patients undergoing cardiopulmonary bypass for cardiac surgery have historically had their bypass circuits primed with fresh frozen plasma (FFP). The intention has been to decrease perioperative bleeding and subsequent transfusions. Studies looking at both plasma-primed circuits and crystalloid-primed circuits have yielded conflicting results.
This is the first double-blind randomised controlled trial looking at bleeding risk in patients whose circuits were primed with crystalloid versus FFP. The study found that there was no difference in postoperative bleeding or the number of transfused blood products between the FFP group and the crystalloid group.
This data suggests that for this specific patient population the cardiopulmonary bypass circuit can be primed with crystalloid, thus avoiding exposure to FFP.
This study aimed to evaluate the association of transfusion timing (intraoperative versus postoperative) on the rate of postoperative myocardial infarction and death after common inpatient vascular surgery procedures. The authors wanted to evaluate this association because, although perioperative transfusions have been associated with increased mortality, in some cases (significant blood loss) they can be life-saving and the timing of the transfusion may be a marker of the necessity for a life-saving transfusion.
The study was a retrospective review of data collected by the Society of Vascular Surgery. There were 1154 total operations included in the analysis, including the following procedures: endovascular aneurysm repair, infra-inguinal bypasses, open abdominal aortic aneurysms, supra-inguinal bypasses, thoracic endovascular aortic repair, and inpatient peripheral vascular interventions. The primary outcome of the study was postoperative myocardial infarction and the secondary outcome was 30-day all-cause mortality.
The authors found that receiving both intraoperative and postoperative transfusion or postoperative transfusion only compared to no transfusion was associated with higher odds of postoperative myocardial infarction; however, intraoperative transfusion only was not. They also found that any intraoperative or postoperative transfusion compared to no transfusion was not associated with mortality.
The authors conclude that the perioperative setting of transfusion is important in its impact on postoperative outcomes and needs to be accounted for when evaluating transfusion outcomes and indications.
Hypofibrinogenaemia is frequently observed in bleeding patients after cardiac surgery with cardiopulmonary bypass. In case of excessive bleeding and acquired hypofibrinogenaemia, the most recent guidelines recommend the administration of either cryoprecipitate or fibrinogen concentrate. Although cryoprecipitate remains the treatment of choice is some countries (e.g. Canada, United States), cryoprecipitate is not longer available in several other countries. So far, very few studies have directly compared those two options in bleeding patients undergoing cardiac surgery.
The FIBRES randomised clinical trial is a large multicentre study in patients undergoing cardiac surgery requiring fibrinogen replacement because of significant bleeding and hypofibrinogenaemia (defined as a plasma fibrinogen <2.0 g/L by the Clauss method or FIBTEM measured at 10 min <10 mm).
Of 827 randomised patients, 372 received fibrinogen concentrate and 363 received cryoprecipitate. Mean 24-hour post-bypass allogeneic transfusions were 16.3 units in the fibrinogen group and 17.0 units in the cryoprecipitate group. Thromboembolic events occurred in 26 patients (7%) in the fibrinogen concentrate group and 35 patients (9.6%) in the cryoprecipitate group.
The results of the study are important for several reasons. Firstly, this large randomised controlled study demonstrated that fibrinogen concentrate can be used as an effective alternative to cryoprecipitate in the presence of bleeding and hypofibrinogenaemia in adults undergoing cardiac surgery with cardiopulmonary bypass.
Secondly, as mentioned in the discussion by the authors, cryoprecipitate is a plasma-derived product that presents some risk of pathogen transmission potentially associated with poor outcomes and increased cost. Although the study did not report any differences in clinical outcome, one should remember that the incidence of those events is extremely low and that an unreasonably large number of patients would be needed to report any significant difference.
Last but not least, it is important to keep in mind that the administration of cryoprecipitate does not only supplement with an inconsistent concentration of fibrinogen, but also with other factors that might not be deficient or might increase the risk of thrombotic complications. In this regard, it may be argued that fibrinogen concentrate offers a more predictable and safer profile than cryoprecipitate.
In summary, the results of FIBRES study demonstrate that fibrinogen concentrate is an effective and safe alternative to cryoprecipitate in patients experiencing excessive bleeding in the presence of hypofibrinogenaemia in the perioperative period of cardiac surgery.
