A semi‐structured interview to assess visual hallucinations in older people

Authors:
Publisher website
Google Scholar
Abstract
Objective: Visual hallucinations are under‐reported by patients and are often undiscovered by health professionals. There is no gold standard available to assess hallucinations. Our objective was to develop a reliable, valid, semi‐structured interview for identifying and assessing visual hallucinations in older people with eye disease and cognitive impairment.Methods: We piloted the North‐East Visual Hallucinations Interview (NEVHI) in 80 older people with visual and/or cognitive impairment (patient group) and 34 older people without known risks of hallucinations (control group). The informants of 11 patients were interviewed separately. We established face validity, content validity, criterion validity, inter‐rater agreement and the internal consistency of the NEVHI, and assessed the factor structure for questions evaluating emotions, cognitions, and behaviours associated with hallucinations.Results: Recurrent visual hallucinations were common in the patient group (68.8%) and absent in controls (0%). The criterion, face and content validities were good and the internal consistency of screening questions for hallucinations was high (Cronbach alpha: 0.71). The inter‐rater agreements for simple and complex hallucinations were good (Kappa 0.72 and 0.83, respectively). Four factors associated with experiencing hallucinations (perceived control, pleasantness, distress and awareness) were identified and explained a total variance of 73%. Informants gave more ‘don't know answers’ than patients throughout the interview (p = 0.008), especially to questions evaluating cognitions and emotions associated with hallucinations (p = 0.02).Conclusions: NEVHI is a comprehensive assessment tool, helpful to identify the presence of visual hallucinations and to quantify cognitions, emotions and behaviours associated with hallucinations. Copyright © 2008 John Wiley & Sons, Ltd.

This publication has 32 references indexed in Scilit: