Clinical implications of collimator exchange effect, relative collimator and phantom scatter

It is well established that the beam fluence on the central axis for rectangular or blocked fields is not always the same as that of its collimator equivalent square field. Many centers still determine monitor unit settings using a single output factor based on the collimator equivalent square field size because they feel the small errors introduced by this approach do not justify the extra data and effort required to correctly determine the monitor unit settings. Analyzing clinical monitor unit determinations for 6 and 15 MV photon irradiation. it is shown that failure to account for these discrepancies in central axis beam fluence may introduce 2.5-3% error in clinical situations. While the error is generally small, it must be eliminated if we intend to reach the commonly stated goal of 5% overall uncertainty in dose determination.