The relationship between transfusion and hypoxemia in combat casualties

Massive Transfusion in the Intensive Care Unit

The transfusion of blood components is a frequent oc currence in intensive care units. Transfusion therapy demands not only a greater ability to meet the needs of acutely ill patients with an enlarging variety of blood components, but also an increasing awareness of previ ously recognized and newly discovered hazards of blood transfusion. Use of the more commonly adminis tered blood components is reviewed, and problems per tinent to the large-volume massively or multiply trans fused patient are discussed.

Trauma team

The introduction of trauma teams has improved patient outcome independently. The aim of establishing a trauma team is to ensure the early mobilization and involvement of more experienced medical staff and thereby to improve patient outcome. The team approach allows for distribution of the several tasks in assessment and resuscitation of the patient in a 'horizontal approach', which may lead to a reduction in time from injury to critical interventions and thus have a direct bearing on the patient's ultimate outcome. A trauma team leader or supervisor, who coordinates the resuscitation and ensures adherence to guidelines, should lead the trauma team. There is a major national and international variety in trauma team composition, however crucial are a surgeon, an Emergency Medicine physician or both and anaesthetist. Advanced Trauma Life Support training, simulation-based training, and video review have all improved patient outcome and trauma team performance. Developments in the radiology, such as the use of computed tomography scanning in the emergency room and the endovascular treatment of bleeding foci, have changed treatment algorithms in selected patients. These developments and new insights in shock management may have a future impact on patient management and trauma team composition.

Nitric oxide scavenging by red blood cell microparticles and cell-free hemoglobin as a mechanism for the red cell storage lesion

BACKGROUND

Intravascular red cell hemolysis impairs nitric oxide (NO)-redox homeostasis, producing endothelial dysfunction, platelet activation, and vasculopathy. Red blood cell storage under standard conditions results in reduced integrity of the erythrocyte membrane, with formation of exocytic microvesicles or microparticles and hemolysis, which we hypothesized could impair vascular function and contribute to the putative storage lesion of banked blood.

METHODS AND RESULTS

We now find that storage of human red blood cells under standard blood banking conditions results in the accumulation of cell-free and microparticle-encapsulated hemoglobin, which, despite 39 days of storage, remains in the reduced ferrous oxyhemoglobin redox state and stoichiometrically reacts with and scavenges the vasodilator NO. Using stopped-flow spectroscopy and laser-triggered NO release from a caged NO compound, we found that both free hemoglobin and microparticles react with NO about 1000 times faster than with intact erythrocytes. In complementary in vivo studies, we show that hemoglobin, even at concentrations below 10 μmol/L (in heme), produces potent vasoconstriction when infused into the rat circulation, whereas controlled infusions of methemoglobin and cyanomethemoglobin, which do not consume NO, have substantially reduced vasoconstrictor effects. Infusion of the plasma from stored human red blood cell units into the rat circulation produces significant vasoconstriction related to the magnitude of storage-related hemolysis.

CONCLUSIONS

The results of these studies suggest new mechanisms for endothelial injury and impaired vascular function associated with the most fundamental of storage lesions, hemolysis.

Pulmonary contusion: an update on recent advances in clinical management

Pulmonary contusion is a common finding after blunt chest trauma. The physiologic consequences of alveolar hemorrhage and pulmonary parenchymal destruction typically manifest themselves within hours of injury and usually resolve within approximately 7 days. Clinical symptoms, including respiratory distress with hypoxemia and hypercarbia, peak at about 72 h after injury. The timely diagnosis of pulmonary contusion requires a high degree of clinical suspicion when a patient presents with trauma caused by an appropriate mechanism of injury. The clinical diagnosis of acute parenchymal lung injury is usually confirmed by thoracic computed tomography, which is both highly sensitive in identifying pulmonary contusion and highly predictive of the need for subsequent mechanical ventilation. Management of pulmonary contusion is primarily supportive. Associated complications such as pneumonia, acute respiratory distress syndrome, and long-term pulmonary disability, however, are frequent sequelae of these injuries.