This very interesting prospective study looked at four patient cohorts undergoing cardiac surgery: patients without aneemia with and without iron deficiency and patient with anaemia with and without iron deficiency. The primary outcomes were 90-day mortality, serious adverse events, major cardiac and cerebrovascular events, transfusions, and prolonged hospital stay.
The study found that 90-day mortality was increased in iron-deficient patients with and without anaemia. Patient with iron deficiency also had more serious adverse events and prolonged hospital stays.
This study suggests that patients with iron deficiency, even in the absence of anaemia, should be treated prior to undergoing cardiac surgery.
Published: Oct 2019
Miles LF, Larsen T, Bailey MJ, Burbury KL, Story DA, Bellomo R
This study aimed to test whether there were differences in outcomes of women undergoing major abdominal surgery with a preoperative Hb <13 g/dL compared to women with a Hb <12 g/dL. Compared to historical definitions of anaemia, there are now updated age-, gender-, and race-adjusted definitions, but the clinical significance of the differences in these definitions has not been well studied. The authors hypothesized that although a Hb 12-12.9 g/dL may be normal, this Hb level may not be physiologically optimal for women undergoing major abdominal surgery.
The study was a retrospective cohort study, and all women undergoing elective surgery at a single institution were included. The authors reviewed all surgeries and defined them as either “major” or “complex major” surgery. Participants were stratified according to their preoperative Hb: <12.0 g/dL, 12.0-12.9 g/dL, and ≥13.0 g/dL. The primary outcome measure was hospital length of stay.
The authors found a negative relationship between logarithmic preoperative Hb and length of stay for an Hb concentration <13.0 g/dL vs. >13.0 g/dL (P = 0.03); with a difference in length of stay approximately 50% greater for women with an Hb of 12.0 g/dL compared with those with an Hb of 13.0 g/dL.
The authors conclude that borderline anaemia increases patients’ length of stay, but the independent contribution of low Hb to perioperative outcome needs further study.
What is the aim of transfusion? Increasing the haemoglobin concentration, increasing oxygen delivery, increasing tissue oxygenation, increasing oxygen consumption, or improving survival? This rather provocative question illustrates one of our problems regarding transfusion. All of our efforts should focus on rock-solid outcome parameters (e.g. survival), however in daily clinical practice we very often look only at the most easily accessible parameter: haemoglobin concentration.
In this study by Aktas and coworkers, another approach tested in preterm infants. The authors investigated the impact of the transfusion of red blood cells on tissue oxygenation in the brain, the abdomen and the kidneys by means of near-infrared spectroscopy.
As many others previously, they were able to demonstrate – in this very specific patient group – that the effect of RBC transfusion on tissue oxygenation is less pronounced than anticipated. Although the overall effects were quite small, they could distinguish between the effects on the brain and the abdomen/kidneys. Tissue oxygenation slightly improved for the abdomen and the kidney, but only small changes were observed for cerebral rSO2. Whether this represents a monitoring artifact or can be explained by changes in regional blood flow remains open for discussion. Furthermore, the authors attempted to distinguish the effects for different degrees of pretransfusion anaemia; however, the differences were more or less negligible.
Although this study has several shortcomings, it clearly demonstrates that the mantra “tissue oxygenation improvement by transfusion” does not necessarily hold true, and therefore the effects typically expected from transfusion cannot always be reached.
Inflamm Bowel Dis 2019; Sep 27 [Epub ahead of print].
REVIEW by S. M. Goobie:
This is a large database report of the incidence and duration of anaemia in the inflammatory bowel disease (IBD) population over the course of their disease lifetime. The authors conducted a retrospective nationwide cohort study among the US veteran population from 2011 to 2018. The primary outcomes were the incidence of anaemia, the number of any anaemia days per year, and the number of moderate or severe anaemia days per year in the IBD population compared to the non-IBD group.
More than half of the patients with IBD (55%) developed anaemia after diagnosis of the disease, with 25% developing moderate-to-severe anaemia. The incidence rate of anaemia was 85% higher in the IBD cohort (>3000) than the matched non-IBD (>5000) patients. This is almost twice the number of days per year in an anaemic state when compared with non-IBD patients.
