Evaluating Which Heart Failure Patients are Likely to be Readmitted

VHHA provides predictive statistics from the inpatient database to determine composite “pictures” of patients likely to return. The predictive models are derived from the diagnosis codes listed for each patient that met CMS’s criteria for readmission. VHHA includes the diagnoses in the readmission dashboards it makes available to members. Next to each diagnosis is the odds ratio (OR), which compares the magnitude of readmission risk associated with the diagnosis:

  • OR=1 The diagnosis does not affect the odds of readmission
  • OR>1 The diagnosis is associated with higher odds of readmission
  • OR<1 The diagnosis is associated with lower odds of readmission

Figure 1 (chart below) compares the 2015 heart failure readmission model to the 2012 model. The five diagnoses with the highest ORs are listed for both. There are three key findings. First, the models are very similar. Readmitted patients with an index admission for heart failure are likely to have renal failure, arrhythmias, fluid imbalances, cardio-respiratory failure and severe hematologic disorders. Second, these diagnoses depict patients with three or more failing body systems. They include cardiac, renal, respiratory and blood/circulatory systems. Third, the model has been stable over three years, though the top five diagnoses are even stronger predictors of readmissions now than in 2012. Overall, the strongest predictor of heart failure readmission is a patient presenting with three or more failing body systems. (7/31/2015)

Comparison of Top Five Diagnosis Odds Ratios 2012 and 2015