Introduction
Numerous preparations of intravenous (IV) fluid are available for the replacement of perioperative fluid losses in patients undergoing major surgery. The selection of a specific fluid may be influenced by multiple factors, such as availability, cost, and tradition. Of late, attention has focused on the possible systemic effects of the various fluid preparations. Additionally, there is awareness that particular fluids may not only influence clinical variables during the intraoperative period but may also affect postoperative outcome. Clinicians’ choice of fluid replacement therapy for patients undergoing major surgical procedures is increasingly influenced by the beneficial or detrimental effects of IV fluids, independent of their efficacy as blood volume expanders.
Many clinical and experimental studies have been carried out to determine the potential clinical effects of IV fluids. Unfortunately, there is a paucity of large, prospective, randomized, blinded clinical studies of the effects of the intraoperative administration of IV fluids on clinical outcomes, despite the fact that approximately 3 million major surgical procedures are performed annually in the United States alone. However, multiple outcomes have been examined in small investigations in numerous and diverse patient populations and in studies of healthy volunteers.
In addition to the fluids themselves, a separate body of literature has looked at the way in which these fluids are administered. Fluid studies are often labeled restrictive, liberal, or goal directed. Goal-directed fluid therapy is a term that relates to the use of an algorithm to guide administration of fluids to optimize hemodynamic status.
To address the question “Does the choice of fluid matter in major surgery?” we will consider the available data from clinical studies of intravascular volume replacement in patients undergoing major surgery. The interpretation of studies of IV fluids is somewhat confounded by their size and design. In many cases, only small numbers of patients have been studied. These trials may not have sufficient power to detect differences in clinically relevant outcomes, and their results are therefore interpreted with this caveat in mind.
Options/Therapies
Traditionally, IV fluids have been classified according to whether they are crystalloid or colloid in nature. Crystalloid fluids comprise electrolyte solutions with or without bicarbonate or one of its precursors, such as acetate or lactate. The colloids contain a complex sugar or protein suspended in an electrolyte solution. A further distinction between IV fluid types may be based on the nature of the solution. Preparations based on 0.9% normal saline (NS) (crystalloid or colloid) contain no electrolytes other than sodium and chloride. In contrast, balanced salt (BS)-based fluids such as lactated Ringer (LR) solution are those that contain other electrolytes with or without bicarbonate or a bicarbonate precursor.
Several types of colloid are available, but three are most commonly used: hydroxyethyl starch (HES), gelatin, and albumin. The HES preparations differ from one another according to their concentration, molecular weight, and extent of hydroxyethylation or substitution, with resultant varying physiochemical properties. HES solutions may be described according to concentration (3%, 6%, 10%), weight-averaged mean molecular weight in kilodaltons (kDa) (high-molecular weight [450-670 kDa], middle-molecular weight [200-270 kDa], low-molecular weight [70-130 kDa]), and the molar substitution (0.4-0.7). HES 670/0.75 and 130/0.4 are available in an NS solution (e.g., HES 130/NS) and in a BS solution (e.g., HES 130/BS). Two forms of gelatin are available: modified (succinylated) and the polygelines. Although all of these colloids are used in Europe, gelatins are not available in the United States, and the only HES preparations approved by the Food and Drug Administration (FDA) are the 670 kDa BS (Hextend, Hospira, Lake Forest, Ill), 670kDa NS (Hespan, B Braun, Irvine, Calif), and 130/NS (Voluven, Fresenius Kabi, Germany) formulations.
Evidence
Does the Choice of Crystalloid Matter in Major Surgery?
Crystalloid therapy is an essential part of all fluid therapy regimens. Electrolyte-containing crystalloid solutions are distributed throughout the extracellular compartment and are used to replace insensible and evaporative losses and urine output during major surgery.
