CURRENT
OPINION Albumin in critically ill patients: the ideal colloid?
Pietro Caironi a,b, Thomas Langer b, and Luciano Gattinoni a,b
Purpose of review
The use of albumin-containing solutions in critically ill patients has been recently revisited, following
evidence on harmful effects of synthetic colloids, and novel randomized controlled trials (RCTs) in sepsis.
Here, we review the most recent findings on albumin administration in acutely ill and septic patients.
Recent findings
The revision of Starling’s theory on microvascular fluid dynamics has highlighted the role of albumin in
preserving intravascular compartment volume. In cirrhosis, albumin may be important in maintaining
immune system reactivity and cardiac contractility. Preliminary analyses indicate albumin as beneficial in
patients with burn, while being associated with increased risk of acute kidney injury after cardiac surgery.
The first RCT (ALBIOS trial) testing the efficacy of albumin replacement in severe sepsis did not show a
survival benefit associated with albumin, although observing at post-hoc analysis a benefit in septic shock.
All the eight meta-analyses performed on albumin in sepsis reveal an absence of harm, and likely a benefit
lower than expected, whereas suggesting an advantage, to be verified, in septic shock.
Summary
Further studies are needed to clarify physiology and clinical impact of albumin in critically ill patients,
considering specific phenotypes and secondary outcomes other than survival, yet clinically relevant.
Keywords
albumin, critically ill, fluids, sepsis
In humans, albumin is the colloid molecule most albumin in severe sepsis [11,12 ]. In the current
representative in the extracellular space, being cru- manuscript, we would like to review the recent
cial in regulating the physiology of microvascular findings related to the physiological rationale of
fluid dynamics [1]. Moreover, albumin, being one of using albumin in critically ill patients, and the last
the molecules synthesized by the liver with the evidence available on its clinical impact.
highest energy requirement, acts as an important
player of several secondary functions, including
ALBUMIN AND VASCULAR COLLOID AND
transportation of endogenous and exogenous mol-
OSMOTIC PRESSURE
ecules, antioxidant and anti-inflammatory proper-
ties [2,3]. This is why its use as intravenous solution Oncotic properties characterizing the intravascular
has been widely applied to critically ill patients, space depend on the concentration of albumin in the
until the publication of a highly ranking meta- capillaries, in which the membrane separating intra-
analysis in 1998 has warned against its use [4]. As vascular and extravascular spaces is impermeable to
the history of medicine goes on, and more solid
evidence becomes available, the publication of the a
Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti, Fon-
SAFE trial has finally clarified its potential disadvan- dazione IRCCS Ca’ Granda – Ospedale Maggiore Policlinico, Università
tages [5] and advantages [6] in specific categories of degli Studi di Milano and bDipartimento di Anestesia, Rianimazione ed
Emergenza Urgenza, Fondazione IRCCS Ca’ Granda – Ospedale
critically ill patients, while assuring its safety in the
Maggiore Policlinico, Milan, Italy
heterogeneous population [7]. Recently, an increas-
Correspondence to Pietro Caironi, MD, Dipartimento di Fisiopatologia
ing attention to the use of albumin-containing Medico-Chirurgica e dei Trapianti, Fondazione IRCCS Ca’ Granda –
solutions in patients admitted to intensive care Ospedale Maggiore Policlinico, Università degli Studi di Milano, Via F.
