Approaches to Fluid Management
The “Classic” Approach to Fluid Management
The “Classic” (read: outdated) approach to management of fluids in the perioperative setting involved trying to predict the amount of fluids needed based on a the duration and severity of a particular operation and empirically replacing fluids based on these estimates. IT IS PRESENTED HERE FOR HISTORICAL INTEREST ONLY AND IS NOT RECOMMENDED.
Calculating Fluid Requirements
Fluids must be given based on an estimation of the following – fluid losses prior to start of anesthesia, maintenance requirements, normal fluid losses that occur during surgery, and response to unanticipated fluid (blood) loss. Furthermore, consider titrating fluid requirements to physiologic measures (ex. CVP, urine output).
Step 1: Calculate Preoperative Fluid Losses
Simply multiply the maintenance fluid requirements (cc/hr) times the amount of time since the patient took PO intake. Estimated maintenance requirements follow the 4/2/1 rule: 4 cc/kg/hr for the first 10 kg, 2 cc/kg/hr for the second 10 kg, and 1 cc/kg/hr for every kg above 20.
Step 2: Calculate Ongoing Maintenance Requirements
Based on patient’s weight, using the same 4/2/1 rule as used to calculate preoperative maintenance requirements.
Step 3: Calculate Anticipated Surgical Fluid Losses
Based on patient’s weight and anticipated tissue trauma. A rough guide can be found in Stoelting:
- Minimal tissue trauma (ex. herniorrhaphy): 2-4 cc/kg/hr
- Moderate tissue trauma (ex. cholecystectomy): 4-6 cc/kg/hr
- Severe tissue trauma (ex. bowel resection): 6-8 cc/kg/hr
Step 4: Adjust for Unanticipated Fluid Losses
A common recommendation is to give 3 cc of crystalloid for every 1 cc of blood loss. Remember to add up lap pads (100-150 cc each) and 4x4s (10 cc each).
Danger Associated with Arbitrary Fluid Administration
The perils of giving critically-ill patients arbitrary amounts of fluid without advanced monitoring (see section on modern fluid management below) was recently hilighted by the FEAST Trial, which included 3141 febrile pediatric patients with impaired perfusion (defined as capillary refill > 3 seconds, a lower-limb temperature gradient, “weak” radial-pulse volume, or severe tachycardia [defined based on age]) and randomized them to 20-40 mL/kg of normal saline, albumin, or no bolus on hospital admission. The primary outcome, death at 48 hours mortality was 10.6%, 10.5%, and 7.3% in the albumin-bolus, saline-bolus, and control groups, respectively. Mortality at 4 weeks was 12.2%, 12.0%, and 8.7%, respectively [Maitland K. NEJM 364: 2483, 2011]
]. Criticisms of this study include a protocol change (which increased the amount of the fluid boluses) midway through the study, lack of control or documentation of fluid management after the first hour, unavailability of monitoring data, and lack of advanced hemodynamic monitoring. However, the 20-40 mL/kg bolus was thought to be relatively conservative in the presence of what appeared to be septic shock and by any account these results are surprising. For comparison, in adults, the Surviving Sepsis Guidelines (2008 version) recommend “Give fluid challenges of 1000 mL of crystalloids or 300–500 mL of colloids over 30 mins. More rapid and larger volumes may be required in sepsis-induced tissue hypoperfusion” (Level of Evidence 1D) [Dellinger R. CCM 36: 296, 2008]
Modern Fluid Management
The modern approach to fluid management is based on the concept of goal-directed therapy (GDT), in which it is believed that interventions should be performed specifically to affect a meaningful clinical variable. The reality is that fluids can be harmful, and should only be given when they are expected to produce some benefit. Management of fluids such that stroke volume is optimized is an extremely well-validated approach that has been shown repeatedly to reduce morbidity [Hamilton MA et al. Anesth Analg 112: 1392, 2011; Gurgel ST and do Nascimento P Jr. Anesth Analg 112: 1384, 2011]. In fact, esophageal Doppler monitoring (EDM) was recently endorsed by the National Health Service as a rational alternative to central venous pressure monitoring in patients undergoing major surgery. A promising alternative to EDM is optimization of respiratory variation, although it is not as well validated – three recent prospective, randomized, controlled trials have suggested that optimization of respiratory variation may have the potential to improve outcomes [Lopes R et al. Crit Care 2007; 11: R100; Benes J et al. Crit Care 14: R118, 2010; Forget P et al. Anesth Analg 111: 910, 2010], although it will take time to accumulate the quantity and quality of data that currently support of EDM. There is essentially no role for “maintenance” IV fluids in modern fluid management – rather, fluids are given as targeted boluses when they are expected to lead to a hemodynamic improvement.