The most commonly used crystalloid in the world is 0.9% NS. However, evidence is emerging that it may cause significant harm. NS is not normal in the physiologic meaning of the word because it contains 154 mmol/L of chloride, which is significantly more than the plasma concentration of 105 mmol/L. Administration of NS and NS-based fluids therefore causes a predictable hyperchloremic metabolic acidosis. This phenomenon has been known for many years, but the question remains, does this acidosis cause any harm?
Shaw and colleagues investigated this risk of harm in a recently published large retrospective database analysis comparing the administration of either NS alone or a balanced crystalloid solution (Plasma-Lyte, Baxter, Ill) alone on the day of surgery. They found that in the NS group the use of special investigations (measurement of arterial blood gases and lactate levels) and treatments (buffers, blood products) was increased, presumably because saline-induced acid–base abnormalities needed to be investigated and managed. When compared with administration of a balanced electrolyte solution, the administration of NS was also found to cause a significantly greater risk of postoperative infection and renal failure requiring dialysis. It is known that hyperchloremia may cause renal vasoconstriction and a decrease in the glomerular filtration rate, which may explain, in part, the mechanism for NS-induced changes in renal performance. Alternatively, the metabolic acidosis itself may induce vasoconstriction and redistribution of intrarenal blood flow with subsequent effects on function.
Several prospective, randomized studies have compared the effects of NS-based and BS-based fluids and have observed a worse urinary output in patients treated with the NS-based fluid preparations. Other investigators have not noted superior renal function after the administration of BS fluids. Intraoperative urine output was greater in patients who received NS than in patients who were given LR during abdominal aortic aneurysm repair. However, NS-treated patients received a larger volume of fluid than patients in the LR group, as well as significant quantities of sodium bicarbonate intraoperatively for treatment of hyperchloremic metabolic acidosis, which suggests that the prevention or treatment of hyperchloremic metabolic acidosis may have negated the negative impact of NS on renal function in some way.
Two studies have looked at the impact of crystalloid choice in patients undergoing kidney transplantation. In both studies patients were randomly assigned to receive either NS or LR for intraoperative fluid resuscitation, and investigators found that hyperkalemia and acidosis were more common in the NS group. In the study by O’Malley and colleagues, eight patients (31%) in the saline group versus zero patients (0%) in the LR group were treated for metabolic acidosis. In addition, five patients (19%) in the saline group had potassium concentrations greater than 6 mEq/L and were treated for hyperkalemia versus zero in the LR group. These results suggest that acidosis-induced extracellular shift of potassium causes a greater risk of hyperkalemia than the small amount of potassium in LR and challenges the dogma that NS should be administered to patients with renal failure.
Other studies suggest NS-based fluids may be associated with more bleeding than BS solutions. Waters and colleagues reported that patients undergoing abdominal aortic aneurysm repair who received LR solution were given smaller volumes of platelets and had less blood product exposure than those treated with NS. HES 670/NS may also be associated with more bleeding than HES 670/BS solution. In a study of 120 major surgical patients, blood loss was greater among patients who received HES 670/NS than in patients who received HES 670/BS.
When differences between fluid types are seen, they may be mediated through impaired platelet function, possibly as a consequence of diminished circulating von Willebrand factor (vWF) antigen and vWF : ristocetin cofactor in patients treated with NS-based rather than BS-based fluids. A second possible explanation is the lack of calcium in NS and related fluids. Calcium is a necessary cofactor at several points in the coagulation process. It is necessary for activation of clotting factors, as well as for normal platelet function. In particular, calcium binding is a prerequisite for the stability and function of the platelet GPIIb/IIIa receptor. This receptor binds fibrinogen and vWF with resultant platelet activation and aggregation. With blood loss and IV fluid administration, ionized calcium levels may fluctuate, and this variation may affect coagulation. Ionized calcium levels may be lower after administration of NS and related fluids rather than after administration of BS fluids. The presence of calcium in IV fluids may maintain more constant plasma calcium levels, avoiding the potential detrimental effect of low or fluctuating ionized calcium levels on coagulation. Thirdly, the role of hyperchloremic acidosis must also be considered because acidosis has been implicated in coagulation derangements.