units (ICUs) has resumed, mainly for two reasons: Sforza 35, 20122 Milano, Italy. Tel: +(02) 5503-3232/3231; fax: +(02)
first, the evidences that most of the available syn- 55033230; e-mail: pietro.caironi@unimi.it
thetic colloids may be harmful in critically ill Curr Opin Crit Care 2015, 21:302–308
patients [8–10]; second, the conclusions of large DOI:10.1097/MCC.0000000000000223
meability nullifies the oncotic properties of albumin Two novel secondary functions have been
at different extent. Therefore, as suggested originally recently investigated. O’Brien et al. [25 ], in an
&
1070-5295 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved. www.co-criticalcare.com 303
extensive translational study, investigated the role Schematically, based upon their electrical character-
of the cyclooxygenase-derived eicosanoid prosta- istics, colloid molecules may be either characterized
glandin E2 (PGE2) as a mediator of the immunosup- by a neutral activity, such as starches and dextrans,
pressive state characterizing advanced cirrhosis. In or by surface negative charges, such as gelatin
&&
their study, the authors observed that plasma of and albumin [30 ]. Consequently, starches and
patients with acute decompensated liver cirrhosis dextrans do not interfere with the electrochemical
had elevated circulating levels of PGE2, which are balance, and solutions containing starches and dex-
able to suppress the pro-inflammatory and the trans follow simple rules described for crystalloid
bacteria-killing capacity of macrophages. Albumin solutions. On the contrary, gelatins and albumin
administration, by reducing circulating PGE2 levels, have surface negative charges. Within the frame-
significantly improved the plasma-induced impair- work of Stewart’s approach, these molecules belong
ment of macrophage pro-inflammatory actions, to the category of weak acids, and therefore have per
&&
indicating that albumin-containing solutions may se an acidifying effect [30 ]. Commercially available
attenuate the PGE2-mediated immunosuppression albumin solutions differ in electrolyte composition,
and the associated risk of secondary infections. In mainly due to the specific purification process
another study, Bortoluzzi et al. [26] elucidated the applied for human plasma and the preparation of
effects of albumin on cardiac contractility in an solutions with different albumin concentration
experimental model of decompensated cirrhosis. (mainly 4%, 5% and 20%), and, consequently, differ
The investigators observed that after the develop- from endogenous albumin, in terms of half-life and
ment of cirrhosis, cardiac contractility was impaired antioxidant properties [31]. Most of studies per-
in association with an increased inducible nitric- formed describe, in fact, the development of mild
oxide synthase (iNOS) and TNF alpha protein metabolic acidosis after the infusion of albumin
&
expression. After administration of albumin, but preparations [32,33 ], although this effect is prim-
not of synthetic colloids, these cellular alterations, arily because of the specific electrolytic composition
in parallel with cardiac functionality, returned to (i.e. the SID of the solvent), rather than the effect of
normal levels, suggesting a potential for albumin the albumin molecule [34].
acting as a positive inotropic agent by contrasting
the effects of iNOS and TNF alpha-pathways.
Although these findings relate to cirrhotic patients, ALBUMIN IN CRITICALLY ILL PATIENTS
their transferability to acutely critically ill patients Since the late 1940s [35], intravenous solutions
may open possible novel therapeutic options of containing albumin have been extensively admin-
great potential. istered to critically ill patients, despite conflicting
&&
results on its effect on survival [36 ].
&
Navickis et al. [37 ] performed a meta-analysis
ALBUMIN-CONTAINING SOLUTIONS AND on four randomized and four nonrandomized stud-
EFFECTS ON ACID-BASE EQUILIBRIUM ies on burn shock resuscitation with albumin
According to Stewart’s approach to acid-base equi- solutions. After the exclusion of two studies at high
librium, pH is regulated by the strong ion difference risk of bias, the authors observed a significant
(SID), defined as the difference between strong reduction in mortality and in incidence of abdomi-
cations and anions, the partial pressure of carbon nal compartment syndrome in patients receiving
dioxide and the concentration of nonvolatile weak albumin. Despite such promising findings, the pauc-
acids (mainly albumin and phosphates) [27]. Infu- ity of available data and the poor quality of RCTs
sion of crystalloid solutions affects plasma pH warrant confirmatory studies. Similarly, in another
&
according to the difference between plasma meta-analysis, Kitsios et al. [38 ] investigated the
bicarbonate concentration (HCO3) and the SID effect of the combined administration of albumin
of the infused crystalloid (SIDINF) [28,29]. If SIDINF and furosemide to overcome diuretic resistance in
is greater than baseline HCO3 concentration, pH patients characterized by hypoalbuminemia. The
will increase (generating alkalosis). If SIDINF is lower authors observed only a minor and transient
than baseline HCO3 concentration, pH will increase in urinary output and sodium excretion
decrease (generating acidosis). If SIDINF approxi- rate in patients receiving the co-administration of
mates the baseline concentration of HCO3, pH will albumin and furosemide, as compared with those
eventually remain constant, regardless of the receiving only loop-diuretics, whereas intermediate
amount of the infusion, and the degree of dilution outcomes did not differ between the two strategies.