Effects of fluid resuscitation with hypertonic saline dextrane or Ringer's acetate after nonhemorrhagic shock caused by pulmonary contusion

BACKGROUND

Injured lungs are sensitive to fluid resuscitation after trauma. Such treatment can increase lung water content and lead to desaturation. Hypertonic saline with dextran (HSD) has hyperosmotic properties that promote plasma volume expansion, thus potentially reducing these side effects. The aim of this study was to (1) evaluate whether fluid treatment counteracts hypotension and improves survival after nonhemorrhagic shock caused by lung contusion and (2) analyze whether resuscitation with HSD is more efficient than treatment with Ringer's acetate (RA) in terms of blood oxygenation, the amount of lung water, circulatory effects, and inflammatory response.

METHODS

Twenty-nine pigs, all wearing body armor, were shot with a 7.62-mm assault rifle to produce a standardized pulmonary contusion. These animals were allocated into three groups: HSD, RA, and an untreated shot control group. Exposed animals were compared with animals not treated with fluid and shot with blank ammunition. For 2 hours after the shot, the inflammatory response and physiologic parameters were monitored.

RESULTS

The impact induced pulmonary contusion, desaturation, hypotension, increased heart rate, and led to an inflammatory response. No change in blood pressure was observed after fluid treatment. HSD treatment resulted in significantly less lung water (p < 0.05) and tended to give better Pao2 (p = 0.09) than RA treatment. Tumor necrosis factor-α release and heart rate were significantly lower in animals given fluids.

CONCLUSION

Fluid treatment does not affect blood pressure or mortality in this model of nonhemorrhagic shock caused by lung contusion. However, our data indicate that HSD, when compared with RA, has advantages for the injured lung.

Pulmonary transfusion reactions

Transfusion reactions remain a common complication of transfusion therapy; reactions affecting the lungs are some of the most serious. Several different mechanisms are responsible for pulmonary transfusion reactions, and most cause adverse effects in addition to lung injury. Fluid overload can lead to pulmonary edema, antibodies reacting with plasma proteins can cause bronchospasm and anaphylaxis, and particulate matter can produce microemboli. These reactions are well understood and usually can be prevented. Transfusions are also associated with acute lung injury and acute respiratory distress syndrome (ARDS), but their etiology is poorly understood and they remain clinically problematic. Neutrophil antibodies cause some of these serious as well as mild pulmonary reactions, but the exact role of leukocyte antibodies in pulmonary reactions remains unclear. Other blood donor, blood component, and transfusion recipient factors likely play a contributing or modulating role in pulmonary transfusion reactions, but prospective studies are needed to better understand their role.

Early onset of acute pulmonary dysfunction after cardiovascular surgery: risk factors and clinical outcome

OBJECTIVE

To define the incidence, risk factors, and clinical outcome of early pulmonary dysfunction after cardiovascular surgery for adults.

STUDY

Inception cohort.

SETTING

Adult cardiovascular intensive care unit (ICU).

PATIENTS

All adult admissions after cardiovascular surgery without preoperative pulmonary parenchyma or vascular disease over a period of 12 consecutive months.

INTERVENTION

Collection of data on demographics, preoperative organ insufficiency, emergency surgery, type of surgical procedure, cardiopulmonary bypass time, transfusion of blood products, postoperative arterial blood gases, and systemic hemodynamics on admission to the cardiovascular ICU.