Wilkes and colleague have also implicated NS-based fluids in splanchnic ischemia and postoperative nausea and vomiting. Elderly surgical patients were randomly assigned to receive either a combination of HES 670/BS and LR or a combination of HES 670/NS and NS for intraoperative fluid replacement. In the BS-based-fluid–treated group there was a smaller intraoperative increase in the CO 2 gap, indicating that BS-based fluids are associated with superior splanchnic perfusion as compared with NS-based fluids. It was postulated that impaired gut perfusion or hyperchloremia associated with NS-based preparations might have caused an impairment of splanchnic perfusion in the patients who were administered NS and HES 670/NS. Of note, the poor splanchnic perfusion in patients treated with the NS-based regimen may have been mediated by generalized vasoconstriction (perhaps secondary to metabolic acidosis), given that these patients also exhibited other evidence of vasoconstriction such as lower urine flow rates and lower peripheral-to-core temperature gradients (reflecting peripheral vasoconstriction).
It is also interesting to note that in a randomized crossover study of healthy volunteers, subjective deterioration in mental status (i.e., lassitude and difficulty in abstract thinking) was reported only by individuals who received NS and not by those who received LR. The possible effect of different IV fluid preparations on central nervous system function has not yet been explored in prospective, randomized clinical studies of patients undergoing major surgical procedures.
The one area of practice in which the use of NS may be prudent is in traumatic brain injury. A post hoc subset analysis of the SAFE study for patients with traumatic brain injury ( n = 460) revealed a lower mortality rate in patients treated with NS compared with albumin (33.2% versus 20.4%). Otherwise, the use of balanced crystalloid solutions seems to cause no harm and could possibly be beneficial. Unfortunately, no significant studies have compared different balanced regimens. In particular, there are no data comparing the different anions (i.e., lactate, acetate, and gluconate) used as alternatives to chloride.
Does the Choice of Colloid Matter in Major Surgery?
Colloids have a smaller volume of distribution than crystalloids, and the majority of the solution remains in the intravascular space. Colloid therapy is therefore frequently used to restore intravascular volume depletion from blood loss or protein-rich shifts to the interstitial space during major surgery.
Impact of Colloids on Coagulation
Albumin has not been associated with significant clotting abnormalities or perioperative bleeding. The gelatins have been associated with minor abnormalities in coagulation, possibly because of derangements in fibrinogen polymerization. However, this has not been associated, other than on an anecdotal basis, with clinically significant perioperative bleeding.
There have been consistent reports of coagulation impairment with the older HES preparations since they were introduced into clinical practice. HES macromolecules interact with platelets and the coagulation cascade, causing a decrease in factors such as factor VIII and vWF. In a study of patients undergoing off-pump cardiac surgery, HES 670/BS increased bleeding and transfusion requirements compared with 4% albumin. Two large observational studies found HES 670/0.7 was an independent risk factor for postoperative hemorrhaging.
The effect of third-generation starches on coagulation is controversial. Voluven130/NS (Fresenius Kabi, Bad Homburg, Germany) was designed with a better coagulation profile in mind and has a molar substitution of 0.4 and a C 2 /C 6 ratio of approximately 9 : 1. Third-generation starches formulated in BS solutions are available around the world but are not commercially available in the United States at the current time.
Several in vitro studies have demonstrated that third-generation HES products have a lesser effect on coagulation. However, in vitro studies can be misleading because they cannot mimic the in vivo environment that will occur during progressive hemodilution. The administration of a large volume of any type of IV fluid will cause dilution of platelets and coagulation factors and may lead to coagulopathy. Because of the multifactorial etiology of bleeding during surgery, it is impossible to know, in any given patient, whether the type of fluid administered is a cause of bleeding independent of the impact of hemodilution. Only properly designed, randomized, clinical trials can determine fluid-specific effects on bleeding and other clinical outcomes. Although many studies report some clinical outcomes related to bleeding, a large number focus on measurements or markers of coagulation and have not been designed to explore outcomes of more clinical relevance such as blood product usage and surgical re-exploration for bleeding.