Despite experiencing a higher incidence of anaemia (adjusted risk ratio 1.85) and spending a longer time anaemic, only 40% of anaemic IBD patients received iron replacement therapy. The authors conclude that more efforts should be made to recognise and treat anaemia among patients with IBD.
This is a retrospective study of patients with inherited platelet disorders undergoing surgical procedures. No protocol was followed, and mechanical and pharmacological thromboprophylaxis was used at clinical discretion.
There were patients with clinical VTE. There were patients who bled. Disappointingly, only a minority got perioperative tranexamic acid. It emphasises the need for RCTs in this area.
Published: Sep 2019
Maeda Y, Ogawa K, Morisaki N, Tachibana Y, Horikawa R, Sago H
Int J Gynaecol Obstet 2019; Sep 26 [Epub ahead of print].
REVIEW by S. M. Goobie:
Is there a relationship between anaemia and depression in the postpartum period? Certainly perioperative anaemia has been independently associated with increased morbidity and mortality across many different specialities in surgical and critical care. It has been linked to prolonged hospital length of stay, excessive health resource utilization and reduced disease-free survival.
Maternal iron deficiency anaemia has been associated with poor maternal and infant outcomes in observational studies and randomised trials. This report by Maeda and colleagues measures a very important quality of life outcome: postpartum depression.
The authors conducted a prospective cohort study of over 1100 pregnant women. The primary outcome was postpartum depression assessed using the Edinburgh Postpartum Depression Scale (EPDS) at postpartum week 4. Postpartum anaemia was significantly associated with increased postpartum depression risk (adjusted odds ratio 1.63, 95% confidence interval 1.17-2.26), the risk increasing with lower haemoglobin levels.
Therefore, according to this report (and others, e.g. Azami M et al. Caspian J Intern Med 2019;10:115-24; Wassef A et al. J Psychosom Obstet Gynaecol 2019;40:19-28), the association between anaemia and postpartum depression is a modifiable variable that should not be ignored.
This is the first ever bleeding assessment scale developed for the evaluation of bleeding in critically ill children.
The BASIC tool was developed by twenty-four clinical content experts using a modified Delphi method. The tool stratifies clinically significant bleeding into three categories: minimal bleeding, moderate bleeding, and severe bleeding.
The authors anticipate that, once validated, this scale will serve to standardise communication and provide a valuable tool for clinical research.
Jackman and colleagues evaluated transfusion-associated microchimerism (TA-MC) in a prospective cohort study of trauma patients. Index samples were collected upon admission, prior to transfusion and at periodical intervals up to one year. TA-MC was detected by real-time quantitative allele-specific PCR assays at the HLA-DR locus and at several polymorphic insertion deletion sites screening for non-recipient alleles.
A total of 378 adult trauma patients (324 transfused and 54 non-transfused) were studied (80% suffered blunt trauma). The median number of red cell units transfused was 6 units. The investigators identified only one case of long-term (>180 days) TA-MC in transfused subjects, and short-term TA-MC in 6.5% of transfused and 5.6% of control patients.
Interestingly, in contrast to previous studies, persistent TA-MC was not observed in this cohort of trauma subjects, and short-term TA-MC was detected at a lower frequency. The reduction in TA-MC occurrence may be attributable to changes in leukoreduction or other blood processing methods, which are highly variable across centres.
Published: Sep 2019
de Bruin S, Scheeren TWL, Bakker J, van Bruggen R, Vlaar APJ
In order to elucidate whether data from transfusion trials has impacted current intensive care transfusion management, de Bruin and colleagues performed an online, anonymous, worldwide survey among ICU physicians, evaluating red blood cell, platelet and plasma transfusion practices.
Of 725 completed surveys, a hospital transfusion protocol was available in 53%, with only 29% using ICU-specific transfusion guidelines. The reported haemoglobin threshold for the general ICU population was 7 g/dL. Higher transfusion thresholds were cited for ECMO and brain injured patients (8 g/dL [7.0–9.0]). Platelets were transfused at 20 x 10(9) cells/L (IQR 10-25) in asymptomatic patients, and at a higher count prior to invasive procedures. Plasma transfusion practice was highly variable. Anaesthesiologists transfused more liberally than internal medicine physicians.
