Determination of serum albumin in older adults. A need for specification of the assay method.

OBJECTIVES: To assess the difference between two colorimetric methods for serum albumin determination, bromocresol purple (BCP) and bromcresol green (BCG), in patients hospitalized in a geriatric hospital. DESIGN: Prospective study of consecutive blood samples. SETTING: A 200-bed Geriatric Division of a university-affiliated geriatric-psychiatric hospital. PARTICIPANTS AND MEASUREMENTS: All serum albumin determinations were obtained from patients hospitalized in the geriatric division during a 3-month period, November 1999 through January 2000. All albumin determinations were performed by two methods: BCP and BCG. RESULTS: A total of 326 serum albumin determinations were performed during the study period. The average serum albumin levels observed were 28.2+/-7.5 g/L and 37.6+/-7.2 g/L for the BCP and BCG methods, respectively. Using the BCP method, 80.1% of serum albumin levels were below 35.0 g/L (cutoff point for hypoalbuminemia) versus 41.1% using the BCG method. CONCLUSIONS: The widespread use of both BCG and BCP and the marked clinical implications of their alternative use underscores the importance of specifying the method employed in determination of serum albumin in routine laboratory analyses. Moreover, physicians need to be aware of which procedure is being used in their hospital.