Detection of errors in individual patients in radiotherapy by systematic in vivo dosimetry.

We report 5 years of systematic measurements of the dose delivered to each patient undergoing radiotherapy treatment with photon beams in order to detect any systematic error that may have escaped the different checks performed at each step of planning and calculation prior to the first treatment session, or may have arisen during the set-up or the treatment delivery. For each patient the target-absorbed dose is derived from the entrance and exit doses measured by silicon diodes, on the beam axis at the patient's skin. Depending on the discrepancies observed between the measured and expected doses we have set decision levels for the corrective actions to be taken. In addition these measurements allow us to obtain information on the overall accuracy or on the quality of a specific treatment. During 5 years, 7519 patients have been measured and 79 errors were detected. Half could have induced a variation of over 10% in the dose delivered. Seventy-eight out of 79 errors were of human origin. As part of an overall quality assurance programme, it is of the utmost importance to check the dose delivered for each patient undergoing radiotherapy treatment in order to avoid systematic underdosing or overdosing.