The author’s work would indicate that red cell transfusion practice for the general ICU population is restrictive, with higher levels targeted in those with brain injury and on ECMO. Plasma and platelet transfusion practice varies.
Roquet et al. report the results of a large retrospective trauma registry study on optimal FFP to RBC transfusion ratios. As stated by the authors themselves, these retrospective studies suffer from survivor bias (patients that do not survive the initial resuscitation just have no time to receive plasma and end up with a low FFP to RBC ratio), which has been identified previously.
Correction of this bias mitigates the previously reported positive effects of high FFP to RBC ratio in trauma care. Hence, this study has limited value and the only randomised trial in this field was negative for the primary endpoint.
This is an excellent multicentre prospective randomised trial evaluating the efficacy of fibrinogen concentrate (FC) compared to cryoprecipitate to achieve haemostasis for infants post-cardiopulmonary bypass.
The primary outcome was the difference in the number of intraoperative blood product transfusions. Secondary outcomes included 24-hour chest tube output, mechanical ventilation time, adverse events, intensive care unit length of stay, hospital length of stay, postoperative thrombosis, and death within 30 days of surgery.
The FC group actually received fewer blood products than the cryoprecipitate group. There was no difference in any of the secondary outcomes. This study supports the use of FC to achieve haemostasis in infants undergoing cardiopulmonary bypass.
Due to population ageing, hip fracture is becoming a major health concern. Indeed, short-term postoperative mortality is high (3-5%) and postoperative delirium is a common complication after hip fracture surgery, and their incidence is increased in case of preoperative anaemia. However, the detection of perioperative anaemia may be delayed.
This study evaluated the impact of noninvasive continuous Hb monitoring using the SpHb (Masimo Corp., USA) for the detection of severe perioperative anaemia (i.e. Hb <10 g/dL) in hip fracture patients. This was a 6-month prospective observational study conducted in a University hospital in Denmark. The authors analysed 42 patients (aged 78 ± 8 years, 42.5% male, 65% ASA I/II) among 51 enrolled ones (inclusion criteria: age ≥65 years and confirmed hip fracture).
The majority of patients were operated under epidural anaesthesia (started preoperatively for analgesia). All patients were monitored using a Radical-7 pulse oximeter and Rainbow sensors (Masimo Corp., USA) that give SpHb (i.e. non-invasive measurement of Hb) from 12 hours before surgery until 24 hours after. The monitor was blinded to all caregivers. A blood count was obtained daily for the first 3 postoperative days.
Twenty-two (52%) patients had at least one lab Hb value below 10 g/dL (the predefined transfusion trigger). This low Hb was detected in 15 (75%) patients by the SpHb (3 patients had not enough SpHb data and 4 had an SpHb >10 g/dL). The delay for detection of a low Hb was only 1.07 ± 2.84 hours earlier using SpHb. Interestingly, patients with postoperative delirium had a more prolonged cumulated time with low SpHb (162 [30-819] vs. 22 [0-70] minutes; P = 0.034).
Unfortunately, the investigators were not able to conduct the study as initially planned – with a monitoring period beginning at admission and lasting until the third postoperative day – because of too many missing data. The “improved” delay they reported does not appear to be clinically significant.
This study indicates that SpHb should probably not be used to replace invasive blood count surveillance, but it could provide some helpful additional information, for example to request a blood sample in case of decreasing SpHb. In addition, the investigators did not take into account the information given by the PVI (pulse variability index), which could help distinguish between haemodilution (low SpHb, low PVI) and bleeding (low SpHb, high PVI).
The authors noted that large academic centres are very likely to have a massive haemorrhage protocol (MHP) but smaller non-academic hospitals may not. Using a Modified Delphi Method, they convened a group of experts to agree on evidence-based recommendations to be included in a standardised regional MHP.
The group found consensus around 42 statements and 8 quality indicators. The resulting “toolkit” will be used to establish MHPs in all of the regional hospitals in Ontario, Canada.
This paper demonstrates an ideal approach to creating standardised care around MHPs across multiple hospitals. The recommendations put forward can be used as a framework for any institution developing a MHP.
This study by Padmanabhan et al. is a well-conducted retrospective analysis that confirms that anaemia before cardiac surgery is strongly associated with long-term mortality.