MEASUREMENTS AND MAIN RESULTS

Early postoperative pulmonary dysfunction was defined by mechanical ventilation with a PaO2/FIO2 ratio of < or = 150 torr (< or = 20 kPa) and chest radiography on admission to the cardiovascular ICU. Secondary outcome included postoperative renal and neurologic dysfunction, nosocomial infections, length of mechanical ventilation, hospitalization, and death. A total of 3,122 patients were evaluated and 1,461 patients satisfied the entry criteria of the study. Early postoperative pulmonary dysfunction was present in 180 (12%) patients on admission to the cardiovascular ICU. Preoperative variables: age of > or = 75 yrs (odds ratio 1.69, 95% confidence interval [CI] 1.06 to 2.65), body mass index of > or = 30 kg/m2 (odds ratio 1.60, 95% CI 1.09 to 2.32), mean pulmonary arterial pressure of > or = 20 mm Hg (odds ratio 1.60, 95% CI 1.13 to 2.28), stroke volume index of < or = 30 mL/m2 (odds ratio 1.57, 95% CI 1.08 to 2.26), serum albumin (odds ratio 0.71, 95% CI 0.49 to 0.97), history of cerebral vascular disease (odds ratio 1.81; 95% CI 1.08 to 2.96); operative variables: emergency surgery (odds ratio 2.12, 95% CI 1.01 to 4.51), total cardiopulmonary bypass time of > or = 140 mins (odds ratio 1.54, 95% CI 1.0 to 2.34); and postoperative variables (on admission to cardiovascular ICU): hematocrit of > or = 30% (odds ratio 2.46, 95% CI 1.71 to 3.56), systemic mean arterial pressure of > or = 90 mm Hg (odds ratio 1.67, 95% CI 1.13 to 2.42), and cardiac index of > or = 3.0 L/min/m2 (odds ratio 2.09, 95% CI 1.44 to 3.01) were predictors of early postoperative pulmonary dysfunction. Pulmonary dysfunction was associated with a postoperative increase of serum creatinine (1.36 +/- 0.4 vs. 1.24 +/- 0.4 mg/dL, p < .02), neurologic complications (3% vs. 1.6%, p < .001), nosocomial infections (3% vs. 1.6%, p < .001), prolonged mechanical ventilation (2.2 +/- 5.9 vs. 1.7 +/- 5.6 days, p < .001), length of stay in the cardiovascular ICU (4.4 +/- 12.2 vs. 2.6 +/- 6.2 days, p < .001) and hospital (14.8 +/- 13.1 vs. 10.5 +/- 8.0 days, p < .001), and death (4.4% vs. 1.6%, p < .001).

CONCLUSIONS

The incidence of early postoperative pulmonary dysfunction is uncommon; however, once developed, it is associated with increased morbidity and mortality after cardiovascular surgery. Advanced age, large body mass index, preoperative increased pulmonary arterial pressure, low stroke volume index, hypoalbuminemia, history of cerebral vascular disease, emergency surgery, and prolonged cardiopulmonary bypass time are risk factors for early onset of severe pulmonary dysfunction after surgery. Postoperative hematocrit and systemic hemodynamics suggest that early postoperative pulmonary dysfunction can be a component of a generalized inflammatory reaction to cardiovascular surgery.

Pulmonary contusion: review of the clinical entity

Pulmonary contusion is a common lesion occurring in patients sustaining severe blunt chest trauma. Alveolar hemorrhage and parenchymal destruction are maximal during the first 24 hours after injury and then usually resolve within 7 days. The diagnosis of traumatic lung injury is usually made clinically with confirmation by chest x-ray films. The chest computed tomography scan is highly sensitive in identifying pulmonary contusion and may help predict the need for mechanical ventilation. Respiratory distress is common after lung trauma, with hypoxemia and hypercarbia greatest at about 72 hours. Although management of patients with pulmonary contusion is supportive, pneumonia and adult respiratory distress syndrome with long-term disability occur frequently.

Resuscitation of pulmonary contusion: effects of a red cell substitute

OBJECTIVE

To determine the impact of a vasoactive red cell substitute, diaspirin cross-linked hemoglobin, on respiratory derangements after traumatic lung injury.

DESIGN

Randomized, controlled animal experiment.

SETTING

Large-animal laboratory.

SUBJECTS

Mechanically ventilated, anesthetized young Yorkshire male swine (15 to 20 kg).

INTERVENTIONS

Pigs (n = 6/group) received two pneumatic blasts to the right thoracic cage at baseline, were hemorrhaged 30 mL/kg from t = 0 to 20 mins, resuscitated with 0.9% saline (group 1, 90 mL/ kg) or diaspirin cross-linked hemoglobin (group 2, 15 mL/kg) from t = 20 to 40 mins, and then observed to t = 240 mins.

MEASUREMENTS AND MAIN RESULTS

Serial pulmonary and systemic hemodynamic measurements, total thoracic compliance assessment, spiral three-dimensional computed tomography scan, and lung weights (n = 3/group) were used to assess lesion size and lung water. Mean arterial pressure was restored in both animal groups. Mean pulmonary arterial pressure was significantly higher after resuscitation in animals receiving the red cell substitute. Oxygenation worsened mildly in both groups. Compliance diminished in both groups but was significantly worse at the end of the experiment in animals infused with diaspirin cross-linked hemoglobin. Right lung weights and right thoracic computed tomography scan volume were higher with diaspirin cross-linked hemoglobin than with saline.

