Dynamic hepatic scintigraphy: the effect of using high administered activity on reproducibility of hepatic perfusion index value.

The hepatic perfusion index (HPI) may be of value in the detection of both overt and occult hepatic metastases in patients with colorectal carcinoma. Using the standard technique, the low counts obtained in each region of interest can produce statistical unreliability in the curve generation and render many studies unsuitable for analysis. In this study the effects of a high administered activity (400 MBq) on the inter- and intra-observer reproducibility of dynamic hepatic scintigraphy was investigated in 18 consecutive patients with colorectal cancer. All 18 patient studies were suitable for analysis by two observers. The correlation coefficient of measurements between observers was 0.98 (P less than 0.000 01 linear regression analysis), with a root mean square difference of replicate HPI measurements between observers of 0.045. The median range of HPI for four reprocessings by one observer was 0.035, with a maximum range of 0.11. Although high administered activity improves count statistics, for the individual patient study the activity used is not the major source of uncertainty in the derived HPI value.