Résumé
OBJECTIVE: To characterize the participation of clinical laboratories in the second biennial of an external quality assurance scheme for urinalysis in Mexico. METHODS: The participants received 32 control samples in 8 shipments of 4 controls each during the biennial 1999-2000. The controls were in the form of tablets to be dissolved in 12 mL of water to mimic urine. The participants performed dipstick assays of pH, specific gravity (SG) and 8 analytes (proteins, glucose, ketones, blood, bilirubin, nitrites, urobilinogen, leukocytes) in the controls. Expected control values were assigned by 10-15 replicate automated assays using dipsticks and apparatus of a single provider (Bayer). Participants' results were transformed to scores ranging from zero (perfect result) to 400 (worst possible). Score differences were analyzed using nonparametric methods (Kruskal-Wallis and Mann-Whitney). RESULTS: The best results were seen for nitrites, ketones and proteins (mean global scores = 3-5) and the worst for SG (score = 45) with intermediate scores of 10-13 for the other six. CONCLUSIONS: The scores improved for the second biennial from a grand mean of 17 in the first significantly down to 12 in the second. The number of participants has been increasing yearly and there has been a decrease of participants who do not report at least 3 out of 4 yearly shipments. We confirmed that automated readings yield better scores than visual readings.