Hospital characteristics associated with unplanned readmissions.

The rate of unplanned readmission of patients to hospitals has been introduced into the Australian Council on Healthcare Standards accreditation program as an internal flag of problems in patient management and outcome. An emphasis, in the indicator definition, is placed on the unexpected nature of the admissions to exclude those which are unplanned but simply due to progression of a disease, and are therefore not 'unexpected'. The association of hospital characteristics with unplanned readmissions was examined using logistic regression on the data collected from hospitals surveyed in 1993. The risk of unplanned readmission was significantly higher in public hospitals than in private hospitals. Hospital bed-size also reflected differences in the risk of unplanned readmission, being significantly higher for hospitals with over 200 beds than for those with 1-100 beds. In rural areas, the risk of unplanned readmission was significantly lower in hospitals with 101-200 beds and over 200 beds compared to hospitals with 1-100 beds (p for trend = .004). However, in metropolitan areas, the risk of unplanned readmission increased with the size of the hospitals (p for trend < .0001). Monitoring of unplanned readmissions prompted internal clinical review and action in 31 per cent of hospitals, demonstrating the indicator's usefulness as an internal quality tool. However, the use of unplanned readmissions as an external performance measure must take into account a hospital's characteristics and will remain of limited value in the absence of clinical information about the expected or unexpected nature of the readmissions.