What accuracy is needed in dosimetry

An attempt is made to formulate requirements for the accuracy in the delivery of absorbed dose to a patient during radiation therapy. These requirements are mainly based on the relative steepness and separation of the dose-effect curves for local tumour control and normal tissue damage. All the clinical data which were reviewed, including some results from the ''Patterns of Care Study'', indicate a close correlation between outcome of therapy (control rates, complications), and dose level, inaccuracy or errors in dosimetry and patient-machine positioning. This has been reported for external beam therapy as well as for brachy- and interstitial therapy. Only the clinical results will allow us to select the optimal treatment conditions (e.g. selection and definition of the planning target volumes, dose levels and beam arrangement) but they could be interpreted correctly only to the extent that the treatment execution would be correct. This gives its full credit to the Quality Assurance Programmes, in which the clinicians and physicists should be fully involved. Lastly, the outcome of a treatment can only be interpreted meaningfully if the parameters of the irradiation, in particular, the distribution of dose in space and time can be accurately correlated with the type and extent of the disease. It is essential that clear, well defined and unambiguous concepts and parameters be used for reporting purposes to ensure a common language between collaborating centres. This is one of the major tasks that the ICRU has undertaken since several years. 48 refs, 5 figs, 17 tabs