the VISs (criticized by some as being too simplistic and not inclusive enough) and the vaccine package inserts, which may have too much technical information for some people to understand and process effectively. To bridge the gap, it was suggested that vaccine risk communicators consider preparing intermediate messages about vaccine risks and benefits that have more detailed information than the VISs but that are less technical than the package inserts. The package inserts might also be reorganized so as to be more accessible to consumers.
Another possibility is an interactive or computerized information system, in which users can access the level of information that is appropriate for their needs and abilities. Such a system was developed by researchers at Ohio State University and the Mayo Clinic (Raman et al., 1996). The system was arranged in a hierarchical fashion with five levels of increasingly complex information, all written at an eighth-grade reading level. The system was tested in a middle-class medical practice, and although 13 percent of the parents declined to participate, there was great variability in the level of information requested by those parents who did participate. Those who agreed to use the system expressed a high degree of satisfaction with it (Ramen et al., 1996).
Several speakers and audience participants suggested that vaccine risk communications include references to the scientific literature, so that readers who question the information presented or who wish to know more can examine the original source for each statement. It was also suggested that all the evidence regarding a specific topic be presented, including information that does not support the position of the person or organization conveying the information.
As discussed previously, many participants noted that there is often uncertainty about estimates of the risks of vaccines. This uncertainty reflects the generally low risks of serious adverse consequences; if the risks were high, the product would not have been licensed. To promote effective risk communication, several speakers and participants felt that risk messages should acknowledge the uncertainty about the existence and magnitude of many of the potential risks associated with vaccines, with the inclusion of estimates of the incidence of the better-defined risks to the extent possible. A risk communications expert added that "in almost every policy forum that I can think of, the ultimate conclusion has been that it is essential for good public policy to communicate uncertainty, even if it is difficult." She suggested that vaccine risk communications include statements regarding what assumptions were made to define risk estimates and whether there is consensus among various groups of experts and the public about the accuracy or uncertainty of the estimates provided.
Other participants suggested that it be emphasized in vaccine risk communications that research is under way to improve understanding of those risks. A CDC representative suggested that much of the problem is inherent in the gaps in knowledge identified in previous IOM reports (1991, 1994a, 1994c), and