With this in mind, several studies demonstrated that Voluven has fewer adverse effects on coagulation compared with the higher molecular weight starches or has an impact on coagulation similar to gelatin. A recent meta-analysis showed a small reduction in bleeding and blood product transfusion with HES 130/0.4 compared with HES 200/0.5. However, other studies have showed similar effects on coagulation with different HES solutions ( Table 27-1 ).
| Author, Year | Control Fluid(s) | N | Type of Surgery | Outcome |
|---|---|---|---|---|
| Gallandat Huet, 2000 | HES 200/0.5 | 59 | Cardiac | Less effect on vWF with HES 130/0.4 Less blood loss and use of blood products with HES 130/0.4 |
| Hanart, 2009 | 4% albumin | 119 | Pediatric cardiac | No difference in blood loss Higher use of blood products in the albumin group |
| Langeron, 2001 | HES 200/0.5 | 100 | Orthopedic | Less effect on FVIII levels and APTT with HES 130/0.4 No difference in use of blood products |
| Kasper, 2003 | HES 200/0.5 | 120 | Cardiac | No difference in blood loss or use of blood products |
| Sander, 2003 | HES 200/0.5 | 60 | Major gynecologic | No difference in coagulation tests or blood loss |
| Van der Linden, 2005 | 3% gelatin | 132 | Cardiac | No difference in blood loss or use of blood products |
| Gandhi, 2007 | HES 670/0.75 | 100 | Orthopedic | Less effect on FVIII and vWF levels with HES 130/0.4 |
| Mittermayr, 2007 | LR | 66 | Orthopedic | Less clot firmness measured by TEG in the HES group |
| Mukhtar, 2009 | 5% albumin | 40 | Liver transplant | No difference in coagulation tests or use of blood products |
| Ooi, 2009 | 4% gelatin | 90 | Cardiac | No difference in blood loss or use of blood products |
| Schramko, 2009 | HES 200/0.5 4% albumin | 45 | Cardiac | Less clot firmness measured by TEG in both HES groups No difference in blood loss |
| Schramko, 2010 | 4% gelatin Ringer acetate | 45 | Cardiac | Less clot firmness measured by TEG in the HES group No difference in blood loss |
In pediatric cardiac patients, Hanart and colleagues compared pump priming with HES 130 or 4% albumin and showed no difference in bleeding; however, a greater need for allogeneic blood was seen in the albumin group. In another randomized controlled trial (RCT) of pediatric noncardiac surgery patients assigned to HES 130/0.4 or albumin, no differences in blood loss or coagulation variables were seen. Conversely, other studies in cardiac and noncardiac surgery have shown impaired thromboelastographic variables with HES 130/0.4 compared with albumin. Among patients with severe head injuries, high doses of HES 130/0.4 showed no difference in coagulation variables or bleeding compared with a combination of HES 200/0.5 and 5% albumin.
It is therefore difficult to draw firm conclusions on the effect of the third-generation starches on coagulation. HES 130/0.4 may have a lesser effect on coagulation than older starches; however, whether this is comparable with albumin needs to be confirmed in larger studies.
Impact of Colloids on Renal Function and Urine Output
The biggest concern about the administration of colloids, with the exception of albumin, is the effect on renal function. There are no reports of renal dysfunction after administration of albumin, even in severe sepsis. In contrast, some data suggest that gelatins may have some effect on renal function in patients with severe sepsis, although this is far from conclusive. By far the biggest concern, however, is with the HES solutions. The administration of older HES solutions to critically ill patients in the intensive care unit (ICU) is associated with the development of renal dysfunction. The Volume Substitution and Insulin Therapy in Severe Sepsis (VISEP) study showed higher rates of acute renal failure and the need for renal replacement therapy with HES compared with LR. However, the investigators used large doses of a hyperoncotic 10% pentastarch, frequently in excess of the daily dosing limit. Because the adverse effects of hyperoncotic starches on renal function is well-known, the general applicability of these results is open to interpretation.
