Human albumin solutions are obtained from healthy donor plasma after pasteurization at 60°C for 10h, and are used in a broad range of clinical and surgical illnesses, though only some of their indications are authorized. In other cases, the use of albumin solutions must be decided on an individualized basis.1–14

Although many studies have examined the effects of albumin solutions upon mortality in critical patients, the conclusions have been contradictory1. A metaanalysis carried out in 1998 with 30 randomized studies, and two Cochrane reviews,3,4 suggested the use of albumin to be associated to increased mortality among critical patients. Posteriorly, another metaanalysis published in the year 2001,5 as well as more recent studies,6–11 have failed to confirm such an increase in mortality, except in patients with severe traumatic brain injury (TBI).12 Furthermore, the antioxidant actions of albumin and its effects upon maintenance of the plasma redox state must be taken into account for the adequate designing of future clinical studies.15

Albumin solutions are more expensive than other colloids and crystalloids. In addition, although such solutions are usually well tolerated, they are not without risks. In effect, albumin solutions can cause allergic reactions, and very rapid infusion can induce sudden hypotension and, in some cases, heart failure–particularly when used at high concentrations. Moreover, although albumin solutions are blood products that are considered to be safe in terms of the transmission of infections, there are doubts regarding the possibility of prion transmission through such solutions.16

The limited availability of albumin solutions, the risk of adverse effects, and their high cost mean that such solutions must be used only in situations where they are known to be effective. However, the level of evidence on which the recommendations are based is limited, and this is probably reflected in daily clinical practice by a broad range of interventions in which such solutions are used among critical care professionals. With the purpose of assessing the clinical practice in Intensive Care Units (ICUs), the Transfusions and Blood Products and Metabolism and Nutrition work groups of the Spanish Society of Intensive Care and Coronary Units (Sociedad Española de Medicina Intensiva y Unidades Coronarias, SEMICYUC) have conducted a survey to evaluate the knowledge and behaviors referred to the use of albumin solutions in ICUs.

A survey was developed on the use of albumin solutions in critical patients. The Transfusions and Blood Products and Metabolism and Nutrition work groups of the SEMICYUC designed the questionnaire in 2012, and invited Spanish and South American ICUs belonging to both public and private centers to participate via e-mail. The professionals who received the questionnaire were personal contacts of the authors, no preference being assigned to members of the work groups of the SEMICYUC. Within each ICU, only a critical care professional answered the questionnaire.

Fifty-seven ICUs were invited, and all agreed to answer the questionnaire (Appendix A).

The submitted questionnaire and results obtained are shown in Table 1. The questionnaire consisted of 35 questions, most with multiple and closed answers, and some with open answers. The questions were grouped into different sections covering data referred to the responders and their work centers, albumin use in different disease conditions; and aspects referred to clinical practice/evidence. The aim of the questionnaire was to record the following aspects: general characteristics of the ICU and of the professionals (specialty, years of experience, and work setting: type of hospital center, number of hospital and ICU beds, and annual admissions to the ICU), usual clinical practice referred to the use of albumin solutions in different disease conditions (nutritional-metabolic, hypovolemia, septic shock/sepsis, and hepatorenal disorders), degree of knowledge of the available scientific evidence, type of albumin solutions used, and cost.

The survey was carried out before the publication of the latest update on the Surviving Sepsis Campaign.17

The data are expressed as frequencies and percentages in the case of qualitative variables, and as means and standard deviations in the case of quantitative variables.

A total of 57 questionnaires from 54 Spanish ICUs and three South American ICUs (Brazil, Chile and Venezuela) were analyzed. All the invited ICUs answered the questionnaire. In turn, all the items of the questionnaire were answered, except three questionnaires that failed to answer the section referred to clinical practice/evidence, and one questionnaire that failed to answer the sections referred to hypovolemia and sepsis/septic shock.

The number of completed questionnaires constitutes a large sample reflecting routine practice in Spain, though it might not be fully representative of the true situation, since the ICUs were not selected on a random basis. Nevertheless, the answers were very homogeneous in practically all the analyzed areas, and this in turn suggests that the survey effectively reflects routine practice regarding the use of albumin solutions. The South American sample is of course too small to allow us to draw such conclusions.

It would be interesting to repeat this study some years after its publication, with a larger number of ICUs, in order to determine whether the recent studies17,18 on the utilization of albumin in septic patients lead to changes in the way in which albumin is used.

The questionnaires were mainly answered by experienced intensivists working in medium to large centers and with a considerable number of ICU admissions each year. In this setting, albumin solutions were used sporadically or were not used in critical patients. This observation persisted in the different sections of the questionnaire, with the sole exception of the hepatorenal section. Thus, albumin was little used or not used at all in denutrition, hypoalbuminemia, hypovolemia and sepsis/septic shock. This is consistent with the general recommendations on albumin use. Table 2 shows the level of evidence referred to the use of albumin solutions in the different sections of the questionnaire.

