147-LB: Targeted Intervention for At-Risk Patients with Diabetes to Reduce Readmissions

With benefits to patient outcomes, provider burden, and health system costs, the goal of reducing hospital readmissions for patients with diabetes mellitus (DM) is multifaceted. Current studies have neither elucidated a clear approach to accurately identify at-risk patients nor implemented reliable interventions to significantly reduce readmissions. We trialed a Quality Improvement project utilizing a diabetes Nurse Practitioner (NP)/Certified Diabetes Care and Education Specialist (CDCES) to see at-risk patients during initial hospitalization and in close clinic follow-up after discharge with an endpoint of reducing future readmissions. Over 21 months, 79 hospitalized patients with DM were identified as high-risk for readmission. Identifying factors included previous admission, DKA/HHS, new DM diagnosis, new to insulin, high-dose steroid use, or provider discretion. The NP/CDCES provided DM management, education, and coordinated discharge. Patients were called shortly after discharge and had a 2-4 week clinic follow-up with the NP/CDCES. The primary endpoint was number of readmissions 180 days post-discharge after NP/CDCES intervention versus number of previous readmissions pre-intervention. Change in hemoglobin A1C, time to post-hospital phone call, and time to outpatient follow-up were also evaluated. Pre- and post-intervention readmission rates remained unchanged; average pre-intervention readmissions per patient was 1.7 versus 1.88 post-intervention readmissions (p=0.74). However, a secondary endpoint of reduction in A1C was statistically significant, with a mean decrease in A1C of 0.97% per patient (p=0.0025). Time to post-hospital phone call and early follow-up did not affect readmission rates. While targeted inpatient-to-outpatient intervention for high-risk patients with DM did not reduce readmission rates, our data suggests coordinated inpatient intervention with close hospital follow-up may improve A1C in high-risk patients with diabetes. Disclosure A. Harris: None. K. Grdinovac: None.