Trailblazers: how albumin was proven to be safe


Albumin, along with hemoglobin and fibrin, were among the first proteins to be identified in the human body. Considering the putative benefits including maintenance of colloid osmotic pressure and more efficient filling of the intravascular compartment, it evolved into an effective resuscitation fluid. Widespread use followed the initial introduction into clinical practice in the face of exigencies of resuscitation encountered on the battlefield. However, questions were soon raised regarding its safety, which culminated in the landmark SAFE randomized controlled trial (RCT). Apart from answering key clinical questions, this trial proved to be a significant milestone in the history of critical care medicine and raised the bar for future clinical trials. 

Early history: albumin in the battlefield

During the first world war, it was clear that most battlefield casualties occurred due to uncontrolled hemorrhage. By the spring of 1940, as yet another world war loomed on the horizon, rapid replacement of hemorrhagic loss with appropriate substitutes assumed top priority. Edwin Cohn, a biochemist at Harvard, used a special technique of fractionation to separate albumin from the plasma, allowing storage for an extended period (1). On the early morning of December 7, 1941, Japanese bombers attacked Pearl Harbor inflicting heavy casualties on American troops. Isidor Ravdin, a surgeon, was summoned by the US Government to Hawaii to evaluate the wounded soldiers at Pearl Harbor. He brought along with him an inventory of the albumin solution developed previously by Edwin Cohn (Fig. 1). This solution was initially used on seven sailors who suffered severe burn injury. The effect of albumin administration appeared dramatic. Burn victims who were barely conscious started speaking, and the generalized edema gradually resolved. All seven of the albumin-treated sailors survived (2). 

Figure 1. Left: Edwin Cohn, Right: Isidor Ravdin

The human albumin program and early studies 

Early success with albumin as a resuscitation fluid in trauma victims and burns led to the “human albumin program” in the US (3). Following the initial experience among battlefield casualties, albumin use became widespread in civilian hospitals. Over the years, refinements in the manufacturing process resulted in a purer and safer product (4). 

Skillmann et al. reported the first RCT on the clinical use of albumin in 1975 (5). Sixteen patients who underwent abdominal vascular reconstructive surgery were randomized to an intraoperative fluid regimen of albumin or crystalloid. Patients randomized to the crystalloid regime were infused with a much larger volume; however, the postoperative plasma volume was lower in this group compared to those treated with albumin. Postoperative serum albumin levels and serum oncotic pressure were also higher with albumin use. The authors concluded that plasma volume expansion was much more efficient with the use of albumin without causing leakage into the extracellular fluid compartment. 

1998: The contentious Cochrane meta-analysis 

In 1998, the Cochrane Injuries Group Albumin Reviewers (CIGAR) performed a meta-analysis that kindled considerable controversy among the medical fraternity and the lay press (6). The authors identified 30 RCTs that evaluated the use of albumin in critically ill patients. A total of 1419 patients were included comparing albumin with a control group that received no albumin or underwent crystalloid administration. The studies evaluated patients with hypovolemia, burns, or hypoalbuminemia. The authors observed an increased risk of mortality with albumin use in each of these categories. The relative risk of mortality with albumin use was 1.46 (0.97–2.22) for hypovolemia, 2.40 (1.11 to 5.19) for burns, and 1.69 (1.07 to 2.67) for hypoalbuminemia. On a fixed effects model, the pooled risk of death was 6%, suggesting one additional death for every 17 patients receiving albumin. 

The CIGAR meta-analysis raised a storm, with The Times (London) suggesting that up to 30,000 deaths in the UK may be attributable to albumin administration. Following this meta-analysis, there was a steep fall in the use of albumin in the UK by approximately 40–45% (7). This was followed by the Expert Working Party of the Committee of Safety of Medicines recommendation for an urgent multicenter RCT to evaluate the safety of albumin administration (8).

The findings of the CIGAR meta-analysis were criticized for the omission of relevant trials, the heterogenous groups of patients included, and failure to take into account the methodological quality of the studies analyzed (9). Three years later, the findings of another meta-analysis revealed contrasting findings (10). Wilkes et al. analyzed 55 RCTs, including 3504 patients with data available on mortality. The relative risk for mortality could be evaluated in 42 trials. The pooled relative risk for mortality was not significantly different in patients who received albumin compared to those who did not. This meta-analysis appeared to support the safety of albumin in critically ill patients. However, there was a compelling need for a multicenter RCT to evaluate the safety and efficacy of albumin in critical care. 