&& &
[30 ]. Uhlig et al. [39 ] performed a meta-analysis to
Colloid solutions are, de facto, crystalloid evaluate the impact on mortality and lung function
solutions containing oncotic macromolecules. of the use of either crystalloid or albumin-
containing solutions in adult acute respiratory dis- exception of the time to suspension of vasoactive
tress syndrome (ARDS). According to their analysis, agents, being shorter in the albumin as compared
pooled risk of death was equal between the two with the crystalloid group. In contrast, in a post-hoc
groups, whereas oxygenation appeared to be subgroup analysis on patients with shock at the time
superior in patients receiving albumin. The authors of enrolment, we observed a significant reduction in
interpret this finding as the result of a reduced 90-day mortality rate in the albumin group as com-
alveolar-capillary leakage, although this effect did pared with the crystalloid group (N ¼ 1121, 43.6 vs.
not translate in an improvement of survival. Lastly, 49.9%, P ¼ 0.03). Taken together, these findings
&&
Frenette et al. [40 ] performed a retrospective showed that albumin supplementation in addition
cohort study including patients undergoing on- to crystalloids to correct hypoalbuminemia does not
pump cardiac surgery to assess the risk of acute improve 90-day survival in patients with severe sep-
kidney injury (AKI) associated with either albumin sis, as compared with the use of crystalloids alone. In
or starches administration. The authors found albu- contrast, they suggest a possible benefit in the sub-
min-containing solutions to be associated with a group of patients with septic shock. As noted by
&
dose-dependent risk of AKI development. In the Flannery et al. [43 ], the trial presented certain limita-
context of previous observational [41,42] and recent tions, mainly relying on the un-predefined nature of
&&
findings observed in the ALBIOS trial [12 ], with no the post-hoc analysis performed, and the open-label
evidence of an increased risk of AKI even with the design. Nonetheless, it is worth mentioning that the
use of hyperoncotic albumin solutions, this issue significant reduction of 90-day mortality in patients
merits further investigations. with shock was still confirmed even after adjustments
&
for clinically relevant variables [44 ].
1070-5295 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved. www.co-criticalcare.com 305
Number of Types of
Year of studies included studies
Authors publication Inclusion criteria (sample size) Comparison included Primary outcomes Results; OR or RR (95% CI)
Delenay et al. 2011 [45] Sepsis of any severity 17 (N ¼ 1977) Albumin vs. RCTs All-cause mortality at the Sepsis: OR 0.82 (0.67–1.00),
Intravenous fluids
www.co-criticalcare.com
crystalloid longest follow-up (0.79–1.02), P ¼ 0.11
or colloid available
&
Wiedermann et al. 2014 [48 ] Severe sepsis (adults) 3 (N ¼ 3791) Albumin vs. Large-scale All-cause mortality at the Severe sepsis: RR 0.92
crystalloid RCTs longest follow-up (0.84–1.00), P ¼ 0.046
available
&&
Rochwerg et al. 2014 [49 ] Severe sepsis (adults). 14 (N ¼ 18916) Any fluid RCTs All-cause mortality at the Severe sepsis: albumin vs.
Predefined subgroup strategy longest follow-up crystalloids, OR 0.83
and network meta-analysis compared available (0.65–1.04)
to a different
fluid strategy
&&
Patel et al. 2014 [50 ] Sepsis of any severity 16 (N ¼ 4190) Albumin vs. RCTs All-cause mortality at the Sepsis: RR 0.94 (0.87–1.01),
(adults). Predefined crystalloid longest follow-up P ¼ 0.11; severe sepsis
subgroups or colloid available without shock: RR 0.95
(0.85–1.06), P ¼ 0.35
(N ¼ 2070); septic shock:
RR 0.92 (0.83–1.02),
P ¼ 0.10 (N ¼ 1962)
&
Jiang et al. 2014 [51 ] Sepsis of any severity 15 (N ¼ 6998) Albumin vs. RCTs All-cause mortality at the Sepsis: RR 0.94 (0.87–1.02),
(adults and pediatrics). crystalloid longest follow-up P ¼ 0.15; severe sepsis without
Predefined subgroups or colloid available shock: RR 0.95 (0.85–1.07),
P ¼ n.s; septic shock: RR 0.89
(0.80–0.99), P ¼ 0.04
&
Xu et al. 2014 [52 ] Severe sepsis (adults). 6 (N ¼ 3658) Albumin vs. RCTs and All-cause mortality Severe sepsis: OR 0.88
Predefined subgroup crystalloid parallel trials (including 28-, 90-day (0.76–1.01), P ¼ 0.08; shock:
mortality, or at other OR 0.81 (0.67–0.97), P ¼ 0.03
time points)
of studies included vary between meta-analyses Is therefore albumin the ideal colloid for crit-
&& &
[50 ,52 ]. ically ill patients? With the exception of patients
What can we derive from these data? Evidence with traumatic brain injury [5], ideally, the answer
gathered so far indicates that in a general population may be ‘yes,’ but, in practice, we have no solid
of patients with severe sepsis, the administration of clinical evidence for its cost-effective benefit, and,
albumin-containing solutions is not associated more importantly, many questions to be answered.
