How the implementation of an in-vivo dosimetry protocol improved the dose delivery accuracy in head and neck radiotherapy.

Doses were measured in-vivo at the entrance using semiconductor detectors for patients with head and neck tumors. Regular measurements started on January 1st and continued till June 30th, 2001. Then the evaluation of the discrepancies between the measured and calculated doses was made, and it resulted in the changes in the protocol of in-vivo dosimetry made effective by February 1st, 2002. The collection of the measurements was censored by January 30th 2003. The number of patients in the two groups was 285 (1st) and 407 (2nd), respectively. The results exhibited the not-Gaussian distribution of the measured doses in both groups. The average number of dose checks per patient increased from 4.9 to 6.0 (1st vs. 2nd group). The mean relative difference between the measured and calculated doses was: -1.5% vs. -0.5%, whereas the standard deviation (1 SD) decreased from 6.1% to 5.6%. The Mann-Whitney U test detected a significant difference between the 1st and the 2nd group (p=0.00000), which justified the conclusion that the changes implemented in the protocol improved dose delivery accuracy and reproducibility of irradiation.