Characteristics of the recurrently hospitalized adult. An information synthesis

Hospital use in the US is concentrated in the approximately 2% of the population that is repeatedly hospitalized. Using the research integration technique of information synthesis, empirical studies of readmission were assembled and analyzed to determine the demographic and clinical characteristics of the repeatedly hospitalized adult. The authors' synthesis of articles that met preset criteria of relevance and validity indicated that clinical variables show a more consistent association with readmission than demographic variables. Diagnosis, prior use, and disability appear to be strong predictors of an individual's level of hospital use. Demographic variables with consistent positive associations with readmission include being widowed, living with relatives other than a spouse, decreasing income, living in the South, and living in an urban area, but available work does not indicate whether these demographic variables are independent predictors of readmission or merely markers of an increased prevalence of chronic disease. Findings indicate that future investigations of the predictors of this costly pattern of hospital use will be most illuminating if illness-related factors such as diagnosis, disease severity, functional status, and usage history are carefully specified and examined.