Reply

Abstract

interventions to reduce readmissions in high-risk patients where there will likely be the most yield per dollar spent. In fact, we have used this risk score for an ongoing pilot study wherein patients at high risk for readmission (>75th percentile of risk score) undergo more frequent follow-up visits and targeted intervention.Although we continually search to improve the statistical performance of our risk models in cardiac surgical procedures, we must also continually remind ourselves of the potential clinical and programmatic implications. In this particular case, we believe that the model has enough separation of readmission risk in the low and high ends of the spectrum to make its use meaningful for directing interventions and resources. The results of the pilot study based on this risk score will, it is hoped, yield further insight into the potential utility of this risk stratification tool.