Background: Current guidelines suggest general practitioners should screen their diabetic patients for microalbuminuria. There is a range of possible tests. We looked for studies that compared a timed urine sample (the gold standard) with a random spot sample.
Method: Systematic review and meta analysis of studies comparing albumin to creatinine ratio (ACR) on a random specimen to albumin excretion rate from an overnight or 24 hour timed sample. Studies were identified using Medline and EMBASE to June 2003. Studies were pooled using diagnostic odds ratios and were checked for heterogeneity.
Results: Ten studies covering 1470 patients were included. Use of the ACR in screening 100 diabetic patients would miss only two out of the 20 patients who would be expected to have microalbuminuria, while there would be 13 false positives. A timed specimen would be required to clarify the diagnosis for 31 patients.
Discussion: The marginal benefit of using a timed urine collection over a spot ACR to detect microalbuminuria in the screening of diabetic patients is small, and not worth the cost and inconvenience of collecting a timed sample.