Optimizing prediction of new-baseline glomerular filtration rate after radical nephrectomy: are algorithms really necessary?