Multi-infarct dementia (MID) may be the second most common form of dementia in later life. A commonly used aid in the clinical diagnosis of MID is the Hachinski Ischaemic Score (HIS). The usefulness of this score is controversial, and we hypothesized that this is because many items of the HIS are open to a wide range of interpretations. We therefore canvassed 45 research and academic centres in the United Kingdom and Ireland with an interest in dementia to assess the variability of interpretation of the HIS. Five template cases were constructed, in which were embedded items which were felt to be potentially contentious. Fifty-five out of 94 (59%) respondents replied. There was a very wide variation in the scores assigned to each vignette. In only five of 65 items was there complete agreement among replies: in general there was a very large range for each item. Thirty of the items showed less than 90% agreement. The apparent simplicity of the HIS conceals possibilities for ambiguous interpretation of individual items. This is a property common to many 'simple' rating scales. It should not lead to outright rejection of these scales, but rather to a refinement and clarification of the scoring and assessment techniques.
