The availability of constantly more sophisticated algorithms and metho
ds for dose planning (denoted 'level 3' procedures by the ICRU) makes
it possible to use accordingly more complex treatment techniques. Such
procedures have the potential of reducing irradiation of organs at ri
sk and other healthy tissue. However, they suffer from a substantially
more difficult set-up of the patient and verification of the treatmen
t. This will introduce additional sources of error, systematic as well
as random, which will be of importance for dose reporting. The purpos
e of this paper is to point out some of these problems and to suggest
a method for reporting which is the least sensitive to these 'new' sou
rces of error.