Reducing avoidable hospital readmissions has been a focus of value-based care. The New England Journal of Medicine found that from 2007 to 2015, risk-adjusted rates of readmission for targeted conditions declined from 21.5% to 17.8%. Given that readmission are stressful for patients and adds significant cost to the healthcare system, providers have been looking for tools that identify when to intervene to attempt to prevent an unnecessary readmission.
This month Remedy introduced the Readmission Risk Predictor as a new feature in Episode ConnectTM. It automates a previously manual process, and identifies a manageable set of high risk patients and improves the accuracy of the risk score.
The Readmission Risk Predictor automatically calculates a risk score for every patient based on data elements in the patient’s Episode Connect record, including socioeconomic factors as well as DRG codes, previous ER visits, admitting physician specialty and other clinical inputs. The Readmission Risk Predictor then calculates a risk score using a proprietary machine learning algorithm based on insights derived from analyzing a dataset of more than 270,000 admissions.
The new Readmission Risk Predictor calculates the probability of 90-day readmission to a degree of accuracy higher than 75%. The predictive algorithm was tested on 90,000 unique admissions and fine-tuned before release.
Healthcare providers cannot practically apply every readmissions risk intervention to every patient, and the patient history available upon admission alone is not sufficient to identify high risk patients. With this new tool from Remedy, providers can see where best to focus their resources.