Currently, the main controversy is regarding the effect of third-generation starches on renal function. Very few RCTs have studied the impact of HES 130 on renal function in the perioperative period. Two small studies compared HES 130/0.4 with gelatin in cardiac and major vascular surgery and showed no differences in renal function. Godet and colleagues suggested that HES 130/0.4 was comparable with gelatin in patients with pre-existing renal impairment undergoing abdominal aortic surgery ( Table 27-2 ).
| Author, Year | Control Fluid(s) | N | Type of Surgery | Outcome |
|---|---|---|---|---|
| Gallandat Huet, 2000 | HES 200/NS | 59 | Cardiac | No difference in urine output or serum creatinine |
| Langeron, 2001 | HES 200/NS | 100 | Orthopedic | No difference in urine output |
| Mahmood, 2007 | HES 200/NS Gelatin/NS | 62 | Abdominal aortic aneurysm | No difference in urine output. Less derangement in markers of glomerular and tubular function with HES 200/NS and HES 130/NS |
| Ooi, 2007 | 4% gelatin | 63 | Cardiac | No difference in estimated glomerular filtration rate |
| Godet, 2008 | 3% gelatin | 65 | Abdominal aortic aneurysm | No difference in urine output, creatinine clearance, or adverse renal events |
| Mukhtar, 2009 | 5% albumin | 40 | Liver transplant | No difference in serum creatinine, creatinine clearance, or cystatin C levels |
Therefore most of our information on effects of HES 130/0.4 on renal function comes from observational studies in the ICU. In a retrospective analysis of data from 3147 critically ill patients in the Sepsis Occurrence in Acutely Ill Patients (SOAP) study, HES 130/0.4 was not an independent risk factor for adverse effects on renal function, even in the 822 patients with severe sepsis or septic shock. However, in another large observational study of 2911 surgical ICU patients, HES 130/0.4 at doses greater than 33mL/kg was an independent risk factor for acute renal failure.
As this book goes to press, the 6S (Scandinavian Starch for Severe Sepsis/Septic Shock) Trial has just been published. This large RCT compared fluid resuscitation on the ICU in patients with severe sepsis. Patients were assigned to either HES 130/0.4 or Ringer acetate at a dose of up to 33 mL/kg of ideal body weight per day. Patients given HES 130/0.4 had an increased risk of death at day 90 and were more likely to require renal replacement therapy compared with those receiving Ringer acetate. Therefore the use of any hydroxyethyl starch in severe sepsis cannot be recommended. Crystalloid versus Hydroxyethyl Starch Trial (CHEST), a multicenter trial comparing 90-day all-cause mortality rates after infusion of HES 130/NS or NS alone in 7000 ICU patients, has also just been completed. Results were expected in late 2012.
Other Effects of Colloid Solutions
In a randomized study of 40 patients receiving either HES 200/NS or a gelatin solution during elective infrarenal aortic aneurysm repair, lower levels of inflammatory markers (i.e., C-reactive protein, microalbuminuria, and plasma vWF) were observed in the HES-treated group than in the gelatin-treated group after cross-clamp removal. These data suggest that HES may mediate the inflammatory response after major surgery.
Hyperamylasemia is associated with the administration of HES but not with other fluid types. Amylasemia is caused by HES through the formation of an HES–amylase complex with consequent reduction in elimination of amylase by the kidneys. This effect is greater with HES 200/NS than with HES 130/NS, which is consistent with the pharmacokinetics of different HES preparations. Intraneural deposition of HES has been purported to cause pruritus after HES administration. Small retrospective studies, in a number of patient populations, have reported a high incidence of HES-induced pruritus. However, no large epidemiologic studies examining this phenomenon have been performed in patients undergoing major surgery who have received large volumes of fluid. Interestingly, the incidence of postoperative pruritus in a prospective study of 750 surgical patients was similar (10%) in patients who received 500 mL of HES 200/NS and in patients who received 1000 mL of LR. The most important potential adverse effect of IV fluids is the occurrence of possibly life-threatening anaphylactic or anaphylactoid reactions. The incidence of severe anaphylactic reactions is 0.038% to 0.345% with gelatins, 0.0004% to 0.058% with HES administration, and 0.099% in patients who receive albumin.