The impact of preoperative anaemia seems to be stronger than and outweigh the risk of RBC transfusion, especially with low volumes (1-2 units) of RBCs that were not associated with a higher risk of mortality. The importance of preoperative anaemia is underscored by the interaction analysis of preoperative anaemia and transfusion, which showed no association.
However, these results need to be interpreted with great caution before it can be concluded that RBC transfusions are safe. There are too many data out showing different results (strong impact on adverse reactions) for RBC transfusions. While confirming preoperative anaemia as a risk factor in cardiac surgery, this study fuels the debate on the impact of RBC transfusions.
Finally, a large and well-designed prospective randomised study comparing causal anaemia treatment to RBC transfusions in anaemic cardiac surgery patients is needed to answer this question.
August 2019
Published: Aug 2019
Savarese G, Jonsson Å, Hallberg AC, Dahlström U, Edner M, Lund LH
This study aims to establish a reliable prevalence estimate of anaemia in patients with different severities of heart failure (HF), and to identify whether the severity of HF and anaemia predicts either mortality or hospitalization from HF. While data from previous studies has established an association between anaemia and HF, recent European updates to HF classification prompted this study and the question of whether there may be a differential effect of anemia based on HF severity.
A total of 42 895 patients from a large registry of HF patients in Sweden were included. Patients were stratified according to the ejection fraction, with 23% having HFrEF (EF <40%), 21% with HFmrEF (EF 40-49%) and 55% with HFpEF (EF ≥50%).
There was a higher prevalence of anaemia in the HFpEF group (41%) than the HFmrEF group (35%) or the HFrEF group (32%). Additionally, while anaemia was independently associated with mortality and increased hospitalization across all HF groups, there was a higher risk of mortality or hospitalization in the HFpEF (OR 1.24, P < 0.001) and HFmrEF (OR 1.26, P < 0.001) groups compared to patients with HFrEF (OR 1.14, P < 0.001).
While the authors are careful to note the limitations of this observational study, their sample size is a real strength and this study should prompt awareness of anaemia and its adverse consequences in HF patients that have been generally considered less sick.
Acta Obstet Gynecol Scand 2019; Aug 26 [Epub ahead of print].
REVIEW by Susan M. Goobie:
This is a very well-designed retrospective study using a large American surgical database (NISQP) reporting the a high incidence of anaemia (19.5%) in patients undergoing elective laparoscopic hysterectomy. The following factors were associated with a high risk of preoperative anaemia: higher body mass index, younger age, Black or African American race, longer operative times, and multiple other medical comorbidities.
The take-home message is that preoperative anaemia is independently associated with morbidity, extended length of stay, readmission and composite increased morbidity after surgery in these women presenting for elective laparoscopic hysterectomy. This important report adds to the wealth of data on the negative impact of preoperative anaemia on patient mortality and morbidity in our surgical patients.
Published: Aug 2019
Porter SB, White LJ, Osagiede O, Robards CB, Spaulding AC
J Arthroplasty 2019; Aug 17 [Epub ahead of print].
REVIEW by Micah T. Prochaska:
While Patient Blood Management (PBM) is becoming a new standard worldwide with evidence-based and expert consensus guidelines to promote the three PBM pillars, measuring the performance of these efforts remains a challenge. PBM programmes need to measure transfusion reductions and patient-specific outcomes, and report these to drive change.
Here is one such impactful report supporting the feasibility and associated transfusion reductions of a PBM programme. This is an interesting retrospective quality improvement report from Dr. Spahn and co-authors of the University Hospital of Zurich who investigated the impact of their PBM programme on blood transfusion practices, costs and sustainability. Blood transfusion practices for each department were monitored and feedback was provided on a regular basis regarding adherence to set transfusion thresholds based on a Hb <9 g/dL, platelet count <100 G/L and PT >12.7 or Factor V activity <20%.
The authors included over 200 000 patients; about 60 000 from the pre-PBM programme year (2012-13) as a baseline, 44 000 from year one of the PBM programme (2014) and over 110 000 from the time period 2015-17. The feasibility, sustainability and associated decreased allogeneic blood transfusion (by 35%) with a corresponding reduction in costs is reported.
