, the cutoff points recommended by guidelines and the optimal cutoff points found in this study) are presented in Figure 1a for men and Figure 1b for women. This shows that the fractional area under the ROC curve is 0.94 (95% confidence interval 0.90 to 0.97) for men and 0.98 (95% confidence Ganetespib order interval 0.95 to 1.00) for women. This high area under the ROC curve reflects the fact that the gain in sensitivity obtained by changing the cutoff level is associated with a relatively modest loss of specificity. Figure 1. (A) ROC curve for UACR predicting an albuminuria ��30 mg/d in men and (B) ��30 mg/d in women. *Cutoff point usually recommended by guidelines (15�C19). **Cutoff point recommended as an alternative by some guidelines (14,17,18). ***Optimal … Discussion Microalbuminuria is typically defined as a UAER �� 30 mg/d.

The study presented here shows that the commonly used cutoff of 30 mg/g for ACR as a test for the detection of microalbuminuria is not optimal because of its low sensitivity in the renal transplant recipient population. Data presented here suggest that in this population, lower and gender-specific cutoffs should be adopted for the detection of microalbuminuria. The performance of the optimal cutoff points determined in this study (21 mg/g for men and 24 mg/g for women) and the performance of the cutoffs proposed by Warram et al. (23) and adopted by some guidelines (17 mg/g for men and 25 mg/g for women) were analyzed in the study presented here. Their performances were found to be very close.

Microalbuminuria predicts the development of chronic kidney disease (1,2,3,12), cardiovascular disease (3�C7,10,11), and the risk of death (2,8,9) in the diabetic and nondiabetic population as well as the risk of graft loss and death in renal transplant recipients (13). The important prognostic significance of microalbuminuria and the therapeutic implications associated with its presence led to the adoption of guidelines recommending the measurement of UAER in high-risk populations using UACR. However, there is no consensus among the guidelines about the definition of a ��normal UACR.�� Some guidelines have adopted 30 mg/g of creatinine as the upper normal limit (15�C19) or lower gender-associated cutoffs (14,17,18,20), whereas others recommend the use of either (17,18). The diversity of these recommendations reflects the uncertainty regarding the optimal definition of UACR.

The study presented here shows that a UACR cutoff of 30 mg/g is not optimal in the renal transplant population. These results are in line with those obtained by other investigators (21,22,24,25) in the general and diabetic population. Connell et al. (24) first showed that the accuracy of UACR as a test to estimate the UAER was gender dependent and suggested that different cutoff values should be defined for men and women. Data presented by Warram et al. (23) and Mattix et al. (25) supported this Batimastat approach.