Abstract Introduction Use of albumin is suggested for some patients with shock, but preferences for its use may vary among intensive care unit (ICU) physicians. Methods We conducted an international online survey of ICU physicians with 20 questions about their use of albumin and their opinion towards a randomised trial among adults with shock comparing the use versus no use of albumin. Results A total of 1248 respondents participated, with a mean response rate of 37%, ranging from 18% to 75% across 21 countries. Respondents mainly worked in mixed ICUs and 92% were specialists in intensive care medicine. The reported use of albumin in general shock varied as 18% reported ‘almost never’, 22% ‘rarely’, 34% ‘occasionally’, 22% ‘frequently’ and 4% ‘almost always’ using albumin. In septic shock, 19% reported ‘almost never’, 22% ‘rarely’, 29% ‘occasionally’, 22% ‘frequently’ and 7% ‘almost always’ using albumin. Physicians’ preferences were more consistent for haemorrhagic‐ and cardiogenic shock, with more than 45% reporting ‘almost never’ using albumin. While the reported use of albumin for other purposes than resuscitation was infrequent (40%–85% reported ‘almost never’ for five other indications), the most frequent other indications were low serum albumin levels and improvement of the efficacy of diuretics. Most respondents (93%) would randomise adult ICU patients with shock to a trial of albumin versus no albumin. Conclusions In this international survey, the reported preferences for the use of albumin in adult ICU patients with shock varied considerably among surveyed ICU physicians. The support for a future randomised trial was high.

Preferences for albumin use in adult intensive care unit patients with shock: An international survey