Assessing urinary albumin excretion in pre-eclamptic women: which sample to use?


* Dr H. Kieler, Department of Medical Epidemiology, Karolinska Institutet, PO Box 281, SE-171 77 Stockholm, Sweden.


Objective To evaluate whether the gold standard of 24-hour urine collection for measuring albumin excretion in pre-eclamptic women could be substituted by shorter collection periods.

Design Prospective study.

Setting Fetal maternity ward, university hospital.

Participants Thirty women with pre-eclampsia and a positive urinary test strip for protein of at least 2+.

Methods From each woman, within a 25-hour period, three spot, two 12-hour (day and night) and one 24-hour urine sample were collected. Urine albumin concentrations in milligrammes per litre were analysed by rate nephelometry on a Beckman Array protein system. The urinary albumin concentrations in the spot and the 12-hour samples were compared with the concentration in the 24-hour urine collection.

Main outcome measures Urinary albumin concentrations in spot and 12-hour samples measured against the standard 24-hour albumin excretion.

Results Albumin concentrations in the day and night collection fitted closely with the concentrations of the 24-hour collection. The median difference between the 24-hour and the day collection was −3 mg/L (interquartile range −264 to 116 mg/L). The median difference between the 24-hour and the night collection was 17 mg/L (interquartile range −186 to 210 mg/L). The association of urinary albumin concentration in the 24-hour collection and the spot samples was much weaker. Of the spot urine samples, the albumin concentration in the sample taken on the morning after admission to hospital was closest to the 24-hour urinary albumin excretion, with a median difference of −62 mg/L (interquartile range −1131 to 285 mg/L).

Conclusion The gold standard of 24-hour urinary excretion for assessment of albuminuria in pre-eclamptic women can be substituted with a 12-hour collection. Spot urine samples were inaccurate and are therefore not recommended for quantification of albumin excretion.