Jornal Brasileiro de Patologia e Medicina Laboratorial (Dec 2007)
Avaliação de diferentes métodos imunoturbidimétricos para determinação de albumina urinária: impacto na classificação dos estágios da nefropatia diabética Evaluation of different immunoturbidimetric methods to measure urinary albumin: impact in the classification of diabetic nephropathy stages
Abstract
INTRODUÇÃO: A nefropatia diabética (ND) acomete até 40% dos pacientes diabéticos e o diagnóstico precoce pode evitar a evolução para estágios avançados. O rastreamento deve ser realizado pela medida de albumina urinária (AlbU) utilizando-se o método quantitativo sensível. OBJETIVO: Avaliar diferentes métodos imunoturbidimétricos de determinação de AlbU para a classificação dos estágios da ND. MATERIAL E MÉTODO: A albumina foi dosada em 167 urinas (65 urinas de 24 h e 102 amostras casuais) por dois métodos imunoturbidimétricos: kit Aptec-BioSys, ADVIA® 1650 Bayer (AlbUAdvia) e kit MAlb Urin-Pack Bayer®, CobasMira® Roche(AlbUCobas). AlbUCobas foi definido como método comparativo e utilizado para classificar as amostras em normoalbuminúricas (albuminúria 300 mg/24 h ou > 174 mg/l, n = 31). Os coeficientes de variação (CV) intra e interensaio, sensibilidade e linearidade dos métodos foram calculados. As concordâncias analítica e diagnóstica foram analisadas por regressão Deming, gráficos de Bland-Altman e por coeficiente kappa. RESULTADOS: Os CVs intra e interensaio foram BACKGROUND: Diabetic nephropathy (DN) affects up to 40% of diabetic patients and must be screened by the measurement of urinary albumin with a sensitive quantitative method. OBJECTIVE: To evaluate the impact of different immunoturbidimetric methods to measure albuminuria in the classification of DN stages. MATERIAL AND METHOD: Albumin was measured in 167 urine samples (65 24 h samples and 102 casual samples) by two immunoturbidimetric methods: Aptec BioSys, ADVIA® 1650 Bayer (AlbUAdvia) and Malb Urin-Pack Bayer®, CobasMira® Roche (AlbUCobas). AlbUCobas was definedas the comparative method used to classify the samples in: normoalbuminuric (albuminuria 300 mg/24 h or > 174 mg/l; n = 31). The intra and interassay coefficients of variation (CVs), sensitivity and linearity of each method were calculated and the analytical and diagnostic agreements were analyzed by Deming's regression and Bland-Altman plots and by Kappa coefficient, respectively. RESULTS: The intra and interassay CVs were < 5% for both methods. The sensitivity was 5 mg/l for both methods and the linearity was 160 mg/l for AlbUCobas and 200 mg/l for AlbUAdvia. The analytical agreement between the two methods was satisfactory (mean differences between methods = -7.68 mg/l (-0.21 - 15.2); r = 0.989; p < 0.001) and the kappa coefficient (0,914; p < 0,001) was excellent. Only nine urine samples (5,4%) were in disagreement with AlbUCobas classification: six microalbuminuric samples were misclassified as normoalbuminuric (n = 3) and macroalbuminuric (n = 3) by AlbUAdvia. Three normoalbuminuric samples were misclassified as microalbuminuric by AlbUAdvia. The albumin values of these samples were in the highest diagnostic adopted cut-off point for each DN stage. CONCLUSION: The immunoturbidimetric methods analyzed may be interchangeable, without significant misclassification of the different stages of DN.
Keywords
