Decision analysis assessment of a national medical study.

Abstract

To decide whether or not to undertake an expensive national survey to determine the effectiveness of infection control, we devised a quantitative decision model to analyze the costs and probabilities of successful study outcomes. The result allowed us to determine whether the proposed study method and design would provide sufficient statistical power to ensure meaningful conclusions from the research. The model was robust in assessing the adequacy of method accuracy and, within the range of assumptions specified, it suggested that the project should be undertaken. The results helped to secure official approval and funding for this large-scale research project. A novel approach to evaluating sensitivity analysis is included. As constructed, the model is applicable to other projects in applied research and, with some modification, to projects in basic research as well.