Liberal vs. Restricted Therapy
Data Favoring “Restrictive” Perioperative Fluids
Brandstrup et. al.
In a randomized, observer-blinded multicenter study, Brandstrup et al. compared a liberal vs. restrictive fluid strategy in 172 patients undergoing colorectal surgery. The liberal patients received 500 cc of 6% HAES and 500 cc NS loading, followed by NS at 7 cc/kg/h for one hour, then 5 cc/kg/hr for two hours, then 3 cc/kg/hr afterwards, with 500 cc blood loss replaced by NS, 500-1500 cc EBL replaced with 6% HAES, and over 1500 cc replaced with blood components. The restrictive group, by contrast, received only 500 cc of D5W (minus whatever oral intake occurred during fasting) and volume to volume blood loss with 6% HAES up to 1500 cc EBL. Total IV fluids average 5.4 L for the liberal group and 2.7 L for the restrictive group. The restrictive regimen appeared to reduce the incidence of major and minor complications (ex. anastomotic leakage, pulmonary edema, pneumonia, and wound infection). More specifically, the numbers of both cardiopulmonary (7% versus 24%, P = 0.007) and tissue-healing complications (16% versus 31%, P = 0.04) were significantly reduced. No patients died in the restricted group compared with 4 deaths in the standard group (0% versus 4.7%, P = 0.12). Despite a perioperative decrease in urine output, acute renal failure did not occur in any patient. Unfortunately, Brandstrup’s data was confounded by the introduction of colloids, as colloids were predominantly given to the restrictive group and the liberal group received > 5 L crystalloids. [Brandstrup B et. al. Ann Surg 238: 641, 2003]
Nisanevich et. al.
Nisanevich et al. randomized 152 patients undergoing various abdominal procedures to liberal (10 cc/kg bolus followed by 12 cc/kg/hr) vs. restrictive (4 cc/kg/hr) of lactated ringers. They found decreased postoperative morbidity (including improved GI recovery and a shortened hospital stay), under a protocol-based, more restrictive fluid therapy (1.2 L vs. 3.7 L). [Nisanevich V et. al. Anesthesiology 103: 25, 2005]
On “Liberal” Perioperative Fluids
Some authors have suggested that liberal fluids improve PONV and tissue oxygenation. Upon reviewing the data, Chappel et. al. state that “These data, despite being inconsistent, indicate that higher fluid amounts might reduce the risk of PONV and increase postoperative lung function after short operations. Nevertheless, most studies considered only one outcome parameter; therefore, the overall effect on the patient is hard to gauge, because other, potentially more serious parameters may be impacted adversely by the same treatment. These results seem interesting regarding certain collectives, e.g., outpatients during minor surgery, but they cannot account for larger surgery over several hours. Current evidence suggests that liberal fluid is a good idea where major trauma and fluid shifting are unlikely, but more careful fluid management may be beneficial in more stressful operations…” [Chappel D et. al. Anesthesiology 109, 723: 2008]
Chappel’s Synthesis of the Available Data
“Because of a total lack of standardization, the available data do not allow evidence-based recommendations on practical perioperative fluid management… Any perioperative fluid handling seems to be justified. However, this is in clear contrast to daily clinical observations during surgery, suggesting that our various surgical and anesthesiologic standard treatments might contribute to important perioperative problems“. [Chappel D et. al. Anesthesiology 109, 723: 2008]
Reconciliation of Available Data (using the Modern Approach)
There are many problems with the fluid management studies of the last century or so, many pointed out by Chappel (see above). Because the vast majority of studies were based on the teleologically flawed, “classical” approach of trying to predict fluid needs, rather than actually measuring fluid needs, it is no wonder that the data are confusing at best. A lack of standardization with regards to what is “restrictive” and what is “liberal” further complicates the data.
