ABSTRACT
A study was conducted to validate the Inoue protocol in determining the glomerular filtration rate (GFR) of Filipinos. Dynamic posterior planar kidney images of 402 consecutive Filipino patients referred for in vitro GFR determination were reprocessed using the Inoue protocol. Regression and Bland-Altman analyses were done on surface area normalized glomerular filtration rates (GFRSAnorm) generated using the Inoue linear regression model of the sample, Gates' method, and original regression formula published by Inoue, using respective two-point plasma concentration (in vitro) GFRSAnorm values as reference standards. GFRSAnorm results from the three camera-based techniques had strong correlation with those obtained using the in vitro method (i.e. r values of 0.9349, 0.8922 and 0.9349, respectively). However, agreement analysis showed lack of both bias and precision in the results of the Inoue linear regression model of the sample, and presence of bias and lack of precision in the results of both the Gates' method and the original linear regression model published by Inoue when compared to their corresponding in vitro GFRSAnorm results (standard error of 0.6209, 0.8379 and 0.9473, respectively). Thus, the linear regression model of the Inoue protocol is superior to the Gates' method for camera-based GFR estimation, and is population-specific, but is not robust enough to be a replacement for the in vitro technique.