CONCLUSIONS

After pulmonary contusion, resuscitation with diaspirin cross-linked hemoglobin led to pulmonary hypertension, greater pulmonary contusion lesion size, and stiffer lungs in this porcine model.

Adult respiratory distress syndrome: a systemic overview of incidence and risk factors

OBJECTIVE

To determine the published incidence of adult respiratory distress syndrome (ARDS) as well as the clinical evidence supporting a casual association between ARDS and its major risk factors.

DATA SOURCES

The National Library of Medicine MEDLINE database and the bibliographies of selected articles.

STUDY SELECTION

Clinical studies were selected from the English literature, if they pertained to either the incidence of ARDS or its association with one or more commonly identified risk factors.

DATA EXTRACTION

All relevant studies identified by the search were evaluated for strength of design, and risk factors were scored according to established criteria for the strength of causation.

DATA SYNTHESIS

A total of 83 articles were considered relevant: six of incidence and 77 on risk factors. Only 49% of the 83 articles provided a definition of ARDS; a definition of risk factors was given in 64%, and 23% had no definition for either ARDS or risk factors. The published, population-based incidence of ARDS ranges from 1.5 to 5.3/10(5) population/yr. The strongest clinical evidence supporting a cause-effect relationship was identified for sepsis, aspiration, trauma, and multiple transfusions. The weakest clinical evidence was identified for disseminated intravascular coagulation. The following study types were represented by the 77 articles on risk factors: observational case-series (56%); cohorts (23%); case-controls (12%); nonrandomized clinical trials (5%); and randomized clinical trials (3%). Only a single study reported an odds ratio.

CONCLUSIONS

The significant variation in the incidence of ARDS is attributed to differences in the type and strength of study designs, as well as definitions or ARDS. While a substantial body of evidence exists concerning a casual role of ARDS risk factors, such as sepsis, aspiration, and trauma, > 60% of clinical studies employed weak designs. The lack of reproducible definitions for ARDS or its potential risk factors in 49% of studies raises concerns about the validity of the conclusions of these studies regarding the association between ARDS and the supposed risk factors.

Perioperative haemotherapy: II. Risks and complications of blood transfusion

Major life-threatening complications following blood transfusion are rare and human error remains an important aetiological factor in many. The infectious risk from blood transfusion is predominantly hepatitis, and non-A, non-B and hepatitis C (HCV) are the most common subtypes noted. The risk of post-transfusion hepatitis (PTH) appears to be decreasing and this is attributed to both deferral of high-risk donors and more aggressive screening of donated blood. Screening for HCV is expected to decrease this risk further. The risk of HIV transmission following blood transfusion is negligibly small. There are data to suggest that perioperative blood transfusion results in suppression of the recipient's immune system. Earlier recurrence of cancer and an increased incidence of postoperative infection have been associated with perioperative blood transfusion although the evidence is not persuasive. Microaggregate blood filters are not recommended for routine blood transfusion but do have a role in the prophylaxis of non-haemolytic febrile reactions caused by platelet and granulocyte debris in the donor blood. Patients should be advised when there is likely to be a requirement for perioperative blood transfusion and informed consent for transfusion should be obtained.

In vitro evaluation of a new dual screen microaggregate filter

We compared a new second generation 40/150 mum dual screen microaggregate filter with a currently available 40 mum screen microaggregate filter. The evaluation included comparison of filter flow rate, capacity, degree of microaggregate removal, degree of leukocyte removal, and extent of filtration-induced hemolysis. We also studied the effect of both devices on filtration of stored platelet concentrates. The 40/150 mum dual screen microaggregate filter showed results comparable to that of the control screen filter following filtration of various types of units of red blood cells as well as units of stored platelet concentrates. Importantly, mean flow rates with the new 40/150 mum filter of 45 g/min after gravity filtration of 1600 mL of blood, make the filter suitable for use in trauma or other massive transfusion settings. We conclude that this new second generation microaggregate filter is suitable for use in clinical transfusion practice.