The reality is that fluids can be harmful, and should only be given when they are expected to produce some benefit. Management of fluids such that stroke volume is optimized is an extremely well-validated approach that has been shown repeatedly to reduce morbidity [Hamilton MA et al. Anesth Analg 112: 1392, 2011; Gurgel ST and do Nascimento P Jr. Anesth Analg 112: 1384, 2011]. In fact, esophageal Doppler monitoring (EDM) was recently endorsed by the National Health Service as a rational alternative to central venous pressure monitoring in patients undergoing major surgery. A promising alternative to EDM is optimization of respiratory variation, although it is not as well validated.
Types of Fluids
Normal saline can cause a hyperchloremic metabolic acidosis, whereas lactated ringer’s can cause a metabolic alkalosis secondary to metabolism of lactate (which produces bicarbonate). Never use LR with blood products as the calcium will bind to the citrate. Dextrose-containing solutions should be avoided in patients with neurologic injuries as they may cause hyperglycemia, cerebral acidosis, and an osmotic diuresis [Stoelting et. al. Basics of Anesthesia, 5th ed. Elsevier – China, p. 351, 2007]. It is well established that hypotonic fluids cause brain edema (thus do not use Lactated ringer’s for large volume resuscitation), although animal studies suggest that crystalloids increase cerebral edema and ICP only when they result in hypoosmolality. [Crit Care Med 16: 862, 1988; Anesthesiology 67: 936, 1987]
Half-life of albumin is 16 hours. Albumin has been the favored colloid in neurosurgical patients as there are several reports suggesting it may have a beneficial effect on cerebral edema and ICP [Arch Neurol Psych 39: 1277 and 1288, 1938; Lancet 2: 557, 1941; Mayo Clin Proc 213: 89, 1948] as well as potentially beneficial effects on microcirculatory flow [NEJM 339: 321, 1998]. Importantly, subgroup analysis of the SAFE trial (see below) suggested a higher mortality rate associated with albumin (as compared to saline) [Myburgh J et al. NEJM 357: 874, 2007; FREE Full-text at New England Journal of Medicine]
Hydroxyethyl Starch (HES) is a synthetic colloid derived from amylopectin (a polysaccharide extracted from maize). The molecular structure of HES is similar to that of glycogen, and it is usually hydrolyzed by α-amylase (the same enzyme responsible for degrading glycogen). Various chemical modifications are employed to delay hydrolysis, for instance hydroxyethylation of the C2/C6 atoms of the glucose subunits of the starch. The C2/C6 ratio describes the amount of C2 vs. C6 hydroxyethylation substitutions of the glucose subunits but is not reported in the name. The average molecular weight and number of hydroxyethyl substitutions per glucose subunit are reported. Higher molecular weight, higher molar substitution, and higher C2/C6 ratios all theoretically lead to delayed hydrolysis in plasma and increased intravascular time. In reality, if leakage cannot be prevented, this may lead to extravasation and prolonged tissue edema as accumulated in the extravascular space.
Hydroxyethyl starch may be mixed in saline [e.g. Hespan, 6% MW hetastarch] or in balanced salt solution [e.g. Hextend, 6% MW hetastarch]. 90% of hydroxylethyl starch particles last 17 days. Dextran comes in dextran 40 and dextran 70. Larger particles have a half-life in the order of days, thus dextran 70 is generally used for volume resuscitation, while dextran 40 is used to improve blood flow to the microcirculation. Hypersensitivity reactions to colloids are possible, but rare. Note that the dextrans can reduce platelet aggregation and adhesiveness, and that hydroxyethyl starch can reduce factor VIII and vWF, as mentioned in Barron’s review of 113 studies (which stated “Artificial colloid administration was consistently associated with coagulopathy and clinical bleeding, most frequently in cardiac surgery patients receiving hydroxyethyl starch”) [Barron ME et. al. Arch Surg 139: 552, 2004]. All colloids share the following potential downsides – volume overload, coagulopathy (especially Hetastarch), anaphylactoid reactions, and interstitial edema).
