Nurse judgments of hallucinated voice descriptions: relevance for intervention.

The aim of the research was to explore the importance of 58 items of descriptive content represented on two parallel forms of an Inventory of Voice Experiences (IVE). A convenience sample of 317 well-educated psychiatric nurses rated descriptions of IVE item content three times over a period of six months according to how the content might relate to the selection of interventions for the management of verbal auditory hallucinations. Cronbach's alpha, Cohen's kappa, Pearson's r, and Bartko's intra-class correlation coefficients were used to measure the internal consistency of the nurses' judgments as well as the concordance of the judgments with a pre-selected standard. Findings from the study revealed modest-to-moderate support for internal consistency and overall equivalence of parallel item content represented on both forms of the IVE within and across three waves of data collection. Also revealed was a relative lack of concordance of the nurses' judgments with the pre-selected standard, and modest-but-consistent concordance of the nurses' original and subsequent judgments. Eight parallel items represented on the IVE demonstrated potential to serve as important cues for making decisions about intervention. This information shall be used to standardize the language and response categories of items tied to the IVE to permit more-definitive decisions about the management of verbal auditory hallucinations.