Verification films: a study of the daily and weekly reproducibility of breast patient set-up in the START trial.

AIMS: Verification of patient position in radiotherapy by the use of portal images is a current practice in most radiotherapy departments. All patients within the Standardisation of Breast Radiotherapy Trial (START) were required to have on-treatment verification images. Measurements of maximum lung and breast depth from these were used to provide details for the case-report forms. This study was undertaken to validate these measurements, assess the reproducibility of patients' set-up within the START trial and to compare reproducibility on a day-to-day basis with reproducibility over a longer period. Analysis also included a subjective assessment of image quality.
MATERIALS AND METHODS: Fifty-three patients from 29 departments were studied. Some centres used electronic portal imaging and others used film. Where film verification was the method of choice, information was collected about the type of film and cassette used.
RESULTS: The correlation coefficients for inter- and intra-observer variability of breast and lung depth were 0.99. Comparing maximum breast depth on weekly and daily images, the median per cent standard deviation of breast depth was 4% and 3%, respectively. The mean standard deviation of breast depth for all patients, which gives an indication of the individual patient variability, was 5.5 +/- 2.2 mm for weekly and 3.9 +/- 1.5 mm for daily measurements. For lung depth, mean standard deviation was 6.8 +/- 2.5 mm for weekly and 6.8 +/- 2.8 mm for daily measurements. Images taken using Kodak X-omat V film were inferior to those taken with films in corresponding cassettes.
CONCLUSION: Differences between lung depth measured in simulator and on the treatment unit were evident for many patients. There was little difference in the standard deviations of breast and lung depth for daily and weekly verification films on the treatment machine.