Transfusion therapy in emergency medicine

Volume replacement is critical to the resuscitation of the hemorrhaging patient, but this usually can be accomplished quickly and safely with crystalloid and/or colloid solutions. Red cells should be used in addition to asanguinous fluids in the treatment of tissue hypoxia due to anemia. The need for whole blood as opposed to packed red blood cells is controversial. However, plasma should not be used as a volume expander, and its use to supplement coagulation factors during the massive transfusion of red cells should be guided by laboratory tests that document a coagulopathy. Similarly, platelet transfusions are indicated to correct documented thrombocytopenia or platelet dysfunction, and routine prophylaxis after fixed volumes of red cells results is unwarranted. Many anticipated complications of massive transfusions, including hemostatic abnormalities, acid-base imbalances, hyperkalemia, and hypocalcemia, are uncommon or of limited clinical significance. The risks of immune hemolysis and transfusion-transmitted diseases, on the other hand, are significant, and argue for judicious use of blood components. In emergencies in which blood is required immediately before compatibility testing can be completed, O-negative uncrossmatched blood can be requested. Careful blood specimen collection and patient identification prior to transfusion are critical. Practices that emphasize blood conservation, including the use of autologous salvaged blood, are always to the patient's advantage.

The mechanisms and management of noncardiogenic pulmonary edema following cardiopulmonary bypass

Cardiac surgeons have become more cognizant of the syndrome of noncardiogenic pulmonary edema after cardiopulmonary bypass. Although this syndrome is rare, its occurrence can be catastrophic. This article reviews the current understanding of several factors that have been implicated in the cause of this syndrome and discusses the various options for management of the problem once it has arisen.

Role of blood transfusion in organ system failure following major abdominal surgery

Using multivariate probit analysis, the data of 565 patients who underwent major abdominal surgery were retrospectively analyzed, and the etiologic role of blood transfusion in organ system failure (OSF), which includes respiratory failure, gastrointestinal stress bleeding, renal failure, nonobstructive, nonhepatitic jaundice, and coagulopathy, was studied. Apart from the amount of blood transfusion, the following factors were included in the analysis as possible contributors to OSF: age, preoperative hematocrit, organ failure risk (diffuse peritonitis, obstructive cholangitis, liver cirrhosis, terminal cancer, and hemorrhagic shock), operative time, blood loss, and postoperative highest hematocrit. The results showed that, except for preoperative hematocrit, all the factors are statistically significant contributors, blood transfusion being the most significant. There was no statistically significant interaction between blood transfusion and organ failure risk. It is concluded that blood transfusion is an important, independent factor contributing to OSF, and its contribution cannot be attributed to the underlying conditions that require blood transfusion.

The present status of blood filtration

The history of the development of blood transfusion and blood filtration is outlined. Clinical and experimental evidence for the efficacy of microfiltration in both small and large volume transfusions is evaluated. Though microfilters do remove the micro-aggregates from stored blood, the results of clinical studies suggest that both the debris from septic processes in the body and the formation of micro-aggregates in the blood stream triggered by processes such as complement activation play a far more important role in the pathogenesis of adult respiratory distress syndrome. If this is so the enhancement of the reticulo-endothelial system by fibronectin therapy may be indicated. It also follows that the use of microfilters is probably an unnecessary expense and, where exsanguination is a risk, may be positively dangerous. Microfilters have been found useful in the preparation of granulocyte-free transfusions after centrifugation of the blood, but their routine use for transfusions, small or large, remains to be justified.

Intraoperative autologous transfusion

Intraoperative autologous transfusion is a technique that was first used almost 2 centuries ago but that has realized its potential only in the past 5 years. A growing national awareness of transfusion-related morbidity, of the need for alternative blood sources, and of improved methods for red blood cell recovery has led to an increased frequency of use of autologous transfusion. Most hospital programs use semicontinuous flow centrifugation or canister technology for the intraoperative salvage and reinfusion of shed blood. This technique is particularly valuable for cardiovascular surgical procedures but has been useful in many other types of surgical procedures as well. Deleterious effects formerly attributed to this technique have been eliminated by methodologic improvements. Concerns about use of autologous transfusion in patients who have an infection or a malignant lesion persist. Most hematologic aberrations are related to massive transfusions and should not be considered a contraindication to the general use of autologous blood.

Haemotherapy for acute haemorrhage

In this review blood component therapy for acute haemorrhage is summarised. As the haemotherapy is frequently the cornerstone of a successful outcome of haemorrhagic shock, attention to details in relation to the indications, safety and efficiency is essential. Massive blood transfusion brings with it many potential complications which may jeopardise a successful outcome for the patient after skillful medical and surgical care has controlled the basic problem.