Colloids vs. Crystalloids
Important Randomized Controlled Trials
A small (126 patients), prospective randomized controlled trial in patients with cirrhosis and spontaneous bacterial peritonitis did show a significant mortality benefit of albumin administration [Sort P et al. NEJM 341: 403, 1999; FREE Full-text at New England Journal of Medicine]. In lieu of additional data to refute this study, colloid administration in this patient population seems justified. Note that the benefits of albumin in this patient population (suppression of plasma renin activity, expansion of central blood volume) may not be translatable to other colloids [Fernández J et al. Hepatology 42: 627, 2005]
More recently, three large, prospective, randomized, controlled trials have attempted to determine the impact of colloids on mortality in critically-ill patients. In the Australian SAFE trial, which randomized 6997 critically ill patients to albumin or saline, there was no difference in mortality [Finfer S et al. NEJM 350: 2247, 2004 ; FREE Full-text at NEJM]. The VISEP trial, a two-by-two factorial trial comparing strict to conventional glucose control as well as lactated ringer’s to low-molecular-weight hydroxyethyl starch (HES, using a normal saline carrier) in severe sepsis, randomized 537 patients prior to being stopped early [Brunkhorst FM et al. NEJM 358: 125, 2008; FREE Full-text at New England Journal of Medicine]. As compared to Ringer’s lactate, HES in a normal saline carrier was associated with higher rates of acute renal failure and renal-replacement therapy in this German study. The even more recent 6s Trial randomized 804 critically-ill patients to Ringer’s acetate versus HES 130/0.42 in a balanced electrolyte solution and found increased rate of death and renal replacement therapy in the colloid group [Perner A et al. NEJM 367: 124, 2012; FREE Full-text at NEJM]. The even more recent CHEST trial, which randomized 7000 critically-ill patients to 130/0.4 tetrastarch in normal saline (Voluven) versus normal saline found a trend towards increased mortality in the tetrastarch group (relative risk 1.06, p = 0.26) but a statistically significant increase in the need for renal replacement therapy (relative risk 1.21, p = 0.04) [Myburgh JA et al. NEJM 2012 (Epub ahead of print); FREE Full-text at NEJM]. Based on these data it is difficult to defend the use of colloids in critically ill patients that do not have cirrhosis and concomitant spontaneous bacterial peritonitis
While the ratio of crystalloids to colloids is classically described as 3:1, this is based on animal data and may not be applicable to critically ill humans. Based on data from the SAFE Trial (NS vs. albumin in 6997 patients), the ratio was 1.3:1. A smaller study of 383 children with dengue shock suggested the ratio was 1.6:1. According to a review by Hartog et al., four additional randomized controlled trials report ratios ranging from 1.6-2.1:1 [Hartog CS et al. Anesth Analg 112: 156, 2011]. The recently completed 6s Trial which compared Ringer’s acetate to 130/0.42 hydroxyethyl starch in 804 patients found no difference in the absolute volumes administered, with more fluid administered in the colloid group when corrected for body weight
Sum of Available Data
A review of all randomized trials comparing the two found no difference between crystalloids and colloids [AIM 151: 901, 1991]. A more recent review of randomized studies suggested that colloids may actually increase mortality in trauma patients [Crit Care Med 27: 200, 1999]. An illuminating review by Hartog, Bauer, and Reinhart suggests why – the purported “long term” expansion caused by colloids is a myth (in reality, colloid expansion lasts only a few hours, after which the colloids begin to accumulate extravascularly, staying there for weeks), the risk of edema is no different between colloids and crystalloids, the true equivolume ratio is probably 1:2 or even as low as 1:1.6 [Hartog CS et al. Anesth Analg 112: 156, 2011].
Article of the Month for January 21001 with Dr. Konrad Reinhart, co-author of the article The Efficacy and Safety of Colloid Resuscitation in the Critically Ill.
Lastly, according to a Cochrane Database Review, “There is no evidence from RCTs that resuscitation with colloids reduces the risk of death, compared to resuscitation with crystalloids, in patients with trauma, burns or following surgery. As colloids are not associated with an improvement in survival, and as they are more expensive than crystalloids, it is hard to see how their continued use in these patients can be justified outside the context of RCTs” [Perel P et. al. Cochrane Database Systemic Review 4: CD000567, 2007]
Several experimental studies and animal reports suggested that HTS would be beneficial in terms of cerebral edema and ICP, however initial studies of head injury patients [Arch Surg 126: 1065, 1991] and multitrauma patients [Ann Surg 213: 482, 1991] were disappointing. More recent studies have been more promising.
