Plasma volume measurement in septic patients using an albumin dilution technique: comparison with the standard radio-labelled albumin method.

OBJECTIVE: To investigate a technique using 20% albumin for measurement of plasma volume in critically ill patients. DESIGN AND
SETTING: Laboratory and clinical investigation in the adult intensive care unit and anaesthetic laboratories of a university hospital. PATIENTS: 12 patients fulfilling ACCP/SCCM criteria for septic shock. INTERVENTIONS AND MEASUREMENTS: Each patient received (125)I-labelled albumin, and the volume of distribution was measured at 1 and 10 min. The accepted standard plasma volume measurement (98% of the 10-min volume of distribution) was calculated. Immediately thereafter 200 ml 20% human albumin was given. Albumin concentrations were measured before and 1 min after this 40-g bolus, and the volume of distribution calculated using a formula that corrected for the 200 ml fluid in which the albumin was dissolved.
RESULTS: Plasma volumes measured using the albumin dilution technique at 1 min were smaller than the standard technique by 110+/-280 ml; limits of agreement were from -660 to +440 ml (-16% to +11%). Plasma volumes measured by (125)I-albumin at 1 min were smaller than the standard by 120+/-110 ml; limits of agreement were from -330 to +100 ml (-8 to +2%).
CONCLUSIONS: Non-labelled albumin can be used easily and quickly to measure a plasma volume in ICU patients and gives a moderately accurate estimate when compared with the (125)I-labelled albumin methods. The normal isotope method over-estimates plasma volumes in septic patients because excessive transcapillary escape of albumin is inadequately compensated for by the standard correction factor.