The recently published consensus document of the European Society of Intensive Care on the use of colloids in the critical patient contemplates the possibility of administering albumin solutions in septic patients, though their efficacy has not been firmly established.18 The more recent review of the management guides of the Surviving Sepsis Campaign recommend the administration of albumin in patients with septic shock who require large amounts of crystalloids, with level of evidence 2C according to the Grading of Recommendations Assessment, Development and Evaluation (GRADE).17 This document has been published after the present survey, and therefore could not be taken into account by the responders of the questionnaire. In any case, the evidence on the effectiveness of albumin in sepsis lacks the quality needed to allow its recommendation in severe sepsis/septic shock.19

The indications of albumin are supported by a greater level of evidence in the case of hepatorenal disease.20,21 This is also reflected by the findings of our study. In fact, most of the responders used albumin solutions in patients with liver disease and hypoalbuminemia, following evacuating paracentesis or in hepatorenal syndrome. In deciding to use albumin, the professionals mainly considered albuminemia, diuresis and the volume of ascitic fluid collected at paracentesis. Indeed, the volume of ascitic fluid collected at paracentesis establishes the indication of albumin administration. Cirrhotic patients with spontaneous bacterial peritonitis or type I hepatorenal syndrome could also benefit from the administration of albumin solutions, though there are no other indications for albumin use in these patients.19,20 However, although there is no scientific evidence warranting the indication of albumin associated to furosemide in patients with anasarca, the results of our study show that almost one-half of the responders used this combination.

Plasmapheresis was available in a large percentage of the centers in our study, and in over one-half of the cases albumin solutions were used to replace the extracted plasma. Based on the available evidence, the use of albumin during plasmapheresis involving the extraction of over 20ml of plasma/kg body weight in a single session, or 20ml/kg/week in several consecutive sessions, is considered adequate. When the volumes are smaller, the indication of albumin should be established on an individualized basis.16,22 In our survey we requested no information on the indications of plasmapheresis with albumin in relation to the needs referred to volume exchange. However, over 28% of the responders claimed that plasmapheresis sessions were always or almost always performed with albumin. We do not know whether there was precise knowledge of the technique on the part of the responders in each center.

Most of the responders considered that albumin use complies with the current recommendations referred to critical patients without liver disease.18 Furthermore, most of the responders were of the opinion that albumin has no positive impact upon the patients, and indeed could even have a negative effect. The available literature does not offer conclusive proof that albumin reduces the mortality rate among patients with hypovolemia, burns and hypoproteinemia, although there is also no proof that it might increase the mortality risk. Consequently, the scant use of albumin in these patients is consistent with the scientific evidence. The degree of dissatisfaction on the part of the professionals regarding the scientific literature supports the need to conduct well-designed studies with homogeneous populations and adequate sample sizes.

Although the studies conducted to date have not confirmed the positive or negative effect which albumin use might have, in the subgroup of patients with severe traumatic brain injury (TBI) of the SAFE study,12 the administration of albumin was associated to greater mortality than resuscitation performed with physiological saline solution (OR 1.8; 95%CI 1.31–2.70; p<0.001). In our survey we did not ask about the management of patients with TBI, though considering the scant use of albumin in the resuscitation of critical patients, similar considerations probably could also be extrapolated to severe TBI.

In critical patients with liver disease, the levels of evidence referred to albumin use are higher.20,21 This is related to the fact that in this case there are more studies with an adequate design and with more homogeneous populations. In this regard, our study shows that in the case of critical patients with liver disease, the evidence was more widely known among the responders and guided albumin use more often. The belief that at least in some cases albumin use could have a positive effect was more prevalent in the case of these patients, and the belief that it might have a negative impact was less prevalent.

The need for more studies with adequate designs, the low levels of evidence referred to albumin use in most critical situations, and the lack of knowledge of the effects of albumin upon mortality undoubtedly condition the habits regarding albumin use and which are still found in the routine practice of some ICUs, despite the absence of adequate degrees of recommendation.

In order to obtain more information on routine clinical practice, we asked about the albumin vials commonly used in the ICUs. In this respect, the vials were found to contain 50ml of albumin solution at a concentration of 5 or 20%. In general, the available presentations and concentrations of albumin were known to the professionals. Few of the responders in our study used concentrations not found on the market.

Regarding knowledge of the cost of albumin use, the questionnaire asked about the price of the vials. The price of a 50-ml vial containing 20% albumin ranges from 37.45 to 44.87 euros (updated prices in Medimecum 201023). A large percentage of the responders did not know the cost of albumin, and only 15.8% offered an adequate price estimate of between 30 and 50 euros. The cost of the albumin solutions therefore does not seem to be particularly important in deciding to use such products, since most of the professionals were unaware of the price of the vials.

Among the substitutes of albumin solutions, a large percentage of the responders (43.9%) cited colloids (starches, dextrans and gelatins) as an alternative. Colloids are used in routine clinical practice mainly for resuscitation in hypovolemia (absolute or relative), although there is controversy regarding the type of fluid indicated in such cases, and at present colloids are not considered to be better than crystalloids. This is another issue requiring further clarification.18

The use of human albumin solutions is not recognized as a common practice among physicians attending critical patients, except in the case of subjects with liver disease. Knowledge of the available scientific evidence is reported to be widespread among these physicians. In general, the opinions on the use of albumin solutions are consistent with the established recommendations, particularly in patients with liver disease, and their utilization in the ICU is guided by the available evidence. Professional experience conditions the decision to use albumin–the mentioned recommendations prevailing only among a minority of professionals.

The authors declare that they have no conflicts of interest.