Development and Validation of a Structured Telephone Interview for Dementia Assessment (STIDA): The NIMH Genetics Initiative

Abstract

As part of the NIMH Genetics Initiative Alzheimer's Disease (AD) Study Group, a brief structured telephone interview to distinguish individuals with normal cognitive functioning from those with changes in cognition and daily functioning suggestive of early AD was developed. The Structured Telephone Interview for Dementia Assessment (STIDA), yields a dementia score between 0 and 81 (higher scores indicating greater impairment). Subscales corresponding to the subscales of the Clinical Dementia Rating Scale (CDR) can be derived. The STIDA performed well as a screening instrument for mildly demented individuals. When a score of 10 or more (based on informant interview and subject testing) was used to identify mildly impaired individuals, the STIDA had a sensitivity of .93 and a specificity of .92 for a clinician-derived CDR of 0.5 or more. The STIDA was also capable of accurately assessing the level of dementia. STIDA-derived CDR ratings agreed with clinician-derived CDR scores in 23 of 28 cases, corresponding to an unweighted kappa of .71 and a weighted kappa of .81. A much-abbreviated short STIDA that could be administered directly to the subject was able to detect possible impairment with a sensitivity of .93 and a specificity of .77. These results suggest that the short STIDA provides a sensitive and fairly specific telephone screen for dementia, and that the full STIDA, consisting of an interview with a knowledgeable informant and subject testing, approximates the success of a face-to-face clinical interview, and provides reliable and valid screening and staging of dementia over the telephone.