Objective: To review literature identifying key components for measuring avoidable readmissions, their prevalence, risk factors, and interventions that can reduce potentially avoidable readmissions.

Data sources and extraction: Literature search using Medline, PubMed and the Cochrane Library up to June 2010, using the terms "avoidable", "preventable", "unplanned", "unnecessary", "readmission", and "rehospitalization".

Study selection: A total of 48 original papers and review articles were selected for inclusion in this review.

Data synthesis: Although hospital readmission seemed to be a term commonly used as an outcome indicator in many studies, it is difficult to make valid comparison of results from different studies. This is because the definitions of terms, methods of data collection, and approaches to data analysis differ greatly. The following criteria for studying hospital readmissions have been recommended: (a) identify hospital admissions and define relevant terms, (b) establish a clinical diagnosis for a readmission; (c) establish the purpose for a readmission, (d) set a discharge-to-readmission timeframe, and (e) identify the sources of information for assessing readmissions. Studies to identify avoidable readmissions usually involve medical records and chart reviews by clinicians using the classification scheme developed by the authors. The proportion of all readmissions assessed as preventable varies from 9 to 59% depending on the population of patients studied, duration of follow-up, type and methodology of the study and case-mix-related factors. A number of studies classified risk factors for readmission into four categories: patient, social, clinical, and system factors. Home-based interventions, intensive education/counselling, multidisciplinary care approaches, and telephone follow-up were the main types of interventions to address potentially avoidable readmissions.

Conclusions: A standard instrument to identify avoidable readmission is important in enabling valid comparisons within the system and at different timelines, so as to permit robust evaluation of interventions. The assessment of preventable risk factors for readmissions also provides a basis for designing and implementing intervention programmes.