Self-Administered and Interviewer-Administered Instruments for Dementia Research

Objective: To compare data produced with instruments used to obtain information from caregivers of demented older adults, when the instruments are self-administered at home or administered by an interviewer. Design: The functional status of patients with Alzheimer's disease was determined at baseline and six months later with three instruments using the patients' caregivers as informants. The instruments were both self-administered and administered by trained interviewers. Three other instruments were administered by interviewers only. Setting: A geriatric clinic in an urban university teaching hospital and residences of the patients. Participants: Thirty-four Alzheimer's patients and their caregivers completed the study. Measurements: Three instruments were compared: an activities of daily living instrument, a sleep problem instrument, and a dysfunctional behavior instrument. Validity was determined by Pearson correlations and agreement with intraclass correlations (ICC) on base-line assessments. Instrument responsiveness was detcrmined by t-test comparisons of pre-post change scores. Results: Correlations between instruments were consistent with predictions, and agreement between versions of the instruments as determined by ICC ranged from 0.71 to 0.89. The instrument used for the measurement of activities of daily living was more responsive to change over the six months when self-administered compared to interviewer-administered (p = 0.007). A reduction in sample size requirements was obtained as a result of the increased responsiveness of the self-administered version of the Lawton. Conclusion: The results suggest that these self-administered instruments when completed at home by caregivers are valid, in agreement with and may be more responsive than interviewer-administered instruments.