The fact that the authors did not find a change in in-hospital mortality and did not include an analysis of any adverse outcomes are limitations in part due to the retrospective nature of the report. Prospective well-designed multicentre studies with patient-centred outcome measures are needed to drive this message home – PBM programmes improve care, decrease complications and decrease costs.
JAMA Intern Med 2019; Aug 5 [Epub ahead of print].
REVIEW by Charles-Marc Samama (NATA):
The Perioperative Anticoagulation Use for Surgery Evaluation (PAUSE) cohort study enrolled 3007 atrial fibrillation patients treated with apixaban, dabigatran etexilate or rivaroxaban and scheduled for an elective surgery or procedure. A simple standardised perioperative DOAC therapy interruption and resumption strategy based on DOAC pharmacokinetic properties, procedure-associated bleeding risk and creatinine clearance levels was implemented.
The DOAC regimens were omitted for 1 day before a low-bleeding-risk procedure and 2 days before a high-bleeding-risk procedure (4 days for dabigatran in patients with a creatinine clearance <50 mL/min). The DOAC regimens were resumed 1 day after a low-bleeding-risk procedure and 2 to 3 days after a high-bleeding-risk procedure.
The 30-day postoperative rate of major bleeding was 1.35% in the apixaban cohort, 0.90% in the dabigatran cohort and 1.85% in the rivaroxaban cohort. The rate of arterial thromboembolism was 0.16% in the apixaban cohort, 0.60% in the dabigatran cohort and 0.37% in the rivaroxaban cohort. In patients with a high-bleeding-risk procedure, the rates of major bleeding were 2.96% in the apixaban cohort and 2.95% in the rivaroxaban cohort, which is quite significant. In addition, a high proportion of patients (>90% overall; 98.8% of those at high bleeding risk) had a minimal or no residual anticoagulant level at the time of the procedure.
This large prospective cohort validates the most recent international recommendations for the interruption and resumption of DOAC therapy in patients undergoing elective surgery or procedures. As already suggested in current recommendations, no bridging was necessary and coagulation tests were not used to modify the strategy. However, these results can only be discussed in light of the unbalanced proportion of low-bleeding-risk procedures (roughly two thirds) versus high-bleeding-risk procedures (one third).
The authors conducted a retrospective study to evaluate outcomes of severely injured trauma patients who received 4-factor prothrombin complex concentrate (4-PCC) + fresh frozen plasma (FFP) compared to FPP alone. For this, they analysed two years (2015–2016) of the American College of Surgeons-Trauma Quality Improvement Program database. A large number of adult, seriously traumatised patients (86% with blunt trauma) were matched (234 in each group).
Four-PCC + FFP were associated with a decreased requirement for packed red blood cell and FFP transfusion compared to FFP alone. Platelet transfusion rates were similar in both groups. Patients who received 4-PCC + FFP had a lower mortality (17.5% vs. 27.7%; P = 0.01), lower rates of ARDS and AKI and a shorter hospital LOS. There was no difference in the rates of DVT and PE (both objectively demonstrated) between the two groups.
This interesting, albeit retrospective, study suggests that the addition of 4-PCC to FFP improves outcomes (mortality and severe adverse events) in patients with trauma-induced coagulopathy. Thrombotic complications were not increased by the addition of 4-PCC to FFP. Further randomised controlled studies are required to evaluate the efficacy and safety of the addition of PCC to the massive transfusion protocol.
This is a retrospective report of 10 years of obstetrical cell salvage use in a single hospital. The authors report on 1170 patients who had a median of 231 mL of blood reinfused. This article contributes to the growing amount of information on the safety of cell salvage use in obstetrics.
Over the 10 years of their study, they observed a reduction in allogeneic blood use from 1.4% in 2008 to 0.4% in 2017. During the same period, they saw a steady increase in the amount of autologous blood re-infused. It is difficult to directly correlate the increasing use of cell salvage with the decreases in use of allogeneic blood because they also implemented an intravenous iron programme and started using tranexamic acid during the same study period. What can be concluded is that the three strategies together resulted in a significant reduction in allogeneic blood exposure.
In this patient population of young women, the long-term consequences of allogeneic transfusion are poorly understood. The best way of eliminating this unknown risk is through a comprehensive patient blood management programme as is illustrated by this article.