Finally, significant cost reduction (32% to 35%) has been shown when HES was used for intraoperative fluid replacement rather than 5% albumin.
Perioperative Fluid Management
It is common for patients to receive IV fluid amounts that greatly exceed perioperative losses. Perioperative IV fluid regimens in abdominal surgery can lead to patients receiving 3.5 to 7 L of fluid on the day of surgery, leading to a 3- to 6-kg weight gain. This is done to replace assumed preoperative deficits, as well as insensible perspiration, “third space” losses, and urine output. Chappell and colleagues reviewed the evidence and suggested that losses via insensible perspiration have been grossly exaggerated and are probably no more than 1 mL/kg/h during major abdominal surgery. Additionally, a third space, as it was originally described, does not exist. There is no evidence for a nonfunctional space in which fluid is sequestered. Fluid is simply shifted perioperatively from the intravascular space toward the interstitium.
Brandstrup and colleagues found that a combination of crystalloid and colloid designed to avoid fluid overloading and maintain fluid balance, guided by body weight, significantly reduced postoperative complications and length of hospital stay after colorectal surgery. In the postoperative period, water and salt restriction to less than 2 L and 70 mmol, respectively, per day caused earlier return of bowel function and reduced weight gain and length of stay.
In noncardiac surgical patients the administration of HES 450 was associated with less edema, postoperative nausea, vomiting, and antiemetic use than the administration of LR solution.
Superior gut function in patients who receive a combined crystalloid and colloid fluid regimen for intraoperative volume resuscitation might be explained by the presence of less intestinal edema than in patients who receive crystalloids alone. More severe periorbital edema was observed after the administration of LR than after intraoperative HES administration in patients who underwent major abdominal surgery. It seems likely that edema may also occur in the gastrointestinal tract and that this may influence gut function in patients undergoing gastrointestinal and nongastrointestinal surgery. Indeed, more intestinal edema was seen in patients undergoing a Whipple operation who received LR solution rather than HES 450/NS or 20% albumin/NS for intraoperative fluid replacement.
Several other trials have compared restrictive and liberal fluid or sodium regimens. The results are not uniform, and comparison is difficult because administered volumes and electrolytes in both arms differed substantially, which reflects nonuniform standard practice. Therefore it has been suggested that “future studies should focus on the effects of individualized ‘goal-directed’ fluid administration strategies rather than fixed fluid amounts on postoperative outcome.”
Goal-Directed Fluid Therapy
Goal-directed fluid therapy (GDFT) is a term that refers to the use of an algorithmic approach to fluid management, whereby tissue oxygenation and intravascular volume status is optimized by assessing and responding to each patient’s individual hemodynamic response to fluid boluses. GDFT has been associated with improved outcomes after moderate to major surgery, with shorter hospital stays, fewer ICU admissions, and earlier return of bowel function ( Table 27-3 ). Most studies use crystalloids and colloids in combination, with background crystalloid infusions to replace extracellular losses augmented by colloid boluses to maintain central euvolemia. A recent meta-analysis suggests that this approach reduces morbidity and mortality rates for high-risk surgical patients.
Related posts:
Is a Preoperative Screening Clinic Cost-Effective?
Which Patient Should Have a Preoperative Cardiac Evaluation (Stress Test)?
What Is the Optimal Airway Management in Patients Undergoing Gastrointestinal Endoscopy?
What Works in a Patient with Acute Respiratory Distress Syndrome?
Are There Special Techniques in Obese Patients?
Is Propofol Safe If Given by Nonanesthesia Providers?
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
