The Effect of Albumin Administration in Critically Ill Patients: A Retrospective Single-Center Analysis

Abstract

Objectives: The aim of this study was to analyze the development of albumin administration in patients admitted to the adult ICU. In addition, we assessed the impact of albumin administration on serum hemoglobin concentration.

Design: We conducted a retrospective single-center study including all patients who were admitted to the ICU from January 2013 to December 2021 and stayed at least 24 hours.

Setting: The study was conducted in an academic hospital (University Hospital Basel, Switzerland).

Patients: A total of 20,927 admissions were included, of which 3748 received albumin at least once during their ICU stay. To analyze volume expansion, 2006 admissions met the inclusion criteria, namely at least two hemoglobin measurements within 12 hours, one albumin delivery, and experienced no bleeding, dialysis, or transfusions during this period.

Interventions: None.

Measurements: We examined the hemoglobin levels before and after albumin administration and compared them with a matched control group to assess the amount and duration of volume expansion.

Main results: From 2013 to 2021 the proportion of critically ill patients treated with albumin rose from 5.0% to 32.5%. An overproportioned increase in albumin use could be seen in surgical patients (4.7-47.2%) and in those receiving RBC transfusion (13.7-72.6%). In those patients receiving albumin, a significant drop in hemoglobin of around 5 g/L on average could be observed following treatment with albumin.

Conclusion: Hemodilution was observable for at least 12 hours after albumin administration and may have caused a decrease in hemoglobin concentration of greater than 8 g/L when isooncotic albumin solution (5%, 25 g in 500 mL) was administered. This makes albumin, especially in its isooncotic form, an ideal colloid to achieve long-lasting volume expansion. However, RBC transfusions may increase under albumin therapy, as transfusion thresholds may be undershot after albumin administration.

Figure 1.

Yearly development of albumin per subgroup. A, Surgical versus nonsurgical patients showing a trend in more albumin administration among surgical patients. B, Within the surgical fields fewer neurosurgical patients received albumin than their other surgical counterparts. C, More patients with comorbidities such as diabetes, hypoalbuminemia, and kidney failure received albumin; an increase of 20.5% in 2013 to 82.6% in 2021 can be observed in septic shock patients; patients with liver cirrhosis remain at around 40%. D, Patients undergoing RBC transfusion increased from 13.7% to 72.6% within the 9-year study period.

Figure 2.

Yearly development of hydroxyethyl starch (HES), albumin, and crystalloids. Average amount per case of all patients. Although HES has phased out in clinical use, albumin is used in more patients, and on average more albumin is administered per patient per year (5.0–25.8 g). The average amount of crystalloids administered has halved in the study period (6933–3298 mL).

Figure 3.

Hemoglobin (Hb) changes after albumin administration over the course of 12 hours. Isooncotic albumin (12.5 g in 250 mL and 25 g in 500 mL) causes a more significant decrease in hemoglobin compared to the control groups. Additionally, double the amount causes around double the change. Hyperoncotic albumin (20 g in 100 mL) leads to a less significant drop in hemoglobin. For all albumin cases, the hemoglobin decreased around 5 g/L and remained at this level for 12 hours.