Published: Aug 2019
Namara H, Kenyon C, Smith R, Mallaiah S, Barclay P
This important study by McNamara et al. reports on four years of observational data from a large UK hospital and tertiary referral unit, following the introduction of a rotational thromboelastometry-guided algorithm for treatment of coagulopathy in major obstetric haemorrhage. This was compared to a matched cohort from the year prior when a “shock pack” was dispatched in a massive bleeding scenario.
Their work demonstrates the magnitude of “real-world” outcome improvements through the implementation of a thromboelastometry-guided algorithm for treatment of coagulopathy in obstetric haemorrhage. In the algorithm group, there was a significant reduction in the number of units (P < 0.0001) and total volume (P = 0.0007) of blood products transfused (in particular fresh frozen plasma), with a reduction in transfusion-associated circulatory overload (P = 0.002).
This study highlights the importance of fibrinogen as a marker/predictor of obstetric haemorrhage and in addition the usefulness and value of fibrinogen concentrate as treatment option. Compared with cryoprecipitate, availability is immediate, dosing is consistent, safety enhanced, and wastage minimized.
The analysis of the thrombelastometry results underpins the “individuality” of the obstetric haemorrhage and the subsequent need for a personalised treatment approach, also outlined in an excellent accompanying editorial by Shah and Collis (Anaesthesia 2019;74: 961-4).
For some time, there has been a lack of support for viscoelastic haemostatic assays (VHA) outside of cardiac surgery. Evidence is now growing that VHA is beneficial in obstetric haemorrhage. Therefore, a recommendation was made in the latest British Society for Haematology Guideline (Curry NS et al. Br J Haematol 2018;182:789-806) and an increasing number of hospitals (like my own) have added this diagnostic tool successfully to their PBM modalities.
The work by McNamara et al. will reassure clinicians who already have the ability of VHA and may reflect their own experiences with obstetric haemorrhage. Health care settings without VHA should be encouraged to work towards the implementation of this test in the future.
Patient blood management (PBM) programmes are recommended, notably to save blood products, but their impact on clinical outcomes has not been widely evaluated. This single-centre before-and-after study conducted at Johns Hopkins Hospital (USA) reports a significant 23% lower odds of being transfused (OR = 0.77, 95% CI 0.657-0.896, P = 0.001) in the “post-PBM” period. This PBM programme was based primarily on implementation of restrictive transfusion guidelines using education and monthly feedback through dashboards (with different indicators such as transfusion triggers, transfusion rates, etc.).
Although RBC transfusion exposure was associated with a higher postoperative mortality and morbidity (including longer LOS for the index admission), mortality and morbidity were not reduced in the post-PBM period. This highlights the importance of taking into account the different dimensions of PBM (i.e. the 3 pillars). In addition to restrictive triggers, it is also very important to optimise the patient’s red blood cell mass and to minimise blood losses (both pre- and postoperatively). When these different aspects are taken into account, even mortality may be reduced following implementation of PBM as shown by Leahy et al. (Transfusion 2017;57:1347-1358).
This study shows that implementing restrictive transfusion strategies is feasible and that feedback using “dashboards” is helpful. This allows for a reduction in blood product use.
Pediatr Crit Care Med 2019; Jun 27 [Epub ahead of print].
REVIEW by Jill M. Cholette:
This retrospective cohort study identified factors associated with bleeding events during paediatric extracorporeal membrane oxygenation (ECMO). Over a 2-year period, 122 children on ECMO (median age 17 weeks, 46% with congenital heart disease) were studied. Bleeding days were defined by surgical exploration, rFVIIa administration, GI, pulmonary or intracranial hemorrhages. Logistic regression assessed factors associated with bleeding days.
Bleeding days comprised 179 (16%) of 1121 ECMO days. By day 4, 50% experienced a bleeding day. Factors significantly associated with bleeding included: central cannulation, older age, higher lactate and lower platelet counts (platelet function was not assessed).
Patients who experienced more frequent bleeding (>75th percentile) had significantly fewer ventilator-free and hospital-free days in the 60 days after cannulation and higher in-hospital mortality (68 vs. 34%; P < 0.001).
There is no current reliable definition of bleeding on ECMO to guide practice and further work regarding bleeding and transfusion guidelines in this critical population is warranted.
