The predictive validity of the consensus criteria for DLB have not yet been fully established. When retrospectively applied to clinical records of autopsy-confirmed cases(21) they had a sensitivity of 0.75 and specificity of 0.79. Inter-rater reliability was acceptable for all key symptoms with the exception of fluctuation (k = 0.25-0.36). In the first prospective validation study(9) the consensus criteria show a tendency to clinically underdiagnose DLB (sensitivity 0.83); false-positive diagnoses are uncommon (specificity 0.95). Although it is possible within a specialist centre to diagnose DLB with an accuracy similar to that for AD or Parkinson's disease, further clarification of the criteria would assist in their routine clinical use, particularly for the items relating to fluctuating cognition.
Was this article helpful?