Clinical predictors of the adult respiratory distress syndrome

One hundred thirty-six patients meeting our criteria for one or more of eight clinical conditions were prospectively observed for the development of the adult respiratory distress syndrome. A high risk population was identified, including those with sepsis syndrome (38 percent), documented aspiration of gastric contents (30 percent), multiple emergency transfusions (24 percent), and pulmonary contusion (17 percent). The risk from multiple major fractures appeared low but contributed to the risk from other factors. The risk associated with just one factor (25 percent) was compounded by the presence of two (42 percent) and three (85 percent) simultaneous factors, and this finding was more predictive of ARDS than the injury severity score or initial arterial oxygenation. Of the ARDS cases, 76 percent occurred in the initial 24 hours after meeting the criteria. ARDS did not occur after 72 hours unless there was late development of sepsis (3 of 136 patients).

Cardiopulmonary function following transfusion of three red blood cell products in elective abdominal aortic aneurysmectomy

In order to evaluate the importance of red cell O(2) affinity of transfused blood on cardiac performance and adverse effects of transfusion on lung function, a prospective double-blind protocol was used in 27 patients undergoing abdominal aortic aneurysmectomy. Three types of blood were administered: packed red cells (PC), washed red cells (WC) and high 2,3 DPG red cells (2,3 DPG). An average of 4.5 units of blood was used per patient. Transfusion of 2,3 DPG blood resulted in maintenance of in vivo P(50) during surgery and an increase to 31.2 torr after operation (p < 0.001). An intraoperative fall in in vivo P(50) to 23.2 +/- 2.0 torr was observed in patients who were transfused with PC (p < 0.001) and to 25.1 +/- 2.6 torr with WC (p < 0.005). A fall in body temperature averaging 2.2 C intraoperatively was noticed in all three groups. After operation, in vitro P(50) decreased in patients transfused with PC (p < 0.005) and WC (p < 0.005) while it remained unchanged in the high 2,3 DPG group. This was consistent with the decrease of red cell 2,3 DPG in the PC (p < 0.001) and WC groups (p < 0.01) and maintenance in the 2,3 DPG group. Left ventricular stroke work and volume loading Starling type myocardial performance curves were similar for the three groups. Microaggregates measured by Coulter counting and screen filtration pressure were the same for all three products in samples drawn on both sides of the 40 microm transfusion filter. There was no relationship of transfusion volumes or type of blood product to changes in lung function (physiologic shunting, dynamic compliance and pulmonary arterial pressure) in the three groups of patients. The false negative, beta, error of missing a true 25% difference was less than 10%. It is concluded that 2,3 DPG enriched red cells improved oxygen availability, but that a 4.5 unit transfusion of any of the three blood products did not influence lung function or myocardial performance following aneurysmectomy.

An in vivo evaluation of microaggregate blood filtration during total hip replacement

In order to evaluate the effect of microaggregate blood filtration on pulmonary status, hemostatic status, and incidence of infection, a prospective study was performed on patients undergoing elective total hip replacement for osteoarthritis. Forty patients were randomized to either a 260 micron standard filter group or a 20 micron microaggregate filter group. Patients were monitored pre- and postoperatively for changes in arterial blood gases and in vitro test of hemostasis. Postoperative measurements were also made of the total volume of blood collected from the operative wound drain and of the number of infections incurred by patients in the two filter groups. Average transfusion was 4.0 units for the standard filter group and 4.6 units for the microaggregate filter group. Results showed that postoperatively, either immediately or after 48 hours, there were no statistically significant differences (p greater than 0.05) between either filter group for any of the tests of pulmonary or hemostatic function evaluated. For infection no trends were found to suggest that microaggregate filters conveyed any protective effect. These data suggest that routine microaggregate blood filtration of up to 5 units of blood is not required.

Effect of micropore filtration on pulmonary function after massive transfusion

1. This study demonstrates a positive correlation between the number and size of infused microaggregates and the subsequent abnormality in pulmonary function as measured by oxygenation and dead space. 2. No such correlation between the severity of injury and the altered pulmonary function or transfusion volume was demonstrated. 3. We were unable to demonstrate an advantage to the use of a 40 mu micropore filter in preventing the adult respiratory distress syndrome (ARDS) or in improving pulmonary function in our patients. 4. One explanation for the failure to demonstrate such an advantage is the low efficiency of the filter used.