The landmark trial: is albumin SAFE?  

In the middle of the controversy surrounding albumin, the Australian and New Zealand Intensive Care Society Clinical Trials Group conducted one of the first large, multicenter RCTs in critical care medicine. The study included patients who required fluid administration according to clinician judgment. A total of 3497 patients were randomized to receive 4% albumin and 3500 to receive normal saline (11). 

The volume of the study fluid administered was significantly less with 4% albumin compared to normal saline on the first three days of the study. By day 4, the volume of saline administered was 1.4 times that of albumin. Thus, the SAFE study contradicted the historically held belief of the equivalence of three times as much volume of crystalloid compared with colloid during resuscitation. The 28-day mortality was not significantly different between the albumin and the saline groups (20.9% vs. 21.1%, relative risk: 0.99 [0.91–1.09], p = 0.87). Besides, no difference in survival times was observed between the two groups. The length of stay in ICU and hospital was similar between groups. The duration of mechanical ventilation and renal replacement therapy were similar. The incidence of new-onset organ failure was also comparable.  

On subgroup analysis, the authors compared the 28-day survival with albumin compared to saline among patients with trauma, severe sepsis, and acute respiratory distress syndrome. Among the other noteworthy findings, the relative risk of death was higher in trauma patients who received albumin. The difference was entirely due to the higher mortality among patients with traumatic brain injury who received albumin compared with normal saline (24.5% vs. 15.1%, p = 0.009). In patients with severe sepsis, the relative risk of death at 28 days was lower with albumin compared with saline, although the difference fell just short of statistical significance. There was no difference in the 28-day survival between the albumin and the saline groups in patients with acute respiratory distress syndrome.  

The findings of the SAFE trial contrasted with those of the CIGAR meta-analysis. The authors concluded that albumin and saline are clinically equivalent as resuscitation fluid in a heterogenous group of critically ill patients. Although subgroup analysis suggested favorable outcomes with saline in traumatic brain injury, and with albumin in severe sepsis, these findings required further investigation before any conclusions could be drawn. 

In the accompanying editorial, Deborah Cook commented that the SAFE trial “heralded a new era in critical care” with intensivists following the footsteps of cardiologists who had embraced and popularized the randomized trial as a robust methodology to answer key clinical questions. The meticulously conducted trial represented a milestone in the history of critical care medicine (12).

Post hoc analyses of the SAFE trial 

A subsequent post hoc follow-up study was performed on patients with traumatic brain injury who were enrolled in the SAFE trial (13). Overall, the 2-year mortality was significantly higher in patients treated with albumin compared to saline (33.2% vs. 20.4%, p = 0.003). The results were similar in patients with severe traumatic brain injury, with a score of 3–8 on the Glasgow Coma Scale (GCS). The mortality in the albumin group was 41.8% compared with a significantly lower 22.2% in the saline group. In patients with less severe injury, with a GCS of 9–12, the mortality was not significantly different between the two groups. The findings were similar on adjusted analysis using a multivariate logistic regression model. 

In another pre-defined subgroup analysis, the SAFE investigators analyzed the outcomes of  1218 patients with severe sepsis who received albumin compared with saline (14). Among these, 603 were assigned to the albumin group and 615 to the saline group. The total Sequential Organ Failure Assessment (SOFA) score was similar between the two groups; the number of patients who underwent renal replacement therapy was also similar. The overall relative risk of 28-day mortality was similar between the two groups. However, on adjusted analysis using a  multivariate logistic regression model, mortality was lower with albumin (odds ratio: 0.71; 95% CI: 0.52–0.97; p = 0.03). 

Subsequent evidence

The ALBIOS trial evaluated the impact of albumin replacement in patients with severe sepsis or septic shock (15). Patients in the albumin group received 300 ml of 20% albumin daily until 28 days or ICU discharge, targeting a serum albumin level of 3 mg/dl. Crystalloids were administered to this group based on clinician judgment. The control group received crystalloids alone during the study period. The 28-day mortality did not differ significantly between the two groups. On post-hoc analysis, patients with septic shock had a lower risk of mortality with albumin compared to crystalloids alone. 

