There is tremendous interest in determining the patient- and system-level characteristics that are associated with an increased risk of readmission. There have been studies that draw on administrative data to provide large numbers of patients and crucial information about billed diagnoses and services, but these efforts are limited in their ability to examine the role of more specific clinical data. Studies that explored these granular details, including vital sign and lab abnormalities, have, to this point, been small in sample size, limited to one SNF, or included large numbers of patients discharged to home.
We leveraged the rich data on clinical course and comorbidities available through the electronic medical record (EMR) at YNHH to explore the risk factors for both early and late hospital readmissions from SNFs. We are currently completing a regression model to evaluate the risk of readmission for more than 30 possible indicators, including characteristics of hospitalization, discharge vital signs, laboratory values at discharge, measurements of chronic disease burden, and social determinants.