Quality assurance in radiation therapy: clinical and physical aspects. Consensus of best current management: the starting point for clinical quality assessment.

Consensus of best current management developed by a rational and deliberative process can provide the basis for clinical quality assessment. Unfortunately, it is not always possible to arrive at a consensus at all cancer sites, and this generally indicates areas where clinical research is needed. Assessing the quality of care in these situations presents special problems. When it is possible to arrive at consensus in a specific disease, this consensus should detail appropriate pretreatment evaluation and the details of the treatment. Committees of experts for each specific disease site can formulate the consensus and must document their decisions based on information from the current world literature. A carefully thought out and documented consensus can then provide the basis for the development of process based questionnaires in assessing quality. We have observed that individuals formulating consensus of best current management do not strictly follow their own criteria, and that compliance in various strata of practice throughout the United States shows a greater deviation from consensus than anticipated and indeed this deviation crosses all types of practice. It was then necessary to conduct outcome surveys in the same patients to validate the processes of care by showing a correlation of process performance with outcome or indeed to change our concepts of best current management. We recognize from these outcome studies that relatively few processes have direct association with outcome and the majority of our consensus points relate to either good general patient management or items important to individual patients but not to large groups of patients. In addition to validating processes through outcome correlations, we have found that process verification is important. We have observed quite different outcomes for two groups of patients with Hodgkin's disease treated with the same processes (i.e., mantle field technology and adequate radiation dose, etc.). We were unable to identify the reason for an increased failure rate in one group of these patients until we looked at each individual mantle port film from the two groups of patients. We then identified that one facility was not including the Hodgkin's disease in the treatment portal due to poor technical performance. We believe that this program of process verification may be important in evaluating quality for any disease site. Data will be presented that illustrates the above problems.