Skip to main content

Predicting Hospital Readmission

December 09, 2013

Researchers identified a means of predicting hospital readmission among patients within 30 days of their discharge, a new study found.

In order to reduce the number of readmissions into the hospital, caregivers first need to know who is most at risk. To do this, researchers integrated an automated prediction system into the existing electronic health records system. At-risk patients were red-flagged.

The study, both retrospective and prospective, analyzed all adult patients admitted to a healthcare system comprised of 3 hospitals between August 2009 and September 2012.

Their results showed that the past is the greatest predictor of the future. After analyzing the retrospective data, they identified a single risk factor, “more than 2 inpatient admissions in the past 12 months,” had the best balance of sensitivity (40%), positive predictive value (31%), and proportion of patients flagged (18%).

After implementing the program and examining the 12-month period that followed, the results nearly mirrored the original. Using the same single risk factor of "more than 2 inpatient admissions." sensitivity was 39%, positive predictive value 30%, and proportion of patients flagged 18%.

The study was published in the Journal of Hospital Medicine.


--Stephanie Vaccaro



Baillie CA, VanZandbergen C, Tait G, et al. The readmission risk flag: Using the electronic health record to automatically identify patients at risk for 30-day readmission. J Hosp Med. 2013 Nov 13 [Epub ahead of print].

Back to Top