with a significant reduction in mortality rate Without being religiously believers of the benefit of
(at least until the generally accepted follow-up time albumin administration in acutely ill patients, or
of 90 days), as suggested by the two large concluded those with severe sepsis, we need to keep on study-
&&
RCTs [11,12 ]. Of note, the analyses performed ing its physiological mechanisms of action, in the
have always observed a relative risk variation in attempt to individuate subgroups of patients, which
favor of the treatment with albumin, although may benefit, or may even receive harm, from this
not achieving (but being close to) a statistical sig- treatment. Based upon simple notions of physiology
nificance. Such consistency of findings suggests, and the clinical evidence available, the premises are
with a relatively high degree of certainty, that albu- still promising.
min administration to these patients does not cause
harm, and that the beneficial effects related to it, if Acknowledgements
any, may be quantitatively smaller than expected. We would like to thank Eleonora Carlesso, MSc, for her
At the same time, the same evidence indicate that assistance in the preparation of the present manuscript.
there might be a beneficial effect related to albumin
in the subgroup of patients with septic shock, as Financial support and sponsorship
observed also in the post-hoc analysis from the This work was partially supported by funding from the
&& &
ALBIOS trial [12 ,44 ], and that this hypothesis Italian Medicines Agency (AIFA, grant FARM6JS3R5,
merits further confirmations. Such considerations 2006) and from the Italian Ministry of Health (Ricerca
gain even more importance when evaluating the Finalizzata, grant RF-2011–02348358).
potential cost-effectiveness of employing albumin-
containing solutions in critically ill patients, and Conflicts of interest
patients with severe sepsis or septic shock, because
P.C. has received honoraria for lectures from Grifols, CLS
of their relative high cost. In fact, although as esti-
Behring, B. Braun and Baxter. T.L. reported no conflicts
mate, recent cost-effectiveness analyses have
of interest. L.G. is on the Advisory Board of Grifols and
suggested that albumin administration may be the
has received speaker’s honoraria from KCI, B. Braun,
most cost-effective treatment, as compared with
Baxter, Grifols, and Kedrion.
other colloids or crystalloids, in septic patients,
when considering all the medical costs implied in
patient clinical treatment [53]. REFERENCES AND RECOMMENDED
READING
Papers of particular interest, published within the annual period of review, have
CONCLUSION been highlighted as:
& of special interest
It is crucial to consider three further aspects, which && of outstanding interest
1070-5295 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved. www.co-criticalcare.com 307
11. Charpentier J, Mira JP. Efficacy and tolerance of hyperoncotic albumin 36. Vincent JL, Russell JA, Jacob M, et al. Albumin administration in the acutely ill:
administration in septic shock patients: the EARSS study. Intensive Care && what is new and where next? Crit Care 2014; 18:231.
Med 2011; 37 (Suppl 2):S115. Extensive review on albumin administraton in critically ill patients, based upon the
12. Caironi P, Tognoni G, Masson S, et al. Albumin replacement in patients with evidence available and discussing the possible next steps of research on this matter.
&& severe sepsis or septic shock. N Engl J Med 2014; 370:1412–1421. 37. Navickis RJ, Greenhalgh DG, Wilkes MM. Albumin in burn shock resuscita-
This is the first large RCT investigating the effect of albumin replacement in & tion: a meta-analysis of controlled clinical studies. J Burn Care Res 2014.
addition to crystalloids as compared with crystalloids alone in 1818 patients with [Epub ahead of print]
severe sepsis or septic shock. This is the meta-analysis on the effects of albumin-containing solutions in burn
13. Peters T. Metabolism: Albumin in the body. All about albumin. San Diego, patients.
California: Academic Press Limited; 1996; 188–250. 38. Kitsios GD, Mascari P, Ettunsi R, Gray AW. Co-administration of furosemide
14. West JB. The original presentation of Boyle’s law. J Appl Physiol (1985) & with albumin for overcoming diuretic resistance in patients with hypoalbumi-
1999; 87:1543–1545. nemia: a meta-analysis. J Crit Care 2014; 29:253–259.
