The ICRU Report 91 recommends, inter alia, Paddick’s dose gradient index (
GI) for reporting the values of dose gradients in stereotactic radiotherapy. Nevertheless, the
GI shows false positive characteristics on a decreasing physical dose gradient
∙
D. One aim of ICRU Report 91 is to better associate treatment complications with the values of dose gradient indices. In this context, the
GI is not suitable. Therefore, the author developed an advanced dose gradient index (
aGI) to get rid of the deficiencies of the
GI. A function of the volume product of the isodoses of interest was used to define the
aGI instead of the volume ratio that defines the
GI. The dose gradient distributions were quantified by the superficially averaged dose gradient (
SADG). The values of the
SADG,
aGI, and
GI were determined for the linac-based stereotactic radiosurgery and radiotherapy of 13 brain metastases and 25 choroidal melanomas, respectively. The
aGI was proportional to Δ
D/
SADG with true characteristics in both irradiation series; Δ
D = const.
is the dose difference of the isodoses of interest. Pearson’s correlation coefficients were
r ≥ 0.647. Contrary to the
aGI, the
GI showed false positive regression lines with
r ≤ -0.511. The growing
aGI on a decreasing |
SADG| is reasonable because the
aGI is nearly a reciprocal dose gradient measure in the form of a radius difference Δ
r = Δ
D/||
·
D||. The utilisation of the
aGI entails no limitations of the comparability of dose gradient values. The
GI and other dose gradient indices based on volume ratios of the isodoses of interest should no longer be used for reporting dose gradient values.