Comparison of three scales to assess health risk perception

Background: Health risks management consists of quantitative and qualitativ e assessment of risks including risk perception among different samples of the population. Little work has been done to develop and validate scales to measure risk perception. Methods: We conducted, in December, 1999, a study among 1358 French GPs, me mbers of the Sentinels network, in order to compare three scales: a visual analog scale, a verbal scale and a numerical scale. GPs were asked about th eir own perception of two risks: the Creutzfeldt-Jakob disease new variant (vMCJ) and the bug. Results: The response rate was 55%, with no difference between the three gr oups (p=0.85). No statistically significant difference was observed between the distributions of the visual analog scale and the numerical scale (p=0. 11 for the question about the vMCJ and p=0.98 for the question about the bu g). Conversely, distributions of the verbal scale were significantly differ ent from those of the visual analog scale (p<0.0001 for both of the questio ns) and from those of the numerical scale (P<0.0001 for both of the questio ns). Separation between worried and non worried people didn't occur in the middle of the visual analog scale but at 33 millimeters from the left extre mity for the question about the vMCJ and at 41 millimeters from the same ex tremity for the question about the bug. Conclusion: We recommend the use of verbal scales to measure instantaneous perception of a given risk. Visual analog scales and numerical scales are k nown to be the best scales to detect minimum changes in the perception of f unctional signs such as pain. On this purpose, their superiority with regar d to verbal scales has to be confirmed in the field of risk perception.