The variability of processing of technetium-99m DTPA renography. Role of interpolative background subtraction

In quantitative renography, observer-dependent selection of renal outlines and background regions may account for considerable variability. This study of differential renal function with Tc-99m DTPA scintigraphy compares interpolative background subtraction with the authors' routine techniques; the latter involves background-subtracted uptake on data integrated over 1-3 minutes after injection. Other techniques considered were omission of background subtraction and use of 1-2 minute integration. The normal range was established in 24 hypertensive patients who had a normal angiogram and a normal radiometric glomerular filtration rate. The test set of 52 other hypertensive patients included 27 with renal artery stenosis. All techniques correlated well with the routine method (r > or = 0.98); however the interpolative background technique was unique in preserving the depiction of renal asymmetry but reducing the variability of replicate measurements (P < 0.05). Reduced renal function increased the variability of routine measurements, but the interpolative background subtraction method performed better in this instance (P < 0.01). Thus, the new technique appears to improve the definition of renal outlines and increase the reliability of measurements of differential renal function.