15. Ganong W. Dynamics of blood and lymph flow. Review of medical physiology. This is the meta-analysis on the combined effect of albumin and furosemide in
Connecticut: Appleton and Lange; 1995. pp. 525–541. patients with resistance to diuretics and hypoalbuminemia.
16. Grande PO, Holbeck S, Persson J. Microvascular fluid exchange. In: Hahn 39. Uhlig C, Silva PL, Deckert S, et al. Albumin versus crystalloid solutions in
RJA, Prough DS, Svensen CH, editors. Perioperative fluid therapy. Michigan: & patients with the acute respiratory distress syndrome: a systematic review and
1st University of Michigan; 2006. pp. 13–26. meta-analysis. Crit Care 2014; 18:R10.
17. Starling EH. On the absorption of fluids from the connective tissue spaces. This is the systematic review on the use of albumin in patients with ARDS in
J Physiol 1896; 19:312–326. comparison with the use of crystalloids.
18. Woodcock TE, Woodcock TM. Revised Starling equation and the glycocalyx 40. Frenette A, Bouchard J, Bernier P, et al. Albumin administration is associated
model of transvascular fluid exchange: an improved paradigm for prescribing && with acute kidney injury in cardiac surgery: a propensity score analysis. Crit
intravenous fluid therapy. Br J Anaesth 2012; 108:384–394. Care 2014; 18:602.
19. Guidet B, Ait-Oufella H. Fluid resuscitation should respect the endothelial This article is on the possible harmful effects of albumin administration in patients
& glycocalyx layer. Crit Care 2014; 18:707. after cardiac surgery, observing an increased risk of acute renal failure.
On the importance of albumin and albumin-containing solutions in maintaining the 41. Schortgen F, Girou E, Deye N, Brochard L. The risk associated with hyper-
integrity of the endothelial glycocalyx. oncotic colloids in patients with shock. Intensive Care Med 2008; 34:2157–
20. Stamler JS, Jaraki O, Osborne J, et al. Nitric oxide circulates in mammalian 2168.
plasma primarily as an S-nitroso adduct of serum albumin. Proc Natl Acad Sci 42. Wiedermann CJ, Dunzendorfer S, Gaioni LU, et al. Hyperoncotic colloids and
U S A 1992; 89:7674–7677. acute kidney injury: a meta-analysis of randomized trials. Crit Care 2010;
21. King TP. On the sulfhydryl group of human plasma albumin. J Biol Chem 1961; 14:R191.
236:C5. 43. Flannery AH, Kane SP, Coz-Yataco AO. A word of caution regarding pro-
22. Reeves RB. Temperature-induced changes in blood acid-base status: Don- & posed benefits of albumin from ALBIOS: a dose of healthy skepticism. Crit
nan rCl and red cell volume. J Appl Physiol 1976; 40:762–767. Care 2014; 18:509.
23. Quinlan GJ, Margarson MP, Mumby S, et al. Administration of albumin to This article is on the possible limitations of the ALBIOS trial.
patients with sepsis syndrome: a possible beneficial role in plasma thiol 44. Caironi P, Gattinoni L. Proposed benefits of albumin from the ALBIOS trial: a
repletion. Clin Sci (Lond) 1998; 95:459–465. & dose of insane belief. Crit Care 2014; 18:510.
24. Caironi P, Tognoni G, Gattinoni L. Albumin replacement in severe sepsis or Further on the potential effects of albumin administration in patients with septic
& septic shock. N Engl J Med 2014; 371:84. shock, based upon the findings observed in the ALBIOS trial.
