The factors advanced as casual agents of acute confusion span the full spectrum of human illnesses and are of physiologic, psychologic, sociologic, and environmental origins. Although it is generally agreed that acute confusion is the result of multiple factors interacting over time, most investigations into the causes of confusion, however, have utilized univariate, single-subject, or single-time designs. But, if the assumption is to be accepted that multiple interacting agents cause confusion, then studies must be of multivariate, multidimensional, and longitudinal designs. Thus, the purpose of the proposed study is to test a three-wave, longitudinal, causal model of confusion. The three waves, or time periods to be examined are: Time I--Admission to the hospital, Time 2--Onset of acute confusion; and Time 3--Resolution of acute confusion. To test the hypothesized causal/model, 1050 medically-ill patients 65 years of age or older will be interviewed and hospital records reviewed daily while hospitalized. These semi-structured interviews and their hospital record reviews will be used to obtain information about cerebral metabolism, neural and sensory structure and function, social dynamics, affective status, and aspects of the environment; all of which are factors considered casually related to acute confusion. These data will be analyzed within Levine's Conservation Model of Nursing in a three-wave longitudinal casual model using linear structural relationship modeling, or LISREL. Information necessary for developing appropriate and effective strategies for the nursing care relative to acute confusion (specifically detection, prevention, prediction, and intervention) of this patient population will be derived from the resultant model.