A subsequent meta-analysis by Patel et al., including 4190 patients from 16 RCTs evaluated clinical outcomes with albumin administration in critically ill patients. In critically ill patients with sepsis, the authors did not observe any impact of albumin on all-cause mortality, although there was no signal towards possible harm. 


Human albumin has been in use as intravenous fluid therapy since the 1940s. A major impetus for its use occurred during the second world war, in the face of battlefield casualties who suffered massive hemorrhage. There was a strong physiological rationale behind the use of albumin as resuscitation fluid, considering its favorable effect on the colloid osmotic pressure of the plasma, and maintenance of the vascular barrier function, thus reducing capillary leak. Over the decades since the second world war, human albumin was extensively used in critically ill patients with no robust evidence to support improved clinical outcomes. The 1998 Cochrane meta-analysis proved to be a turning point, suggesting the possibility of harm with albumin use across heterogenous groups of critically ill patients. The stage was thus set for one of the largest RCTs in the relatively short history of the specialty of critical care medicine. The trial asked a simple question – is 4% albumin safe compared with normal saline as resuscitation fluid? Clinical outcomes were comparable between the two types of fluid in a heterogenous group of ICU patients. The study also suggested the likelihood of worse outcomes with albumin use in patients with traumatic brain injury. A trend towards benefit was observed in patients with severe sepsis. The SAFE study proved that large, multicentric clinical trials offer the most efficacious methodology to settle contentious treatment modalities in critically ill patients. The study turned out to be the harbinger of future large RCTs addressing key questions in critical care medicine.  


1. Matejtschuk P, Dash CH, Gascoigne EW. Production of human albumin solution: a continually developing colloid. Br J Anaesth. 2000 Dec;85(6):887–95. 

2. Peters T. 1 – Historical Perspective. In: Peters T, editor. All About Albumin [Internet]. San Diego: Academic Press; 1995 [cited 2022 Dec 4]. p. 1–8. Available from:

3. HUMAN ALBUMIN IN MILITARY MEDICINE. J Am Med Assoc. 1942 Nov 28;120(13):1041–2. 

4. Nicholson JP, Wolmarans MR, Park GR. The role of albumin in critical illness. Br J Anaesth. 2000 Oct;85(4):599–610. 

5. Skillman JJ, Restall DS, Salzman EW. Randomized trial of albumin vs. electrolyte solutions during abdominal aortic operations. Surgery. 1975 Sep;78(3):291–303. 

6. Cochrane Injuries Group Albumin Reviewers. Human albumin administration in critically ill patients: systematic review of randomised controlled trials. BMJ. 1998 Jul 25;317(7153):235–40. 

7. Roberts I, Edwards P, McLelland B. More on albumin. Use of human albumin in UK fell substantially when systematic review was published. BMJ. 1999 May 1;318(7192):1214–5. 

8. Vincent JL, Russell JA, Jacob M, Martin G, Guidet B, Wernerman J, et al. Albumin administration in the acutely ill: what is new and where next? Crit Care. 2014;18(4):231. 

9. Horsey PJ. The Cochrane 1998 Albumin Review–not all it was cracked up to be. Eur J Anaesthesiol. 2002 Oct;19(10):701–4. 

10. Wilkes MM, Navickis RJ. Patient Survival after Human Albumin Administration: A Meta-Analysis of Randomized, Controlled Trials. Ann Intern Med. 2001 Aug 7;135(3):149. 

11. A Comparison of Albumin and Saline for Fluid Resuscitation in the Intensive Care Unit. N Engl J Med. 2004;10. 

12. Cook D. Is albumin safe? N Engl J Med. 2004 May 27;350(22):2294–6. 

13. Saline or Albumin for Fluid Resuscitation in Patients with Traumatic Brain Injury. N Engl J Med. 2007;11. 

14. SAFE Study Investigators, Finfer S, McEvoy S, Bellomo R, McArthur C, Myburgh J, et al. Impact of albumin compared to saline on organ function and mortality of patients with severe sepsis. Intensive Care Med. 2011 Jan;37(1):86–96. 

15. Caironi P, Tognoni G, Masson S, Fumagalli R, Pesenti A, Romero M, et al. Albumin Replacement in Patients with Severe Sepsis or Septic Shock. N Engl J Med. 2014 Apr 10;370(15):1412–21. 

Top 35 critical care blogs of 2022:

Leave a Reply