Further on the rationale of the ALBIOS trial, highlighting the importance of effective 45. Delaney AP, Dan A, McCaffrey J, Finfer S. The role of albumin as a resuscita-
secondary functions of albumin. tion fluid for patients with sepsis: a systematic review and meta-analysis. Crit
25. O’Brien AJ, Fullerton JN, Massey KA, et al. Immunosuppression in acutely Care Med 2011; 39:386–391.
& decompensated cirrhosis is mediated by prostaglandin E2. Nat Med 2014; 46. Cui JY, Xu HL, Wang AT, et al. Influence of albumin as a resuscitation fluid on
20:518–523. the prognosis of patients with sepsis: a meta-analysis. Zhongguo Wei Zhong
Excellent translational investigation showing an immunomodulating effect of albu- Bing Ji Jiu Yi Xue 2012; 24:18–23.
min by inhibiting PGE2-mediated immunosuppression in advanced decompen- 47. Leitch A, Craig G, Sadler P. Human albumin solution resuscitation in severe
sated cirrhosis. sepsis and septic shock. J Intensive Care Soc 2013; 14:45–52.
26. Bortoluzzi A, Ceolotto G, Gola E, et al. Positive cardiac inotropic effect of 48. Wiedermann CJ, Joannidis M. Albumin replacement in severe sepsis or septic
albumin infusion in rodents with cirrhosis and ascites: molecular mechanisms. & shock. N Engl J Med 2014; 371:83.
Hepatology 2013; 57:266–276. This is a letter showing a meta-analytic analysis on the three largest RCTs on the
27. Fencl V, Leith DE. Stewart’s quantitative acid-base chemistry: applications in use of albumin in patients with severe sepsis or septic shock.
biology and medicine. Respir Physiol 1993; 91:1–16. 49. Rochwerg B, Alhazzani W, Sindi A, et al. Fluid resuscitation in sepsis: a
28. Carlesso E, Maiocchi G, Tallarini F, et al. The rule regulating pH changes && systematic review and network meta-analysis. Ann Intern Med 2014;
during crystalloid infusion. Intensive Care Med 2011; 37:461–468. 161:347–355.
29. Langer T, Carlesso E, Protti A, et al. In vivo conditioning of acid-base This is an extensive network meta-analysis on fluid resuscitation, and its different
equilibrium by crystalloid solutions: an experimental study on pigs. Intensive aspects, in patients with sepsis.
Care Med 2012; 38:686–693. 50. Patel A, Laffan MA, Waheed U, Brett SJ. Randomised trials of human albumin
30. Langer T, Ferrari M, Zazzeron L, et al. Effects of intravenous solutions on acid- && for adults with sepsis: systematic review and meta-analysis with trial sequen-
&& base equilibrium: from crystalloids to colloids and blood components. Anaes- tial analysis of all-cause mortality. BMJ 2014; 349:g4561.
thesiol Intensive Ther 2014; 46:350–360. This is the most accurate and extensive meta-analysis on RCTs investigating the
Extensive and thorough review on the effects of intravenous solutions, including effects of human albumin administration in patients with sepsis of any severity,
albumin-containing solutions, on acid-base equilibrium. including a trial sequential analysis.
31. Bernardi M, Maggioli C, Zaccherini G. Human albumin in the management of 51. Jiang L, Jiang S, Zhang M, et al. Albumin versus other fluids for fluid
complications of liver cirrhosis. Crit Care 2012; 16:211. & resuscitation in patients with sepsis: a meta-analysis. PLoS One 2014;
32. Morgan TJ, Vellaichamy M, Cowley DM, et al. Equivalent metabolic acidosis 9:e114666.
with four colloids and saline on ex vivo haemodilution. Anaesth Intensive Care Further meta-analysis on the use of human albumin in patients with sepsis of any
2009; 37:407–414. severity.
33. Bihari S, Prakash S, Bersten AD. Early changes in serum electrolytes and 52. Xu JY, Chen QH, Xie JF, et al. Comparison of the effects of albumin and
& acid-base status with administration of 4% albumin. Intensive Care Med & crystalloid on mortality in adult patients with severe sepsis and septic shock: a
2014; 40:1392–1393. meta-analysis of randomized clinical trials. Crit Care 2014; 18:702.
On the effect of 4% albumin aministration on acid-base equilibrium. Further meta-analysis on the use of human albumin in patients with severe sepsis or
34. Waters JH, Bernstein CA. Dilutional acidosis following hetastarch or albumin septic shock.
in healthy volunteers. Anesthesiology 2000; 93:1184–1187. 53. Farrugia A, Bansal M, Balboni S, et al. Choice of fluids in severe septic
35. McClelland DB. Safety of human albumin as a constituent of biologic patients: a cost-effectiveness analysis informed by recent clinical trials. Rev
therapeutic products. Transfusion 1998; 38:690–699. Recent Clin Trials 2014; 9:21–30.