Prospective, Randomized Studies of HTS
Crit Care Med 26:1265, 1998
Severely head injured children (n=32), 1.7% HTS vs. LR
HTS more effective than LR for reducing ICP. Shorter ventilation, decreased ICU stay Crit Care Med 31: 1683, 2003
Comatose TBI (n=20), 7.5% HSS vs. 20% mannitol, 2 mL/kg
Fewer episodes of intracranial hypertension per day (6.9 vs. 13.3), lower daily duration of elevated ICP (67 vs. 131 min, p < 0.01), lower rate of clinical failure (10% vs. 70%, p < .01) Crit Care Med 33: 196, 2005
Patients: ICP > 20 mm Hg (n=9), 7.5% saline and 6% dextran-70 solution (HSD) vs. 20% mannitol
ICP lowered slightly more with HSD (13 vs. 7.5 mm Hg, p = 0.044) and HSD had a longer duration (p = .044) Crit Care 9: R530, 2005
At risk of increased ICP (n=40), 7.2% NaCl/HES 200/0.5 vs. 15% mannitol
HSS worked 2.7 min faster (p < 0.0002), caused a greater decrease in ICP (57% vs 48%; p < 0.01), increased CPP by 12 mm Hg vs. 9 mm Hg (p < 0.0001). No clinically relevant effects on electrolyte concentrations and serum osmolarity .
At least one methodologically flaw
J Trauma 44: 50, 1998
Comatose TBI (n=34), 1.6% HTS vs. LR for SBP < 90 mm Hg
No difference in ICP at any point AFTER administration (note – HTS had initial ICP of 21, vs. 18 in LR cohort). This study was flawed because presenting ICP were vastly different.
JAMA 291: 1350, 2004
PRE-hospital resuscitation of comatose TBI (n=229), 7.5% vs. LR (only 250 cc given)
No difference in mortality or neurologic outcome. Note that this study may be flawed because no endpoints were defined, and only 250 cc were given on top of standard therapy.
Neurosurgery 57: 727, 2005 (Retrospective)
This retrospective review of 13 adult TBI patients with (mannitol vs. 23.4% saline) showed no difference in ICP reduction (p = 0.174) but the duration for hypertonic saline was longer (96 vs. 59 min, p= 0.016) [Neurosurgery 57: 727, 2005]
Hypertonic Saline in Neurosurgical Patients
Hypertonic Saline in Neurosurgical Patients
- Multiple small studies (10-40 patients) , some of them randomized and double-blind, show that hypertonic saline lowers ICP faster , to a greater extent, and for a longer duration than mannitol
- Larger clinical trials have failed to show an improvement in outcome but have been poorly designed. Some use “pseudo”-hypertonic saline (< 7.5%), others don’t use enough or have no physiologic endpoints, and others did not appropriately match the treatment and control groups.
Below Hgb of 7 mg/dL, cardiac output has to increase substantially in order to maintain DO2, thus maintain Hgb > 7.0 (and consider 10.0 in patients with a cardiac history) [Stoelting et. al. Basics of Anesthesia, 5th ed. Elsevier – China, p. 352, 2007]. Beware dilutional thrombocytopenia as it is the most common intraoperative coagulopathy (factor deficiency is rare in the absence of liver disease).
3rd Space: Does it Exist?
From Chappel D et. al. – “In summary, a classic third space was never localized and only “quantified” with one specific method using certain conditions regarding sampling and equilibration times, implying serious concerns and weaknesses. All other methods using various tracers, multiple sampling techniques, longer equilibration times, or analysis of kinetics contradict the existence of a fluid-consuming third space. Taking all this into account, we have to conclude that a classic third space per se quantitatively does not exist. It is currently not more than an ill-defined compartment thought to reflect an otherwise unexplainable perioperative fluid shift. Therefore, we suggest abolishing this mystery and sticking to the given facts: Fluid is perioperatively shifted within the functional extracellular compartment, from the intravascular toward the interstitial space“. [Chappel D et. al. Anesthesiology 109, 723: 2008]