Predicting hospital readmissions in the Medicare population.

摘要: Using a nationally random sample of Medicare beneficiaries, we developed multivariate logistic model that identified series factors predict readmission to an acute care hospital within 60 days discharge. Our regression results show 10 variables are statistically significant predictors readmissions, with the patient's disease history and diagnosis among best predictors. can be used by peer review organizations evaluate quality provided different hospitals as well physicians social workers improve discharge planning.