Implementation of a Perioperative Insulin Protocol in an Advanced Practice Provider (APP)–Led Limb Salvage Hospitalist Team

Abstract

We implemented a standardized perioperative insulin management protocol (PIP) for patients with hyperglycemia and acute diabetic foot problems. Our multidisciplinary quality improvement (QI) project retrospectively compared outcomes in the APP-led hospitalist team (TU) using the PIP vs. other hospitalist teams (OT) managing similar patients. We chose glargine U100 as the basal insulin to simplify preoperative dose adjustments. There were no statistical differences in either arm with respect to mean age (53-57 years), female distribution (26-27%), mean weight (85-94 kg) and race/ethnicity (Hispanic 52-62%, Black 34-21% and non-Hispanic white 11-17%). Patients in TU (n=44) and OT (n=42) were started on basal/bolus insulin 80 v 74% of the time, respectively. Initial basal insulin doses (units) were similar (24 ± 12 v 26 ± 24), while initial prandial insulin amounts were higher in TU than OT (21 ± 10 v 13 ± 12; p<0.01), respectively. Use of insulin types were consistent with PIP protocol and hospital formulary guidelines, with TU using glargine and aspart in 89% and 82% of patients, while OT used NPH and Regular in 69% and 76% of patients, respectively. Mean admission A1C (9.1±2.3 v 9.2±2.2 %) and capillary blood glucose (CBG in mg/dl) (239±117 v 257±123) were similar for TU and OT. Mean CBG during hospitalization was lower for TU (163 ± 31 v 187 ± 44; p<0.01). Pre-op (145 ± 49 v 153 ± 50; p=0.48) and immediately post-op (153 ± 74 v 162 ± 81; p=0.59) CBGs were similar for TU and OT, respectively. Hypoglycemia (CBG<70) occurred in 32% of TU and 19% of OT (p=0.18) with only one CBG<50 recorded in each arm. There were no differences in ER visits or readmissions within 30 days of discharge; there was a trend for shorter hospital LOS in TU (14±8 v 18±9 days; p=0.08). A standardized PIP improved perioperative glycemic control without increasing hypoglycemia risk. Wider implementation of this protocol through incorporation into electronic order sets will be the next step in our QI initiative. Disclosure K. Odedosu: None. K.K. OConnell: None. M.C. Jordan: Employee; Spouse/Partner; Accenture. U. Gunasekaran: None. P.E. Fox: None. J.N. McNulty: None. S. Harder: None. L. Meneghini: Advisory Panel; Self; Novo Nordisk Inc., Sanofi US. Consultant; Self; Sanofi US, Novo Nordisk Inc.. Advisory Panel; Self; Intarcia Therapeutics, Inc.. Other Relationship; Self; American Diabetes Association.