Keynote on lessons from major radiation accidents

Generic lessons have been learned from a relatively large number of accidents in the most relevant practices (a set of analysis have been made on about 90 radiotherapy events, 43 industrial radiography and nine from industrial irradiations); more specific lessons have been drawn from in-depth investigations of individual accidents. The body of knowledge is grouped as follows: a) radiotherapy is very unique in that humans (patients) are purposely given very high radiation doses (20-75 Gy) by placing them in the radiation beam or by placing radioactive sources in contact with tissues. Intended deterministic effects are the essence of the normal radiotherapy practice and relatively small deviation from the intended doses, i.e,, slightly higher or lower than intended may cause increased rate of severe complication or reduce probability of cure. Consequences of major accidents have been devastating, affecting tens, even hundreds of patients and causing death (directly or indirectly) to a large number of them; b) accidents involving industrial radiography are the most frequent cause of overexposure to workers (radiographers); c) accidents with industrial irradiators have lower probability of occurrence, however, they are deemed to be fatal, especially when whole body exposure to panoramic gamma irradiators occur; partial body irradiation from industrial or research accelerator beams has led to amputation of hands and legs; d) when control of sources was relinquished ('orphan' sources) this has resulted in severe injuries, in some cases death and widespread contamination of the environment. A tool for further dissemination of lessons will be an international reporting system of unusual radiation events (RADEV), being introduced world-wide. Accidents were rarely due to a single human error or isolated equipment failure. In most cases there was a combination of elements such as: a) unawareness of the potential for an accident, b) poor education, which usually did not include identifying and facing unusual events, c) an unbalanced striving for resuming or finishing work, which led to ignoring warnings and was often tolerated by management, d) poor maintenance programme or no programme at all leading to poor safety conditions, and non-investigated false alarms leading to distrust in wanting systems. This combination points primarily to an overall managerial failure. Licensing appraisals and inspections should be able to identify the degree of awareness of the management, its commitment reflected in written policy, procedures and supervision. Performance indicators should address these managerial arrangements and aim at identifying whether alertness, due thought and sense of responsibility are praised. Virtually each accident started with a degradation of alertness leading to deterioration of safety, much earlier than the event itself, and virtually all accidents could have been prevented. Preventive measures drawn from lessons learned from reported accidents should be complemented with a more comprehensive approach to identify other potential events that never occurred or were not reported. Prospective assessment of potential exposure with radiation sources, as initiated in ICRP publication 76 will serve to disclose other possible scenarios. This approach has been further pursued by the IAEA and systematic studies are being done for industrial irradiators, industrial radiography, radiotherapy and orphan sources. (author)